Is the Postcode Address File honestly worth £487m? The Royal Mail would like us to believe so. CC-licensed photo by Stuart Orford on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 10 links for you. Double helping, lucky you. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Apple study reveals critical flaws in AI’s logical reasoning abilities • MacRumors

Hartley Charlton:

»

Apple’s AI research team has uncovered significant weaknesses in the reasoning abilities of large language models, according to a newly published study.

The study, published on arXiv, outlines Apple’s evaluation of a range of leading language models, including those from OpenAI, Meta, and other prominent developers, to determine how well these models could handle mathematical reasoning tasks. The findings reveal that even slight changes in the phrasing of questions can cause major discrepancies in model performance that can undermine their reliability in scenarios requiring logical consistency.

Apple draws attention to a persistent problem in language models: their reliance on pattern matching rather than genuine logical reasoning. In several tests, the researchers demonstrated that adding irrelevant information to a question—details that should not affect the mathematical outcome—can lead to vastly different answers from the models.

One example given in the paper involves a simple math problem asking how many kiwis a person collected over several days. When irrelevant details about the size of some kiwis were introduced, models such as OpenAI’s o1 and Meta’s Llama incorrectly adjusted the final total, despite the extra information having no bearing on the solution: “We found no evidence of formal reasoning in language models. Their behavior is better explained by sophisticated pattern matching—so fragile, in fact, that changing names can alter results by ~10%.”

This fragility in reasoning prompted the researchers to conclude that the models do not use real logic to solve problems but instead rely on sophisticated pattern recognition learned during training.

«

These things can’t reason! It’s worth saying again – they’re stochastic parrots. I used ChatGPT the other day because I wanted a list of weather conditions, and separately a list of types of music groups (eg quartet, band, etc). But I’d never trust or expect them to do reasoning about content. People won’t trust the media, with content created by humans who’ve worked at refining their processes for years, but they’ll trust a machine? People are strange.
unique link to this extract

Chinese scientists use quantum computers to crack military-grade encryption — quantum attack poses a “real and substantial threat” to RSA and AES • Tom’s Hardware

Mark Tyson:

»

Chinese researchers claim to have uncovered a “real and substantial threat” to the classical cryptography widely used in banking and the military sectors. According to a report published by the South China Morning Post, the researchers utilized a D-Wave quantum computer to mount the first successful quantum attack on widely used cryptographic algorithms. These algorithms, classed as substitution–permutation network (SPN) cryptographic algorithms, are at the heart of widely used standards like the Rivest-Shamir-Adleman (RSA) and Advanced Encryption Standard (AES).

The Chinese-language research paper is titled Quantum Annealing Public Key Cryptographic Attack Algorithm Based on D-Wave Advantage. The paper outlines how two technical approaches grounded in the quantum annealing algorithm can be used to challenge classical RSA cryptographic security.

The first attack route is “entirely based on D-Wave computers,” explains the paper. It coaxes the Canadian quantum computer into a cryptographic attack by presenting the combination of an optimization problem and exponential space search problem to the computer. The issues are solved using the Ising and QUBO models.

The second proposed attack incorporates classical computing-based cryptographic technology, such as the Schnorr signature algorithm and the Babai rounding technique, layered with a quantum annealing algorithm, to work “beyond the reach of traditional computing methods.”Applying the above techniques, with the help of the D-Wave quantum computer, the team led by Wang Chao of Shanghai University claim to have successfully breached the widely used SPN structure.

«

This feels important, but it’s also pretty impenetrable if you haven’t followed the quantum computing field. And I haven’t.
unique link to this extract

Taylor Lorenz’s plan to dance on legacy media’s grave • The New Yorker

Kyle Chayka:

»

In 2024, [former Verge, Business Insider, Daily Dot, The Atlantic, Washington Post, New York Times journalist Taylor] Lorenz told me, she no longer sees a reason to remain associated with the mainstream media. “I don’t need it for credibility,” she said. “I don’t need it to reach an audience. I don’t know what it does other than connote prestige for a shrinking amount of people.” She added, leaning into exactly the sort of rivalrous drama that plays well online, “Legacy media sucks, it’s crumbling, and, by the way, I’m going to dance on the grave of a lot of these places.”

In some ways, Lorenz’s decision feels belated. Around 2020, a wave of high-profile journalists left traditional outlets to take lucrative deals to launch newsletters at Substack. Lorenz told me that Substack offered her a deal back then, but she turned it down because she felt she needed the imprimatur of an institution to “get more eyes on my work” and persuade people to “take it seriously.” In the years since, the insurgent creator economy has tempted more journalists away to run upstart operations, and consumers have grown increasingly accustomed to paying piecemeal for access to individual writers.

Johnny Harris, formerly a video producer at Vox, developed his own documentary YouTube channel that now has nearly six million subscribers and covers subjects ranging from the criminal investigations of Donald Trump to the threat of China invading Taiwan. In August, the video producer Becca Farsace left the Verge to commit full time to her own YouTube channel, citing the fact that her old employers said she was not guaranteed the rights to content on her social-media channels. “It made me feel like the Verge owned me,” she said in a launch video for her channel, which hosts gadget reviews and now has more than a hundred thousand subscribers. Matthew Yglesias, who decamped to Substack from Vox (a site he co-founded) in 2020, says he has accrued roughly eight-thousand paid subscribers, and, according to Business Insider, he is “likely grossing at least $1.4 million a year.”

«

Like all these systems, a few people are absolutely raking it in, and many, many more are absolutely not.
unique link to this extract

Invisible text that AI chatbots understand and humans can’t? Yep, it’s a thing • Ars Technica

Dan Goodin:

»

What if there was a way to sneak malicious instructions into Claude, Copilot, or other top-name AI chatbots and get confidential data out of them by using characters large language models can recognize and their human users can’t? As it turns out, there was—and in some cases still is.

The invisible characters, the result of a quirk in the Unicode text encoding standard, create an ideal covert channel that can make it easier for attackers to conceal malicious payloads fed into an LLM. The hidden text can similarly obfuscate the exfiltration of passwords, financial information, or other secrets out of the same AI-powered bots. Because the hidden text can be combined with normal text, users can unwittingly paste it into prompts. The secret content can also be appended to visible text in chatbot output.

…To demonstrate the utility of “ASCII smuggling”—the term used to describe the embedding of invisible characters mirroring those contained in the American Standard Code for Information Interchange—researcher and term creator Johann Rehberger created two proof-of-concept (POC) attacks earlier this year that used the technique in hacks against Microsoft 365 Copilot. The service allows Microsoft users to use Copilot to process emails, documents, or any other content connected to their accounts. Both attacks searched a user’s inbox for sensitive secrets—in one case, sales figures and, in the other, a one-time passcode.

When found, the attacks induced Copilot to express the secrets in invisible characters and append them to a URL, along with instructions for the user to visit the link. Because the confidential information isn’t visible, the link appeared benign, so many users would see little reason not to click on it as instructed by Copilot. And with that, the invisible string of non-renderable characters covertly conveyed the secret messages inside to Rehberger’s server. Microsoft introduced mitigations for the attack several months after Rehberger privately reported it. The POCs are nonetheless enlightening.

«

Prompt injection takes many forms; I don’t think SQL injection was vulnerable to it (though control characters could count). New tech, variants of old forms of attack.
unique link to this extract

The Optimus robots at Tesla’s Cybercab event were humans in disguise • The Verge

Wes Davis:

»

Tesla made sure its Optimus robots were a big part of its extravagant, in-person Cybercab reveal last week. The robots mingled with the crowd, served drinks to and played games with guests, and danced inside a gazebo. Seemingly most surprisingly, they could even talk. But it was mostly just a show.

It’s obvious when you watch the videos from the event, of course. If Optimus really was a fully autonomous machine that could immediately react to verbal and visual cues while talking, one-on-one, to human beings in a dimly lit crowd, that would be mind-blowing.

Attendee Robert Scoble posted that he’d learned humans were “remote assisting” the robots, later clarifying that an engineer had told him the robots used AI to walk, spotted Electrek. Morgan Stanley analyst Adam Jonas wrote that the robots “relied on tele-ops (human intervention)” in a note, the outlet reports.

There are obvious tells to back those claims up, like the fact that the robots all have different voices or that their responses were immediate, with gesticulation to match.

It doesn’t feel like Tesla was going out of its way to make anyone think the Optimus machines were acting on their own. In another video that Jalopnik pointed to, an Optimus’ voice jokingly told Scoble that “it might be some” when he asked it how much it was controlled by AI.

«

It’s just so pathetic. Musk has nothing to show, but he feels he needs to show something, so he shows rubbish.
unique link to this extract

“I’m suing the council for £495m because they won’t give me back my bin bag” • Wales Online

Conor Gogarty:

»

A man has filed a court claim against Newport council in a “last resort” to get back almost half a billion pounds’ worth of Bitcoin. A mix-up saw James Howells’ hard drive dumped at a recycling centre in 2013 causing him to lose access to cryptocurrency coins which have since rocketed in value.

WalesOnline has seen a court document that says Mr Howells, 39, is suing the council for £495,314,800 in damages, which was the peak valuation of his 8,000 Bitcoins from earlier this year. But he told us this is not a reflection of “what is really going on” and the point is to “leverage” the council into agreeing to an excavation of its landfill to avoid a legal battle. Mr Howells says he has assembled a team of experts who would carry out the £10million dig at no cost to the council. He is also offering the council 10% of the coins’ value if recovered.

The case is due to be heard in December after what Mr Howells described as more than a decade of being “largely ignored” by the council. “I’m still allocating 10% of the value for the council even though they have been problematic throughout,” he said. “That would be £41m based on today’s rate but in the future it could be hundreds of millions. If they had spoken to me in 2013 this place would look like Las Vegas now. Newport would look like Dubai. That’s the kind of opportunity they’ve missed.”

The hard drive disaster unfolded after a miscommunication between the IT engineer and his then-partner. Mr Howells, who learned about Bitcoin in 2009 by spending time on IT forums, believes he was one of the very first miners of the cryptocurrency. In basic terms he created the 8,000 coins himself and they cost him nothing beyond pennies’ worth of electricity to run his laptop. He stored the private key needed to access the coins on a 2.5in hard drive which he put in a drawer at his home office.

In August 2013 he had a clearout of equipment. Looking through his drawers he came across two hard drives of the same size. One contained the Bitcoin data while the other was blank. Mistakenly he put the Bitcoin one into a black bin liner.

«

Alex Hern’s finest moment (for me) as a reporter was tracking this guy down in the first place, in November 2013, based on a few posts on Reddit. The saga grinds on, it seems.
unique link to this extract

Missing immune cells may explain why COVID-19 vaccine protection quickly wanes • Science

Jon Cohen:

»

Neither vaccinations nor immunity from infections seem to thwart SARS-CoV-2 for long. The frequency of new infections within a few months of a previous bout or a shot is one of COVID-19’s most vexing puzzles. Now, scientists have learned that a little-known type of immune cell in the bone marrow may play a major role in this failure.

The study, which appeared last month in Nature Medicine, found that people who received repeated doses of vaccine, and in some cases also became infected with SARS-CoV-2, largely failed to make special antibody-producing cells called long-lived plasma cells (LLPCs). “That’s really, really interesting,” says Mark Slifka, an immunologist at the Oregon Health & Science University who was not involved with the work. The study authors say their finding may indicate a way to make better COVID-19 vaccines: by altering how they present the spike surface protein of SARS-CoV-2 to a person’s immune cells.

Durability is an age-old bugaboo of vaccine designers. Some vaccines, particularly ones made from weakened versions of viruses, can protect people for decades, even life. Yet others lose effectiveness within months. “We really haven’t overcome this challenge,” says Akiko Iwasaki, a Yale University immunologist who is developing a nasal COVID-19 vaccine she hopes can be given often enough to get around the durability problem.

«

I wondered if this was just about mRNA vaccination; that if you had been infected with SARS-CoV-2 before a vaccination, whether you might have the LLPCs. However:

»

An earlier study of bone marrow from 20 people who had been infected with SARS-CoV-2 but never vaccinated against it also found that they were “deficient” in LLPCs specific to SARS-CoV-2 compared with those for tetanus.

«

unique link to this extract

I wish I went before Mary Shelley in this storytelling contest • McSweeney’s Internet Tendency

Mike Drucker:

»

“‘We will each write a ghost story,’ said Lord Byron; and his proposition was acceded to. There were four of us.” – Mary Shelley, in the introduction to Frankenstein.

Wow, Mary! Wow. Dr. Frankenstein and his monster. I can’t imagine anything more chilling. In fact, it’s so chilling that I think we should probably call off the rest of the storytelling contest right now. I don’t even need to take my turn.

Oh, are you sure?

Still?

Because I kind of wish I had gone first. My thing isn’t even that scary. Or about humankind. It’s just, well, did everyone else do this overnight? Because I feel like Mary Shelley may have pre-written her idea. All I’m saying is it feels pretty fleshed out already. I’m not trying to accuse anyone of anything. It’s just, I thought we were telling stories we came up with in the last twenty-four hours and not workshopping full novel ideas.

No, I didn’t dislike the story. It’s not about a ghost, so it doesn’t fit in the rules laid out by Lord Byron, but I love it! “A Modern Prometheus,” yeah, no, I get it. It’s really smart. And it makes you think about playing God and stuff, even though none of us would be able to play God that way. I know it’s a metaphor, but maybe try a more approachable idea for hubris? Just if you’re trying to pitch this later. I don’t even know that many people who have electricity, so it’s like, who’s going to get the message? Half the people who read this will just want the monster to wave his hands at fire or something.

«

This just gets better and better.
unique link to this extract

TikTok knows its app is harming kids, new internal documents show • NPR

Bobby Allyn, Sylvia Goodman and Dara Kerr:

»

For the first time, internal TikTok communications have been made public that show a company unconcerned with the harms the app poses for American teenagers. This is despite its own research validating many child safety concerns.

The confidential material was part of a more than two-year investigation into TikTok by 14 attorneys general that led to state officials suing the company on Tuesday. The lawsuit alleges that TikTok was designed with the express intention of addicting young people to the app. The states argue the multi-billion-dollar company deceived the public about the risks.

In each of the separate lawsuits state regulators filed, dozens of internal communications, documents and research data were redacted — blacked-out from public view — since authorities entered into confidentiality agreements with TikTok.

But in one of the lawsuits, filed by the Kentucky Attorney General’s Office, the redactions were faulty. This was revealed when Kentucky Public Radio copied-and-pasted excerpts of the redacted material, bringing to light some 30 pages of documents that had been kept secret.

After Kentucky Public Radio published excerpts of the redacted material, a state judge sealed the entire complaint following a request from the attorney general’s office “to ensure that any settlement documents and related information, confidential commercial and trade secret information, and other protected information was not improperly disseminated,” according to an emergency motion to seal the complaint filed on Wednesday by Kentucky officials.

«

unique link to this extract

No, the Postcode Address File should not cost £487m • Peter K Wells

Peter Wells:

»

In 2021 the UK government’s Geospatial Commission prepared a briefing paper for some discussions about address data. It has been released to the journalist James O’Malley after a freedom of information request.

The paper was prepared in response to the long-running campaign asking the government to deliver on political commitments to make the list of UK addresses – and other non-personal geospatial data – freely available. People could then use the data to improve public services or build innovative new businesses.

The paper includes the mind-boggling statement that a government project in 2016 estimated that the cost to the UK government of buying back UK address data [in effect the Postcode Address File, or PAF] from the Royal Mail would be £487m. 

Yes, you read that right. That figure was four hundred and eighty seven million pounds.

It’s a big number and – if true – one that would call into question the whole campaign.

But it doesn’t hold up to critical scrutiny and, unfortunately, the 2021 paper repeats this estimate without questioning it.

The civil service needs to be less credulous when it comes to claims over the financial value of data assets, and the UK government needs some fresh analysis.

«

The PAF generates £30m in revenue and £3m in profit annually. A £487m valuation is 162 P/E and makes the PAF worth 15% of Royal Mail’s entire business. How about the government nationalises it and promises to pay Royal Mail £3m for the next 163 years?
unique link to this extract

Errata, corrigenda and ai no corrida: There should be a double helping of emails. Apologies. WordPress keeps screwing around with the design of its blog pages, and that screws with my scripts which collate and schedule the posts. Fingers crossed.

Just over 30 years ago, Netscape Navigator was released. The mugs have lasted better but didn’t change the world like it did. CC-licensed photo by Steve Bowbrick on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 9 links for you. Directional. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

I’m running out of ways to explain how bad this is • The Atlantic

Charlie Warzel:

»

So much of the conversation around misinformation suggests that its primary job is to persuade. But as Michael Caulfield, an information researcher at the University of Washington, has argued, “The primary use of ‘misinformation’ is not to change the beliefs of other people at all. Instead, the vast majority of misinformation is offered as a service for people to maintain their beliefs in face of overwhelming evidence to the contrary.”

This distinction is important, in part because it assigns agency to those who consume and share obviously fake information. What is clear from comments such as Kremer’s is that she is not a dupe; although she may come off as deeply incurious and shameless, she is publicly admitting to being an active participant in the far right’s world-building project, where feel is always greater than real.

What we’re witnessing online during and in the aftermath of these hurricanes is a group of people desperate to protect the dark, fictitious world they’ve built. Rather than deal with the realities of a warming planet hurling once-in-a-generation storms at them every few weeks, they’d rather malign and threaten meteorologists, who, in their minds, are “nothing but a trained subversive liar programmed to spew stupid shit to support the global warming bullshit,” as one X user put it.

It is a strategy designed to silence voices of reason, because those voices threaten to expose the cracks in their current worldview. But their efforts are doomed, futile. As one dispirited meteorologist wrote on X this week, “Murdering meteorologists won’t stop hurricanes.” She followed with: “I can’t believe I just had to type that.”

What is clear is that a new framework is needed to describe this fracturing. Misinformation is too technical, too freighted, and, after almost a decade of Trump, too political. Nor does it explain what is really happening, which is nothing less than a cultural assault on any person or institution that operates in reality.

«

Warzel’s so right about this: what we’re seeing now is not an effort to change minds any more. Social networks have become places filled with experts, but stuffed with even more inexperts who have a different view and will shout them down.

To prove Warzel’s (and my) point, on Twitter his article was demeaned for wanting government regulation of content (by Marc Andreessen, literary critic, and others).
unique link to this extract

The editors protecting Wikipedia from AI hoaxes • 404 Media

Emanuel Maiberg:

»

A group of Wikipedia editors have formed WikiProject AI Cleanup, “a collaboration to combat the increasing problem of unsourced, poorly-written AI-generated content on Wikipedia.”

The group’s goal is to protect one of the world’s largest repositories of information from the same kind of misleading AI-generated information that has plagued Google search results, books sold on Amazon, and academic journals.

“A few of us had noticed the prevalence of unnatural writing that showed clear signs of being AI-generated, and we managed to replicate similar ‘styles’ using ChatGPT,” Ilyas Lebleu, a founding member of WikiProject AI Cleanup, told me in an email. “Discovering some common AI catchphrases allowed us to quickly spot some of the most egregious examples of generated articles, which we quickly wanted to formalize into an organized project to compile our findings and techniques.”

In many cases, WikiProject AI Cleanup finds AI-generated content on Wikipedia with the same methods others have used to find AI-generated content in scientific journals and Google Books, namely by searching for phrases commonly used by ChatGPT. One egregious example is this Wikipedia article about the Chester Mental Health Center, which in November of 2023 included the phrase “As of my last knowledge update in January 2022,” referring to the last time the large language model was updated. 

Other instances are harder to detect.

«

There’s an impressive one given, which is a long new article about an Ottoman fortress which has tons of detail. Also, never existed.
unique link to this extract

WordPress.org’s latest move involves taking control of a WP Engine plugin • The Verge

Wes Davis:

»

WordPress.org has taken over a popular WP Engine plugin in order “to remove commercial upsells and fix a security problem,” WordPress cofounder and Automattic CEO Matt Mullenweg announced today. This “minimal” update, which he labels a fork of the Advanced Custom Fields (ACF) plugin, is now called “Secure Custom Fields.”

It’s not clear what security problem Mullenweg is referring to in the post. He writes that he’s “invoking point 18 of the plugin directory guidelines,” in which the WordPress team reserves several rights, including removing a plugin, or changing it “without developer consent.” Mullenweg explains that the move has to do with WP Engine’s recently-filed lawsuit against him and Automattic.

Similar situations have happened before, but not at this scale. This is a rare and unusual situation brought on by WP Engine’s legal attacks, we do not anticipate this happening for other plugins.

WP Engine’s ACF team claimed on X that WordPress has never “unilaterally and forcibly” taken a plugin “from its creator without consent.”

«

The WordPress drama is shifting from low key mad to slightly higher key mad, all orchestrated by Mullenweg. I still don’t understand what the justification for his animus is: if WP Engine doesn’t contribute to the open source project, and just makes money from it, that’s completely allowed under the licence. Mullenweg seems to think it should be different. This will not end well: the antipathy that is building up between the well-funded WP Engine and the code-controlling WordPress is going to boil over soon.
unique link to this extract

The scourge of ‘win probability’ in sports • The Atlantic

Ross Andersen:

»

To watch baseball or any other sport is to confront the fundamental unpredictability of the universe, its utter refusal to bend to your wishes, no matter how fervent. In recent years, some broadcasters have sought to soothe this existential uncertainty with statistics.

This season, ESPN announced that a special graphic would appear on all of its Major League Baseball telecasts. In the upper-left corner of the screen, just above the score, each team’s chance of winning the game is expressed as a percentage—a whole number, reassuring in its roundness, that is recalculated after every at-bat. Its predictions may help tame the wild and fearful id of your fandom, restricting your imagination of what might happen next to a narrow and respectable range.

You might think that so insistently reminding fans of their team’s “Win Probability” would be against ESPN’s interests. If your team is down by several runs in the eighth inning, your hopes will already be fading. But to see that sinking feeling represented on the screen, in a crisp and precise-sounding 4%, could make an early bedtime more enticing. The producers of reality shows such as The Amazing Race know this, which is why they use quick cuts and split screens to deceive fans into thinking that teams are closer than they really are, and that the outcome is less certain than it really is.

But ESPN has a more evolved consumer in mind. We got a clue as to who this person might be in March, when Phil Orlins, a vice president of production at the company, previewed the graphic. Orlins said that Win Probability would speak “to the way people think about sports right now,” especially people “who have a wager on the game.”

«

Ugh. Might have guessed it would come down to betting: people trying to arbitrage a wager on the odds of a win. I find the win percentage predictions tedious: I watch sports not because I know the result, but because I don’t, and mad things can happen in sports.
unique link to this extract

The retreat To Muskworld • niedermeyer.io

E.W. Niedermeyer:

»

Almost eight years ago, Elon Musk announced that every Tesla made from that moment forward would be capable of Level 5 autonomous driving with nothing more than a software update. It was a pivotal moment in Tesla’s history, committing the company to not just succeed as an electric automaker, but solve one of the most ambitious AI and robotics challenges possible. To create confidence in that staggering aspiration, Tesla released a video of a Model X driving around Palo Alto autonomously to the Rolling Stones’ “Paint it Black,” claiming that the driver behind the wheel was only there “for legal purposes.”

Eight long and hype-filled years later, Tesla is still looking for ways to build confidence in its ability to deliver a “general solution to self-driving” through hype and spectacle, even as companies like Waymo deliver the reality of 100,000 driverless taxi rides per week. Rather than meeting the competitive challenge from Waymo with real driverless rides on real public streets, Tesla’s latest ploy for credibility sees the firm retreating ever deeper into fantasy, building what can only be described as a temporary theme park on a movie studio lot for its first ever “driverless” demonstration.

This contrast is instructive. The “Paint It Black” video of eight years ago was no more “real” or “fake” than yesterday’s “We, Robot” demonstration, but at least it had the pretense of reality: it depicted a real car on real roads. Tesla’s latest spectacle likely cost orders of magnitude more to produce, but it didn’t even purport to show any actual real-world capability. The entire thing was pure fantasy, in a contained fantasy world, built on a movie theater lot that exists for the sole purpose of producing such spectacles.
This trajectory, from simulating future capability on public roads to creating a fantasy world for fantasy cars to show off fantasy capabilities, should worry Tesla’s supporters.

…Ultimately, Musk’s increasingly-degenerate gambling run is slouching toward one last big coinflip: the 2024 presidential election.

«

That coinflip is not far away. SpaceX had a successful day, demonstrating a self-driving rocket; Tesla didn’t. The question of what happens to Musk if Harris wins remains open.
unique link to this extract

Google’s dominant search business is under attack, from TikTok to AI • WSJ

Suzanne Vranica and Miles Kruppa:

»

Google’s share of the U.S. search ad market is expected to drop below 50% next year for the first time in over a decade, according to the research firm eMarketer.

Amazon is expected to have 22.3% of the market this year, with 17.6% growth, compared with Google’s 50.5% share and its 7.6% growth.

“This space has been ripe for a shake-up for a long period of time,” said Brendan Alberts, head of search and commerce at the ad-buying firm Dentsu. 

Google remains in an enviable position: far ahead of the pack in the search market, with plenty of resources to counter moves by its rivals. Still, advertisers are eager for more competition.

“For the first time in probably 15 years, we will have viable alternatives to Google,” said Nii Ahene, a veteran digital-advertising executive. 

The generative-AI boom is transforming search products, which will increasingly serve up fully formed answers to user queries or summaries of the results. Google this past week rolled out ads in the AI-generated summaries it has begun placing at the top of search results. The ads will only show up on mobile searches in the U.S. at first, Google said.

In one example of how the new search ads might appear, Google showed a listing for a Tide pen that is available on the Albertsons website in an AI overview responding to the query, “How do I get a grass stain out of jeans?”

“We’re confident in this approach to monetizing our AI-powered experiences,” said Brendon Kraham, a Google vice president overseeing the search ads business. “We’ve been here before navigating these kinds of changes.”

«

This was discussed on a recent episode of Dithering, where Ben Thompson made the good point that the Dominant Thing is never disrupted by something that does the same. Google disrupted Microsoft not by doing Windows, but by doing search. Mobile disrupted the desktop web by offering different possibilities such as Uber.
unique link to this extract

ADM’s CCS project can’t seal the deal • CTVC

»

In the first major US initiative of its kind, ADM (Archer-Daniels-Midland) launched CCS [carbon capture and storage] operations at its Decatur, Illinois ethanol plant in 2021. This involves capturing CO2 emissions from the plant, purifying and compressing the CO2 into a liquid-like form, then transporting it into CO2 injection wells to be injected into sealed-off geologic formations ~5,550 feet (1.69km) underground for permanent sequestration. Next to these injection wells, which require federal permits known as Class VIs, are two deep monitoring wells that track the movement of the injected CO2 plume, the integrity of the injection well, and groundwater quality.

However, at the end of last year, ADM said it detected “some corrosion” in a section of one monitoring well, and subsequently plugged the well and no longer uses it. This March, the company reportedly discovered possible leakage in the rock formation above the CO2 injection well, at a depth of 5,000 feet — just above the zone where ADM is permitted to inject (5,553-7,043 feet underground). Tests confirmed the presence of CO2.

By August, the EPA issued a violation notice alleging that the company hadn’t complied with its federal permit. And at the end of September, while investigating, ADM said it discovered more potential movement of fluid “between different formations” 5,000 feet underground, prompting worries that the leak violates the Clean Drinking Water Act (although drinking water wells are only 110 feet deep).

«

CCS has always seemed like a daft idea. Now we’re seeing it’s impractical too.
unique link to this extract

Carbon Mapper releases first emissions detections from the Tanager-1 satellite • Carbon Mapper

Carbon Mapper Inc.:

»

“To meet ambitious climate goals, it is important for philanthropy to lead carefully and follow fast. This is exactly what we have done with our investment in the Carbon Mapper coalition. We were methodical in how we built an emissions monitoring program to drive transparency and actionable emissions insights, and we have delivered,” said Richard Lawrence, Founder and Executive Chairman of High Tide Foundation. “Now is the time to quickly scale up investments to get this data into the right hands so we can accelerate global actions to cut methane and CO2.”

To make this data accessible and actionable, Carbon Mapper makes all of its methane and CO2 detections publicly available for noncommercial use on its data portal, a web platform that is updated on an ongoing basis with observations and emissions data from remote sensing sources. 

“Reducing methane pollution starts with measuring it,” said Michael R. Bloomberg, UN Secretary-General’s Special Envoy on Climate Ambition and Solutions and Founder of Bloomberg L.P. and Bloomberg Philanthropies. “Data from the Tanager-1 satellite is providing us with the real-time data necessary to pinpoint methane leaks at their source and clean them up. This new technology is crucial to curbing emissions from one of the biggest contributors to climate change.”

Emissions data from Carbon Mapper alongside data from other monitoring programs will be critical to helping governments deliver on the Global Methane Pledge, an unprecedented agreement led by the United States and the European Union to reduce global methane emissions by 30% by 2030. It can also be transformative across major emitting sectors such as energy, waste and agriculture, empowering companies to identify and verify emissions reductions across their supply chains and deliver on stated commitments such as the Oil and Gas Decarbonization Charter.

«

What’s obvious from two minutes on the data portal is that the US – California and Texas – have the biggest number of methane plumes. Sort them, you’ve sorted a lot.
unique link to this extract

Mosaic Netscape 0.9 was released 30 years ago on Sunday • jwz

Jamie Zawinski was one of the founders of Netscape and, later, Mozilla:

»

According to my notes, it went live shortly after midnight on Oct 13, 1994. We sat in the conference room in the dark and listened to different sound effects fired for each different platform that was downloaded. At some point late that night I wandered off and wrote the first version of the page that loaded when you pressed the “What’s Cool” button in the toolbar. (A couple days later, Jim Clark would go ballistic in a company-wide email because I had included a link to Bianca’s Smut Shack.)

For those of you who are unaware of these finer details, 0.9 was the first release of the Netscape browser (which begat Firefox) available to the general public. This beta release was an unannounced surprise. Prior to this, everyone assumed that what we were doing was going to be a standard for-sale product where you sent off your $35 and then some time later got a disc in the mail with a license key. That we just said, “Here’s our FTP site, come get it, go crazy” was, at the time, shocking to people.

«

Now, his bio says ” I’m the proprietor of DNA Lounge, a world famous and award-winning all ages dance club and live music venue in San Francisco, and of DNA Pizza, our attached cafe and pizzeria.”

Though the question of whether you had to pay for Netscape did puzzle a lot of people (myself included) for quite a while. If you downloaded it personally but used it in a company was that a.. personal licence? Commercial one? And would they come after you and find out? All obviated eventually by Microsoft making Internet Explorer free.

Arguably, though, this is when the internet really became the internet.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Kickers in American football have discovered the benefits of physical training – and are now making field goals almost from the halfway line. CC-licensed photo by Ed Schipul on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

It’s Friday, so there’s another post due at the Social Warming Substack at about 0845 UK time.

A selection of 10 links for you. Flying. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

The future of live news online sucks • Read Max

Max Read:

»

On Wednesday evening, where did you go to find up-to-the-minute news about Hurricane Milton? If you checked X.com at around 8 p.m., as the storm made landfall in Tampa and Sarasota, and clicked on the prominent “Hurricane Milton” link in the “Happening now” sidebar, you were taken to a landing page with a few videos from Florida Governor Ron DeSantis and an exhortation to “check back later” because there was “nothing to see now.” Three hours later, at 11 p.m., nothing new had been added:

There was, of course, plenty to see on Wednesday night; it was just that it was basically impossible to find any of it on Twitter. The landing page was empty; the FYP feed worse than useless; the machine-curated hashtag pages a mix of days-old posts and influencers I’d never heard of sharing the same handful of images and videos. This was not a problem of “misinformation,” to be clear, so much as one of “no information”: Twitter seemed effectively incapable of serving me even relevant, up-to-the-minute fake stuff, let alone any actual news. Unless you’d already searched out and made a list of local journalists, meteorologists, and storm chasers, it was impossible to tell from Twitter alone where the storm was, how hard it was hitting, what its effects looked like, or how people were responding.

Instead–in place of the professional and citizen journalists, the eager experts, and the volunteer aggregators–what I found was clipped videos from a bunch of fucking freaks.

…It–really!–wasn’t always this way. For most of the Obama-Trump Era–the Long 2010s–Twitter was the website to go to for up-to-the-minute updates about Things That Were Happening–hurricanes, invasions, elections–from journalists, experts, and people on the ground. It was far from perfect; it was relatively easy for hoaxes to spread and morons to gain attention. But it more or less worked: Twitter’s search and sorting mechanisms, helped along by the many users who treated “posting about news” as an uncompensated part-time job, tended to surface interesting and relevant information in a timely fashion.

But “News,” even breaking news, is not really a priority for Twitter anymore, nor is it for any social-media platform.

«

This is a good point. I too was trying to find information about the hurricane; Twitter was basically a waste of time. Nowhere is good. In the comments, people suggested that YouTube is now the place to go for live info.
unique link to this extract

Hacked ‘AI Girlfriend’ data shows prompts describing child sexual abuse • 404 Media

Joseph Cox:

»

A hacker has targeted a website that lets users create their own “uncensored” AI-powered sexual partners and stolen a massive database of users’ interactions with their chatbots.

The data, taken from a site called Muah.ai and viewed by 404 Media, includes chatbot prompts that reveal users’ sexual fantasies. In many instances, users are trying to create chatbots that roleplay child sexual abuse scenarios. These prompts are in turn linked to email addresses, many of which appear to be personal accounts with users’ real names.

“I went to the site to jerk off (to an *adult* scenario, to be clear) and noticed that it looked like it [the Muah.ai website] was put together pretty poorly,” the hacker told 404 Media. “It’s basically a handful of open-source projects duct-taped together. I started poking around and found some vulnerabilities relatively quickly. At the start it was mostly just curiosity but I decided to contact you once I saw what was in the database.”

The administrator of Muah.ai, who used the name Harvard Han, told 404 Media in an email that “the data breach was financed by our competitors in the uncensored AI industry who are profit driven, whereas Muah AI becomes a target for being a community driven project.” The site’s operators detected that it was hacked last week.

Han didn’t provide 404 Media with any evidence for their claim, and the hacker said they do work in the tech industry but not on AI.

“We have a team of moderation staff that suspend and delete ALL child related chatbots on our card gallery, discord, reddit, etc,” Han added

«

Suuuure they have a team that does that. Sure they do. Of course they do. Open source projects running websites always have loads of administrators doing that sort of thing.
unique link to this extract

How to hire a dark net hitman • How to Survive the Internet

Jamie Bartlett:

»

Besa Mafia was a dark net site offering hitmen for hire. It worked something like this: a user could connect to the site using the Tor browser and request a hit. They’d send over some bitcoin (prices started from $5,000 USD for ‘death by shotgun’). Then they’d upload the name, address, photographs, of who they wanted killed. Plus any extra requests: make it look like a bungled robbery; need it done next week, etc. The website owner, a mysterious Romanian called ‘Yura’ would then connect them with a specialist hitman to carry out the commission.   

I know all about the dark net. I wrote a book about it and still jump on often to see what the latest trends are. Yes: you can buy all manner of illicit goods there. Stolen credit cards, identity documents, drugs (of course). You can purchase ransomware-as-a-service and chat with hackers to hire. You can download child sexual abuse images. But there has always been one area shrouded in mystery, even for me. Can you really order a hitman with a few clicks? There have always been hitman sites, I saw one while researching my book. But do they work? Are they real?  

Finally [the podcast] Kill List provides an answer. As he poked around, IT-guy Chris stumbled across a vulnerability on the site, and was able to access the ‘back end’. From there he could see what Yura could see: hundreds of names. Each one, a person someone else wanted dead. And next to each name: photographs, addresses, commute routes, phone numbers. Victims were from all over the world. Switzerland, Spain, the Czech Republic, the US, the UK. And alongside each, gruesome requests:  

Make it look like a road accident  

I would just like his person to be shot and killed. Where, how and what week does not bother me

This person needs to go away, but disposed of without a trace  

Need target killed, make it look like an accident  

And in 175 cases there was also evidence of a successful bitcoin payment. In other words: people had paid real money to have them killed.  

«

It’s now a big BBC podcast, which is a remarkable achievement. Hacking for good.
unique link to this extract

They’re better, longer and more valuable than ever: the NFL’s newest superstars are kickers • WSJ

Andrew Beaton and Rosie Ettenheim:

»

This season, teams are making more field goals per game than ever— even as they’re also hitting them from more far-flung distances. Some 29% of this year’s attempts have been from 50-plus yards, far and away the most in NFL history. At the same time, 76% of those long-distance kicks have gone in, which is easily the highest percentage ever. 

Somehow kickers are becoming more accurate and more powerful at the same time. Last week alone, there were 15 tries from at least 50 yards—and 12 of them went in. 

The shift has completely transformed some of football’s most basic dynamics. Now, offenses are often just a couple first downs away from a spot on the field where they can plausibly try a field goal instead of punting. It’s why field goals are at an all-time high, even though this is also an era when coaches are more inclined to leave kickers on the sideline and go for it on fourth downs. 

Those inside the game say there are a number of forces at work, from the precision of long-snappers to the technology used in training and a ball that flies better through the air. But they also say it boils down to something simple: kickers have gotten both far stronger and more technically sound than prior generations. 

The widespread proliferation of kicking camps has allowed kids to hone their crafts beginning at a young age. And kickers, who weren’t historically known as gym rats, now rigorously approach strength training. That’s why former NFL kicker Jay Feely says it’s “similar to Tiger Woods” and how he changed golf by adding power—forcing his competitors to do the same to keep up. 

“It’s unbelievable,” Feely says. “Now coaches expect to make these kicks, and so they’re more willing to attempt these kicks.”

«

The maths needs a little explanation (for non-gridiron fans): the end zone is 10 yards long, and the goalposts are at the end. So a 50-yard kick is made from the opponents’ 40-yard line, which is 10 yards inside their half. The way that physicality is changing sports shows up in all sorts of subtle ways.
unique link to this extract

How everyone got lost in Netflix’s endless library • The New York Times

Willy Staley:

»

For a company like Apple, where the streaming business is practically an afterthought, the shows it produces are stashed away in the app and hardly spoken of, no matter how big their budget and how impressive the talent involved. In some instances, streamers have shelved finished movies entirely so they could write down the losses on their taxes. These are the incentives of the streaming marketplace pushed to their logical extremes: mass entertainment completely severed from market signal — paradoxically by entities that know more about our viewing habits than ever before.

Which isn’t to say that the streamers don’t make hits and that people don’t watch and enjoy a lot of streaming television, as Netflix’s 183 billion viewer hours in 2023 can attest. But it can certainly account for the rise of so-called Mid TV: shows that look expensive, are reasonably smart and packed with talent and somehow manage to be, in the Times TV critic James Poniewozik’s words, “. . . fine?” There’s no denying that, in the long journey prestige TV has taken from “The Wire” to “The Bear,” a certain slackness has crept in: comedies without many jokes, dramas without any stakes, a pronounced preference for backward-looking plotting that fixates on characters’ traumas, a plague of visibly Canadian filming locations, “Barry.”

The first generation of prestige shows was created by veterans of linear TV who longed for creative freedom but knew the rudiments of the business, the things that kept you watching through the commercial breaks: pacing, structure, believable dialogue. But the leash has been off for a decade now, and eventually you face the same problem [original Netflix series] Richie Rich did: when you’re drowning in cash, it’s always tempting to say yes.

Maybe the most disorienting outcome of this information poverty has been the significant disconnects that can — and do — arise between what people watch and what we think we’re watching. This has been a persistent element of TV criticism since at least “Mad Men,” but it’s hard not to sense that things have gotten worse since then.

«

A thoughtful essay: it’s notable how Netflix is getting absolutely roasted for cancelling “Kaos”, a clever series based on the Greek myths, after one season. Meanwhile Apple keeps turning up with Slow Horses, which will next year hit Season 5. There’s entirely too much Mid TV.
unique link to this extract

Forums are still alive, active, and a treasure trove of information • Aftermath

Chris Person:

»

When I want information, like the real stuff, I go to forums. Over the years, forums did not really get smaller, so much as the rest of the internet just got bigger. Reddit, Discord and Facebook groups have filled a lot of that space, but there is just certain information that requires the dedication of adults who have specifically signed up to be in one kind of community. This blog is a salute to those forums that are either worth participating in or at least looking at in bewilderment.

What follows is a list of forums that range from at least interesting to good. I will attempt to contextualize the ones I know well. This post is by no means supposed to be complete and will be updated whenever I find more good forums.

«

Forums tend to have the problem, for me, that you need to do a ton of digging through to find the useful stuff. Though it’s not that different on social media! Perhaps one forgets the pain.

Anyhow, if you’re in need of a forum about something or other (which probably doesn’t show up in search results), this is for you.
unique link to this extract

Have we reached peak human life span? • The New York Times

Dana Smith:

»

The oldest human on record, Jeanne Calment of France, lived to the age of 122. What are the odds that the rest of us get there, too?

Not high, barring a transformative medical breakthrough, according to research published Monday in the journal Nature Aging.

The study looked at data on life expectancy at birth collected between 1990 and 2019 from some of the places where people typically live the longest: Australia, France, Italy, Hong Kong, Japan, South Korea, Spain, Sweden and Switzerland. Data from the United States was also included, though the country’s life expectancy is lower.

The researchers found that while average life expectancies increased during that time in all of the locations, the rates at which they rose slowed down. The one exception was Hong Kong, where life expectancy did not decelerate.

The data suggests that after decades of life expectancy marching upward thanks to medical and technological advancements, humans could be closing in on the limits of what’s possible for average life span.

“We’re basically suggesting that as long as we live now is about as long as we’re going to live,” said S. Jay Olshansky, a professor of epidemiology and biostatistics at the University of Illinois Chicago, who led the study. He predicted maximum life expectancy will end up around 87 years — approximately 84 for men, and 90 for women — an average age that several countries are already close to achieving.

«

Without some amazing bit of genetic wizardry (telomeres get hyped up every ten years or so as a new group of people rediscover them), it seems like only 15% of women could ever be centenarians, and 5% of men. The question always is, are those extra years the ones you want to live?
unique link to this extract

How North Korea infiltrated the crypto industry • Coindesk

Sam Kessler:

»

• CoinDesk identified more than a dozen crypto companies that unknowingly hired IT workers from the Democratic People’s Republic of Korea (DPRK), including such well-established blockchain projects as Injective, ZeroLend, Fantom, Sushi, Yearn Finance and Cosmos Hub.

• The workers used fake IDs, successfully navigated interviews, passed reference checks and presented genuine work histories.

• Hiring DPRK workers is against the law in the U.S. and other countries that sanction North Korea. It also presents a security risk: CoinDesk encountered multiple examples of companies hiring DPRK IT workers and subsequently getting hacked.

• “Everyone is struggling to filter out these people,” said Zaki Manian, a prominent blockchain developer who says he inadvertently hired two DPRK IT workers to help develop the Cosmos Hub blockchain in 2021.

«

I think this is the same story, in essence, as the Mandiant story a week ago, but showing that it was concentrated in the crypto industry – of course: slack hiring, desperate for workers, and easy access to lots of untraceable money. If you’re a country which is a pariah state then what’s not to like about the crypto industry?
unique link to this extract

The fight that nearly destroyed the Letterboxd community • WIRED

Adam Bumas:

»

THINGS LIKE THIS don’t happen on Letterboxd. It’s supposed to be a place where movie nerds share their love of cinema, a throwback to the internet’s pre-Facebook halcyon days. But lately, it’s been reeling from a disagreement between the site’s users and staff that got so big, major directors started weighing in. To make matters worse, it wasn’t some argument about Marvel movies or Martin Scorsese. It was about anime.

The trouble started on September 9, when Letterboxd’s curators updated the platform’s official list of top-rated movies. Usually, the list changes only when a new movie gets rated highly enough to remove another from the top 250, but Neon Genesis Evangelion: The End of Evangelion had gone from No. 23 overall to off the list entirely. In a comment on the list, curator Dave Vis called the removal “an effort to align our eligibility rules,” made after “careful consideration.”

Letterboxd’s users, by and large, didn’t agree that the effort had been very careful. The comment section of the “official top 100 animation” list, which also removed End of Evangelion, became a pressure chamber of fury, disagreement, and confusion, filled with the kind of negativity and argument the site has made a point of avoiding.

Letterboxd has grown steadily since its 2011 launch, and now boasts more than 15 million users. Until now, it has largely steered clear of growing pains, even as the platform took off during people’s Covid-19-lockdown-induced movie marathons. If anything, it has become a source for memes on other platforms like TikTok and X. But the drawback of creating a community modeled after the 2000s internet is re-creating the same tensions between moderators and users that plagued early social media platforms, which is pretty much what happened when End of Evangelion fell off the top lists.

«

You’re telling me that social patterns repeat again and again on the internet? Amazing.
unique link to this extract

PC shipments stuck in neutral despite AI buzz • The Register

Dan Robinson:

»

The PC market is not showing many signs of a rebound, despite the hype around AI PCs, with market watchers split over whether unit shipments are up or down slightly.

Those magical AI PC boxes were supposed to fire up buyer enthusiasm and spur the somewhat listless market for desktop and laptop systems into significant growth territory, but that doesn’t appear to be happening.

According to the latest figures from Gartner, global PC shipments totaled 62.9m units during Q3 of this year, representing a 1.3% decline compared with the same period last year. However, this does follow three consecutive quarters of modest growth.

“Even with a full line-up of Windows-based AI PCs for both Arm and x86 in the third quarter of 2024, AI PCs did not boost the demand for PCs since buyers have yet to see their clear benefits or business value,” commented Gartner Director Analyst Mikako Kitagawa.

This is perhaps understandable when AI PCs are largely just a marketing concept, and vendors can’t agree on exactly what the the definition of an AI PC should be. Even worse, some buyers of Arm-based Copilot+ machines discovered that their performance isn’t actually very good with some applications.

«

The hope that sprinkling some AI fairy dust on PCs would suddenly turn them into must-replace boxes was always overblown. I feel we’re past the point where they’re a commodity, and nothing exciting will ever happen in PCs again.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The US wants to get rid of all its lead water pipes in ten years – a decade after the Flint scandal. CC-licensed photo by Louise Devitt on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 9 links for you. Like the petrol. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Google faces US government attempt to break it up • The Guardian

Jack Simpson:

»

The Department of Justice has filed court papers that say it is considering enforcing “structural remedies” that would prevent Google from using some of its products such as Chrome, Android and Play, which the DoJ argues give the company an advantage over rivals.

Other actions being considered include blocking Google from paying to have its search engine pre-installed on smartphones and other devices.

Google, which is owned by Alphabet, said it would challenge any case by the DoJ and that the proposals marked an “overreach” by the government that would harm consumers.

The latest filing comes after a court ruling in August in favour of the DoJ that found Google, which controls 90% of the global search market, had violated antitrust laws and spent billions building up an illegal monopoly. The ruling paved the way for the current lawsuit by the justice department that will rule on potential actions to counteract Google’s market domination.

The filing said Google’s conduct had resulted in “interlocking and pernicious harms” to users, and the importance of restoring competition to a market, which was “indispensable” to Americans, could not be overstated.

The judgment said: “Plaintiffs are considering behavioural and structural remedies that would prevent Google from using products such as Chrome, Play, and Android to advantage Google search and Google search-related products and features – including emerging search access points and features, such as artificial intelligence –over rivals or new entrants.”

The move may also prevent Google from being able to pay major phone companies such as Apple and Samsung for Chrome to be the default browser on their devices. In 2021, Google paid companies $26.3bn to ensure its search engine was the default option in the products.

«

To be decided by the judge some time next summer, but the landscape in a few years will likely look very different.
unique link to this extract

Biden sets 10-year deadline for US cities to replace lead pipes nationwide • AP News

Matthew Daly and Michael Phillis:

»

A decade after the Flint, Michigan, water crisis raised alarms about the continuing dangers of lead in tap water, President Joe Biden on Tuesday set a 10-year deadline for cities across the nation to replace their lead pipes, finalizing an aggressive approach aimed at ensuring that drinking water is safe for all Americans.

Biden announced the final Environmental Protection Agency rule during a visit to the swing state of Wisconsin in the final month of a tight presidential campaign. The announcement highlights an issue — safe drinking water — that Kamala Harris has prioritized as vice president and during her presidential campaign. The new rule supplants a looser standard set by former President Donald Trump’s administration that did not include a universal requirement to replace lead pipes.

“Folks, what is a government for if it cannot protect the public health?” Biden asked a crowd of union members at a cavernous Department of Public Works warehouse in Milwaukee. The city has the fifth-highest number of lead pipes in the nation, according to the EPA.

Decades after the dangers of lead pipes were clear, more than 9 million lead pipes remain in use, a fact Biden called shameful.

«

Ten years later, and there’s another ten years to go? I thought Britain might have done better, but then found this June 2023 article which says that there are at least three million lead pipes in the network. It turns out to be a shockingly high proportion – a third of homes have some lead piping.
unique link to this extract

FEMA adds misinformation to its list of disasters to clean up • The Verge

Lauren Feiner:

»

The Federal Emergency Management Agency (FEMA) is fighting misinformation on top of a major storm cleanup in Florida as Hurricane Milton rapidly intensifies just after Hurricane Helene rocked the state.

FEMA Administrator Deanne Criswell told reporters on a call Tuesday that misinformation around the storms is “absolutely the worst I have ever seen,” according to Politico. FEMA posted a rumor response page about the hurricane, and though it’s not the first time it’s taken that kind of approach, Criswell said, “I anticipated some of this, but not to the extent that we’re seeing.”

FEMA’s rumor response page includes fact-checks to claims made by former President Donald Trump, like that the agency will only provide $750 to disaster survivors. FEMA says that’s just the amount provided quickly through “Serious Needs Assistance” for food and emergency supplies, but survivors could still be eligible for other types of funds, too. Other fact-checks include debunking the false claim that FEMA disaster response resources were diverted to border issues. FEMA says “Disaster Relief Fund money has not been diverted to other, non-disaster related efforts.”

Elon Musk, one of Trump’s most prominent tech backers, has also contributed to the misinformation, according to FEMA officials. Musk claimed on X last week that FEMA was “actively blocking citizens who try to help.” FEMA’s acting director for response and recovery, Keith Turi, told ABC that’s “absolutely not true. FEMA does not block anyone from helping or assisting. We do not confiscate supplies and use them for other purposes. In fact, we do the exact opposite.”

«

Personally I’m already wondering what happens to Musk’s interest in X/Twitter if (as increasingly seems possible) Trump does not win. Will he really think it’s worth burning all that money for something which failed to achieve what he wanted in propaganda terms? Will he just let it decline, or will he try to sell it?
unique link to this extract

Xiaohongshu helps Southeast Asia with tourism recovery post Covid-19 • Rest of World

Zhaoyin Feng:

»

At 5 a.m., the air around Ijen volcano in eastern Java, Indonesia, is thick with bright yellow smoke. Undeterred, young Chinese tourists don gas masks and flock to the rim of the active volcano crater at sunrise, eager to capture the perfect photo. Steps away from a sheer drop into the turquoise volcanic lake, some whip out their phones to check Xiaohongshu, a Chinese social media app, for the best vantage point. Within hours, their snapshots may join the thousands already shared on the platform. 

Aang Koen is familiar with this phenomenon. The 48-year-old Indonesian owns a travel agency in Surabaya and organizes tours to Ijen. For years, his clientele was predominantly European, but since early 2023, Koen’s business has undergone a dramatic shift. Now, 60% of his clients are Chinese, most of whom found him on Xiaohongshu. 

“My business is getting famous,” he told Rest of World. Ijen was little-known among Chinese travelers before the end of the Covid-19 pandemic, Koen said, but the volcano rim pictures on Xiaohongshu have made it a popular destination. Almost all of his Chinese clients asked to include Ijen in their tours.

Xiaohongshu is often referred to as “China’s Instagram,” but it offers something that Instagram and other social media apps generally do not: in-depth, user-generated travel advice and itineraries. Travel posts on the app often include comprehensive hotel and restaurant reviews, tips on transportation logistics, and curated lists of shops and attractions, all complemented by stunning snapshots. With over 300 million monthly active users, Xiaohongshu has become a beloved resource among young Chinese travelers, many of whom consider it their go-to travel guide.

«

Literally the rest of the world, and pretty much unknown to almost anyone in the western world. Apart from you, of course.
unique link to this extract

The bill finally comes due for Elon Musk • The Verge

Andrew Hawkins:

»

For almost as long as he’s been CEO of Tesla, Elon Musk has been bullshitting us about self-driving cars. 

In 2016, he said Tesla self-driving cars were “two years away.” A year later, it was “six months, definitely,” and customers would be able to actually sleep in their Tesla in “two years.” In 2018, it was still a “year away” and would be “200% safer” than human driving. In 2019, he said there would be “feature complete full self-driving this year.” There hasn’t been a year go by without Musk promising the imminent arrival of a fully driverless Tesla. 

This week, it’s finally here. Or at least that’s what Musk says.

On October 10th, Tesla will reveal its long-awaited “robotaxi,” a supposedly fully autonomous vehicle that Musk has said will catapult the company into trillion-dollar status. It will be some combination of “Uber and Airbnb,” Musk said during a recent earnings call, allowing Tesla owners to serve as landlords for their driverless cars as they roam about the cityscape, picking up and dropping off strangers. And it will be futuristic in its design, with Bloomberg reporting that it will be a two-seater with butterfly wing doors. Musk has been calling it the “Cybercab.”

The event, which will be held on the film lot of Warner Bros. in Burbank, California, will be the culmination of almost a decade of blown deadlines and broken promises from Musk, a moment when the richest man in the world will finally be forced to stop hiding behind his own bluster and actually show us what he’s been working on. 

«

I detect a certain amount of scepticism on the part of Mr Hawkins here. All, one has to agree, justified.
unique link to this extract

Protein structure prediction wins the Nobel • Ars Technica

John Timmer:

»

Chemically, proteins are a linear string of amino acids linked together, with living creatures typically having the choice of 20 different amino acids for each position along the string. Most of those 20 have distinctive chemical properties: some are acidic, others basic; some may be negatively charged, others positively charged, and still others neutral, etc. These properties allow different areas of the string to interact with each other, causing it to fold up into a complex three-dimensional structure. That structure is essential for the protein’s function.

Typically, figuring out the structure involves laborious biochemistry to purify the protein, followed by a number of imaging techniques to determine where each of its atoms resides. But in theory, all of that should be predictable since the structure is just the product of chemistry and physics. Since any amino acid could potentially interact with any other on the chain, however, the complexity of making predictions rises very rapidly with the length of the protein. Extend it out past a dozen amino acids long, and it could quickly humble the most powerful supercomputers.

A lot of work over the years went into trying to figure out computational shortcuts. DeepMind, by contrast, did what it did best and put an AI on the case. For protein folding, the AI was trained on two large existing data sets. One included every protein structure that had been solved through lab work, allowing it to extract general principles for how different amino acids typically interact. The second was the sequence of every protein we’ve determined, allowing it to identify proteins related through evolution and determine what sorts of flexibility can be tolerated in a given structure.

The net result is software that produced reasonable structural predictions, easily beating every other software package we’d developed in a regular computational challenge. DeepMind has since used it to generate predictions for most of the existing protein-coding genes in the databases (it still struggles with excessively long ones) and has continued to upgrade the software. The predictions aren’t perfect, and some appear to stumble badly, but when the alternative is simply a string of amino acids and a shrug, this represents a major advance.

«

The Nobel Prize in Chemistry went jointly to Demis Hassabis and John Junper of DeepMind for their AlphaFold work (described above), and also to Professor David Baker, who did a variant of that. Notable how quickly this has won the Nobel: AlphaFold only appeared in 2020.
unique link to this extract

Quantifying ‘The Kevin Bacon Game’: a statistical exploration of Hollywood’s most connected actors • Stat Significant

Daniel Parris:

»

This modeling technique also broadly applies to the film industry, where we can create networks using actors as nodes and movies as edges. Consider the following example graph with Jack Nicholson, Adam Sandler, and Leonardo DiCaprio:

Nicholson and Sandler were in Anger Management together.
DiCaprio and Nicholson were both in The Departed.
The resulting graph looks something like this:

We can build a network featuring every actor in Hollywood and then calculate which figures are most central to the film industry’s casting graph. Below is an example network of the 50 most connected actors in entertainment (to use more than 50 examples would render the visual unreadable). The actors (nodes) that are largest and most central, like Morgan Freeman and Bruce Willis, are ideal fodder for the Kevin Bacon game—a cheat code for connecting disparate careers via extensive movie credits.

An example network of actors connected by co-starring film credits.

We’ll use “eigenvector centrality” to measure an actor’s network importance, which considers both direct connections and the influence of second and third-order connections. A higher score means an actor is more central to our casting network.

When we calculate our centrality metric for every actor in the film industry, our top scorers are prolific performers who have appeared in numerous ensemble projects (which is not surprising at all).

«

I won’t spoil the surprise, but see if you can guess who the most connected actor is. (It’s no longer Kevin Bacon, though with all the adverts he’s doing for EE, maybe he could claim still to be. Brm-tish.)
unique link to this extract

How Meta brings in millions off political violence • The Markup

Colin Lecher and Tomas Apodaca:

»

After the attempted assassination of Donald Trump in July, the merchandise started showing up on Facebook.

Trump, fist in the air, face bloodied from a bullet, appeared on everything. Coffee mugs. Hawaiian shirts. Trading cards. Commemorative coins. Heart ornaments. Ads for these products used images captured at the scene by Doug Mills for the New York Times and Evan Vucci for the Associated Press, showing Trump yelling “fight” after the shooting. The Trump campaign itself even offered some gear commemorating his survival.

As the Secret Service drew scrutiny and law enforcement searched for a motive, online advertisers saw a business opportunity in the moment, pumping out Facebook ads to supporters hungry for merch.

In the 10 weeks after the shooting, advertisers paid Meta between $593,000 and $813,000 for political ads that explicitly mentioned the assassination attempt, according to The Markup’s analysis. (Meta provides only estimates of spending and reach for ads in its database.) 

Even Facebook itself has acknowledged that polarizing content and misinformation on its platform has incited real-life violence. An analysis by CalMatters and The Markup found that the reverse is also true: real-world violence can sometimes open new revenue opportunities for Meta.

While the spending on assassination ads represents a sliver of Meta’s $100 billion-plus ad revenue, the company also builds its bottom line when tragedies like war and mass shootings occur, in the United States and beyond. After the October 7th attack on Israel last year and the country’s response in Gaza, Meta saw a major increase in dollars spent related to the conflict, according to our review.

«

You could say: oh but Meta isn’t taking a position! But the difference is that in the age of print newspapers, there would have been questions raised before these adverts would get printed; they might even be rejected on the grounds of poor taste, or profiting from political violence – which implicitly normalises it.
unique link to this extract

Reach bosses in drive to increase websites’ story volumes • HoldtheFrontPage

Paul Linford:

»

Bosses at [UK newspaper group] Reach plc have launched a drive to increase story count across its network of regional websites in a bid to boost online traffic.

Company chiefs want to increase article volume in order to boost page views, which have been badly hit by algorithm changes from major referrers such as Google and Facebook.

While there are no individual targets for journalists, reporters on office-based shifts will be expected to generate more stories in order to increase the overall volume of articles on the group’s ‘Live’ newsbrands.

One internal email, seen by HTFP, suggests an average count of eight stories a shift for reporters working in the office, although this would not apply to those who are sent out on stories.

The email, sent by a senior Reach editor, states: “We need to make more of shifts where people are not going out as drivers of volume. In practice, if you’re on a general shift and you’re not on a job, it should be at least eight stories a shift.”

The latest IPSOS traffic figures published by HTFP last week showed a mixed picture for Reach’s leading websites.

While Birmingham Live increased its page views by a third during August compared to the same month in 2023, other big city brands such as the Liverpool Echo, Chronicle Live, Bristol Post and Nottingham Post all saw significant year-on-year decreases.

The overall thinking behind the latest move was set out out in an email from Paul Rowland, editorial director of the Live network, which has also been seen by HTFP.

«

When I finished at The Guardian (ten years ago), I thought I was doing OK if I got three stories written in a day, as much as anything because there were probably only three stories worth writing. Eight is just bonkers. There truly isn’t that much news, especially not at the local level. And this is all going to be AI-driven, until the AIs take it all over.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The Chagos Islands have beautiful seas – but if they change hands, the .io domain could cease to exist. Then what? CC-licensed photo by Foreign, Commonwealth & Development Office on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 11 links for you. Subject to treaty. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

India holds record for internet shutdowns with 771 since 2016 • Rest of World

Ananya Bhattacharya:

»

On September 10, when the northeast Indian state of Manipur announced a five-day internet shutdown in response to student-led protests, citizens were livid — but not surprised.

India has been a leader in internet shutdowns, by a huge margin, for nearly a decade, according to data shared by digital rights watchdog Access Now.

While some authoritarian regimes, such as those in China, North Korea, and Russia systematically censor, surveil, or limit the internet, India’s citizens have relatively free access. But India is unique among democratic countries for its frequent enforcement of blackouts. Between 2016 and 2023, India shut down the internet 771 times, Access Now’s data shows.

Myanmar and Ukraine ranked second on the list of internet shutdowns in 2023 and 2022, respectively. “These are regions obviously embroiled in conflict. And if you look at the margin between India and the second spot — massive,” Namrata Maheshwari, senior policy counsel at Access Now, told Rest of World. “There is no form of an internet shutdown that is proportionate or necessary.”

Reasons for the blackouts in India have included the government’s attempts to control agitation surrounding the Citizenship Amendment Act, suppress the farmers’ protests, and curb cheating during exams, according to Software Freedom Law Center, India’s tracker on internet shutdowns.

The majority of shutdowns in India have historically occurred in Jammu and Kashmir, a region at the center of a decades-long dispute between India, Pakistan, and China. In August 2019, it experienced  552 consecutive days of internet blackout, the world’s longest shutdown in history.

«

The irony – or perhaps explanation – is that India’s population absolutely loves the internet, and being connected. So being cut off this much is one of the most frustrating things they can experience.
unique link to this extract

The disappearance of an internet domain • Every

Gareth Edwards:

»

On October 3, the British government announced that it was giving up sovereignty over a small tropical atoll in the Indian Ocean known as the Chagos Islands. The islands would be handed over to the neighboring island country of Mauritius, about 1,100 miles off the southeastern coast of Africa. 

The story did not make the tech press, but perhaps it should have. The decision to transfer the islands to their new owner [by signing a treaty] will result in the loss of one of the tech and gaming industry’s preferred top-level domains: .io.

Whether it’s Github.io, gaming site itch.io, or even Google I/O (which arguably kicked off the trend in 2008), .io has been a constant presence in the tech lexicon. Its popularity is sometimes explained by how it represents the abbreviation for “input/output,” or the data received and processed by any system. What’s not often acknowledged is that it’s more than a quippy domain. It’s a country code top-level domain (ccTLD) related to a nation—meaning it involves politics far beyond the digital world.

…Once this treaty is signed, the British Indian Ocean Territory will cease to exist. Various international bodies will update their records. In particular, the International Standard for Organization (ISO) will remove country code “IO” from its specification. The Internet Assigned Numbers Authority (IANA), which creates and delegates top-level domains, uses this specification to determine which top-level country domains should exist. Once IO is removed, the IANA will refuse to allow any new registrations with a .io domain. It will also automatically begin the process of retiring existing ones. (There is no official count of the number of extant .io domains.)

Officially, .io—and countless websites—will disappear. At a time when domains can go for millions of dollars, it’s a shocking reminder that there are forces outside of the internet that still affect our digital lives.

«

IANA will probably – probably – do something to keep this going, since there’s a lot of money sloshing around in those domains.
unique link to this extract

US states sue TikTok, claiming its addictive features harm youth mental health • The Guardian

Johana Bhuiyan, Nick Robins-Early and agencies:

»

More than a dozen states and the District of Columbia filed lawsuits against TikTok on Tuesday, alleging the popular short-form video app is damaging children’s mental health with a product designed to be used compulsively and excessively.

The lawsuits stem from a national investigation into TikTok, which was launched in March 2022 by a bipartisan coalition of attorneys general from several states, including California, Kentucky and New Jersey. All of the complaints were filed in state courts and claim that TikTok’s algorithm is especially dangerous given the platform’s widespread use among young people and its ability to deliver quick hits of dopamine. Design choices such as infinite scrolling, push notifications and in-app purchases prey on youth and create addictive habits among users, prosecutors allege. There are over 170m monthly active TikTok users in the US, and over a billion worldwide.

At the heart of each lawsuit is the TikTok algorithm, which powers what users see on the platform by populating the app’s main “For You” feed with content tailored to people’s interests.

In its filings, the District of Columbia called the algorithm “dopamine-inducing”, and said it was created to be intentionally addictive so the company could trap many young users into excessive use and keep them on its app for hours on end. TikTok does this despite knowing that these behaviors will lead to “profound psychological and physiological harms”, such as anxiety, depression, body dysmorphia and other long-lasting problems, the complaint said.

«

Popcorn time, though probably won’t reach a court until some time in 2025, and properly until 2026. There’s still a US ban due in January 2025 unless ByteDance sells the US division to a US owner.
unique link to this extract

Starlink was offered for free to those hit by Hurricane Helene. It is not actually free • The Register

Brandon Vigliarolo:

»

The free Starlink service Elon Musk and SpaceX so graciously promised for communities devastated by Hurricane Helene in the US is not actually entirely free, according to those living in the aftermath – and the satellite operator’s own signup page.

There is a significant caveat: you are still expected to foot the bill for the hardware.

Starlink’s Twitter account declared last week, in a post with tens of millions of views, that “Starlink is now free for 30 days.” The world’s richest man, with a net worth of approximately $260bn, followed up by saying, in quite the PR coup, that all Starlink terminals would now work automatically “without [the] need for payment in the areas affected by Hurricane Helene.” 

But try to sign up for the ostensibly “free” service in an area Starlink has designated as a Helene disaster zone, and surprise: You still have to pay for the terminal (normally $350, but reportedly discounted to $299 for disaster relief, though that’s not reflected in Starlink’s signup page), plus shipping and tax, bringing the grand total to just shy of $400.

You can see for yourself in the video: putting in the address of city hall in Boone, North Carolina, one of the areas wrecked by the lethal super storm, shows folks recovering from the disaster are still expected to pay hundreds for that that free, month-long Starlink service. Though better than nothing at all, it is not quite the humanitarian aid it was promoted and heralded as.

«

There’s also the question of being signed up to an auto-renewing contract – so one needs to cancel within 30 days.
unique link to this extract

How to delete Your 23andme data • Lifehacker

Beth Skwarecki:

»

So far there is no definitive word about the company being sold, with or without its data. However, it’s reasonable to expect that the company could be sold, and that owner could inherit the data. Something similar happened when MyHeritage bought Promethease, another DNA analysis company, in 2020.

Currently, your data may already be shared with other companies. If you signed up to participate in research studies through 23andme, “de-identified” data about you (including genetic data) has likely been given to research institutions and pharma companies. For example, 23andme has a data licensing agreement with GSK (formerly GlaxoSmithKline) to use the 23andme database to “conduct drug target discovery and other research.” 

This isn’t a possible future scenario, but rather the current operation of the business. Licensing agreements like these are a big part of how 23andme makes money. Or intends to make money. Or possibly once made money. They’re not doing so great at the money making thing these days. 

Deleting your 23andme data doesn’t necessarily withdraw it from studies, especially since the data was “de-identified,” that is, stripped from your name and personal information. It does mean that your data will not be used in future research projects.

«

OK, but how do you delete your data?

»

If you would like to keep any of your data, sign in to your account and visit your user settings page. (You can also opt out of research studies there.). Click View on the 23andme Data card.

You’ll be asked to enter your date of birth to confirm your identity. In theory this is where you can download your data, but I can’t test this—I have a 23andme account, but I must have given the company a fake date of birth all those years ago. The website just tells me to contact Customer Care.

«

Well, huh. Kind of sold us a pup there, Lifehacker.
unique link to this extract

Monday, October 7 2024 • Scripting News

Dave Winer has hit 30 years of blogging:

»

Blogging started out as a programming adventure and eventually became a form of literature. How about that. I’m up for doing more of that if you all are. But please expect to make contributions, don’t expect it all to come to you for free, because as we know nothing really is free. #

Today’s the big day. Thanks to John Naughton’s wonderful piece in the Guardian, I’m hearing from people all over the world about what blogging means to them. I appreciate all of the messages, but would appreciate them even more if they were on your blog. We need to keep using the tech. Blogging is kind of lost, and I would like to see that change. Every time you post something you’re proud of on a social media site, how about taking a moment and posting it to your blog too. And while there, if appropriate, link to something from some part of your post, even though the social media sites don’t support linking, the web is still there and it still does. #

Interestingly, the clock at the bottom of the nightly emails does not agree with the clock on the home page of Scripting News. It’s a hard thing to test in real life. And it’s completely fitting, given the motto of the blog is: it’s even worse than it appears, which could be the motto of all programmers everywhere, and probably bloggers too. We always focus on the bad news, of course — that’s human nature — but always remember, it could actually be worse. #

«

Is blogging going to make a comeback? There are many, many more ways to express oneself nowadays. Is Substack blogging? (I think so.) Is social media? (Arguably not, too difficult to roll back through time.)
unique link to this extract

What happens if a hurricane smashes Tampa, Florida? • The Big Newsletter

Matt Stoller:

»

Tampa hasn’t been hit directly by a big storm since 1921, and it has grown a lot since then.

Hopefully, [Hurricane] Milton doesn’t hit the city directly, and it may not. These big storms almost always tend to avoid the most catastrophic hits. But let’s go over some of the consequences if it does. First, Tampa has an important port, managing 33 million tons of cargo a year. It’s the biggest exporter of fertilizer in America, and is the biggest importer of gasoline and jet fuel used in Florida. So that means we can expect significant supply shocks, and probably environmental damage. Utilities are already stretched because of Helene, so replacing electrical equipment is going to be difficult. All the major supplies for recovery, everything from lumber to ice to drinking water to skilled labor, are already being sucked into North Carolina to deal with the after-effects of Helene.

In addition, MacDill Air Force Base is in Tampa, which is where CENTCOM, the command center for U.S. forces in the Middle East, is located. So we could see modest disruptions to U.S. military operations. Tampa is near important NASA assets like Cape Canaveral, so there are space and defense contractors in the city.

There’s a lot more than that, of course, since Tampa is a major metropolitan area, an important hub for fishing, tourism, medicine, manufacturing and finance. It has convenient rail lines and highways to pair with its deepwater port, serving as a trans-shipment point for moving goods throughout Florida. Beyond Tampa, a good chunk of Florida is in the path of this storm, with unpredictable consequences. For instance, I wrote earlier about the shortage of IV fluids due to problems in North Carolina. It turns out another large IV solutions manufacturing plant in Daytona Beach is in the path of Milton. Yikes.

The insured losses could be massive, and we could face shortages of all sorts of random and important stuff. But more than these elements, we might lose an entire city, an apocalyptic level of destruction. And increasingly in Florida, there is no way to insure anything. In fact, the state itself, through its Citizens Property Insurance Corp., self-insures against natural disasters, because private insurers just won’t do business in Florida anymore. That means Florida property owners – who in aggregate own about $4 trillion – could enter a death spiral where they can’t get insurance, and so can’t get financing.

«

Milton is expected to make landfall on Thursday morning, UK time. Fingers crossed.
unique link to this extract

The connected TV industry’s unprecedented “surveillance” • Ars Technica

Scharon Harding:

»

The companies behind the streaming industry, including smart TV and streaming stick manufacturers and streaming service providers, have developed a “surveillance system” that has “long undermined privacy and consumer protection,” according to a report from the Center for Digital Democracy (CDD) published today and sent to the Federal Trade Commission (FTC). Unprecedented tracking techniques aimed at pleasing advertisers have resulted in connected TVs (CTVs) being a “privacy nightmare,” according to Jeffrey Chester, report co-author and CDD executive director, resulting in calls for stronger regulation.

The 48-page report, How TV Watches Us: Commercial Surveillance in the Streaming Era [PDF], cites Ars Technica, other news publications, trade publications, blog posts, and statements from big players in streaming—from Amazon to NBCUniversal and Tubi, to LG, Samsung, and Vizio. It provides a detailed overview of the various ways that streaming services and streaming hardware target viewers in newfound ways that the CDD argues pose severe privacy risks. The nonprofit composed the report as part of efforts to encourage regulation. Today, the CDD sent letters to the FTC [PDF], Federal Communications Commission (FCC), California attorney general [PDF], and California Privacy Protection Agency (CPPA) [PDF], regarding its concerns.

“Not only does CTV operate in ways that are unfair to consumers, it is also putting them and their families at risk as it gathers and uses sensitive data about health, children, race, and political interests,” Chester said in a statement.

…The report notes “misleading” privacy policies that have minimal information on data collection and tracking methods and the use of marketing tactics like cookie-less IDs and identity graphs that make promises of not collecting or sharing personal information “meaningless.”

“As a consequence, buying a smart TV set in today’s connected television marketplace is akin to bringing a digital Trojan Horse into one’s home,” it says.

«

unique link to this extract

Machine learning pioneers win Nobel prize in physics • The Guardian

Ian Sample:

»

Two researchers who helped lay the foundations for modern artificial intelligence – although one later warned of its potential harms – have been awarded the 2024 Nobel prize in physics.

Inspired by the workings of the brain, John Hopfield, a US professor emeritus at Princeton University, and Geoffrey Hinton, a British-Canadian professor emeritus at the University of Toronto, built artificial neural networks that store and retrieve memories like the human brain, and learn from information fed into them.

Hinton, 76, who is often called “the godfather of AI”, made headlines last year when he quit Google and warned about the dangers of machines outsmarting humans.

The scientists’ pioneering work began in the 1980s and demonstrated how computer programs that draw on neural networks and statistics could form the basis for an entire field, which paved the way for swift and accurate language translation, facial recognition systems, and the generative AI that underpins chatbots such as ChatGPT, Gemini and Claude.

Hopfield, 91, was honoured for building “an associative memory that can store and reconstruct images and other types of patterns in data”, while Hinton invented a method that can “independently discover properties in data”, an important feature of the large artificial neural networks in use today.

In 1982, Hopfield built a neural network that stored images and other information as patterns, mimicking the way memories are stored in the brain. The network was able to recall images when prompted with similar patterns, akin to identifying a song heard only briefly in a noisy bar.

Hinton built on Hopfield’s research by incorporating probabilities into a multilayered version of the neural network, leading to a program that could recognise, classify and even generate images after being fed a training set of pictures.

«

Doesn’t feel like physics, though, does it. Are we in The Three-Body Problem? Has physics stopped?
unique link to this extract

How the US lost the solar power race to China • Bloomberg

David Fickling:

»

Washington blames China’s dominance of the solar industry on what are routinely dubbed “unfair trade practices.” But that’s just a comforting myth. China’s edge doesn’t come from a conspiratorial plot hatched by an authoritarian government. It hasn’t been driven by state-owned manufacturers, subsidized loans to factories, tariffs on imported modules or theft of foreign technological expertise. Instead, it’s come from private businesses convinced of a bright future, investing aggressively and luring global talent to a booming industry — exactly the entrepreneurial mix that made the US an industrial powerhouse.

The fall of America as a solar superpower is a tragedy of errors where myopic corporate leadership, timid financing, oligopolistic complacency and policy chaos allowed the US and Europe to neglect their own clean-tech industries. That left a yawning gap that was filled by Chinese start-ups, sprouting like saplings in a forest clearing. If rich democracies are playing to win the clean technology revolution, they need to learn the lessons of what went wrong, rather than just comfort themselves with fairy tales.

To understand what happened, I visited two places: Hemlock, Michigan, a tiny community of 1,408 people that used to produce about one-quarter of the world’s PV-grade polysilicon, and Leshan, China, which is now home to some of the world’s biggest polysilicon factories. The similarities and differences between the towns tell the story of how the US won the 20th century’s technological battle — and how it risks losing its way in the decades ahead.

… the core questions are often almost impossible to answer. Is [China’s] Tongwei’s cheap electricity from a state-owned utility a form of government subsidy? What about [US company] Hemlock’s tax credits protecting it from high power prices? Chinese businesses can often get cheap land in industrial parks, something that’s often considered a subsidy. But does zoning US land for industrial usage count as a subsidy too? Most countries have tax credits for research and development and compete to lower their corporate tax rates to encourage investment. The factor that determines whether such initiatives are considered statist industrial policy (bad), or building a business-friendly environment (good), is usually whether they’re being done by a foreign government, or our own.

«

Long piece, but worthwhile.
unique link to this extract

Threads knows it has an engagement bait problem • The Verge

Jess Weatherbed:

»

If you’ve noticed more engagement bait appearing across your Threads feed, you’re not alone. Meta is aware of the issue and looking at how to address it, according to Instagram boss Adam Mosseri. “We’ve seen an increase in engagement-bait on Threads and we’re working to get it under control,” Mosseri said on Threads in response to comments flagging the issue.

Engagement bait on Threads typically covers posts with banal questions or invites for open-ended discussions to encourage other users to interact. Because Threads, like Instagram, pushes users to see an algorithmic feed of posts by default, getting more interactions can snowball a simple post into virality.

When asked about comparisons between Threads and X in a recent Decoder interview, Meta CEO Mark Zuckerberg alluded to the fact that comment engagement was a fundamental aspect of a “very good discussion-oriented platform.”

“Not all comments or replies are good,” said Mosseri. “Mark’s comment is more about the Twitter pioneering a format where the reply can be elevated, which is a good thing, but that doesn’t mean that every reply should be.” The thread that Mosseri replied to cites one example of a bait post, featuring a seemingly AI-generated image paired with an incendiary take on politeness that had pulled in more than 17,000 responses.

«

You’d think – you’d think! – that by now Meta would know what does and doesn’t work for engagement, and what reaction clickbait would get. It’s hard to shake a tiny suspicion that Meta doesn’t mind the wild engagement for now because it keeps people on the site.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Set your watches: in about 40 million years Mar’s moon Phobos will spiral down into the planet. CC-licensed photo by Andrea Luck on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 10 links for you. Orbiting. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

US wiretap systems targeted in China-linked hack • WSJ

Sarah Krouse, Dustin Volz, Aruna Viswanatha and Robert McMillan:

»

A cyberattack tied to the Chinese government penetrated the networks of a swath of US broadband providers, potentially accessing information from systems the federal government uses for court-authorized network wiretapping requests.

For months or longer, the hackers might have held access to network infrastructure used to cooperate with lawful US requests for communications data, according to people familiar with the matter, which amounts to a major national security risk. The attackers also had access to other tranches of more generic internet traffic, they said.

Verizon Communications, AT&T and Lumen Technologies are among the companies whose networks were breached by the recently discovered intrusion, the people said. 

The widespread compromise is considered a potentially catastrophic security breach and was carried out by a sophisticated Chinese hacking group dubbed Salt Typhoon. It appeared to be geared toward intelligence collection, the people said. 

Spokesmen for AT&T, Verizon and Lumen declined to comment on the Salt Typhoon campaign.

Companies are generally required to disclose material cyber intrusions to securities regulators within a short time, but in rare cases, federal authorities can grant them an exemption from doing so on national security grounds.

The surveillance systems believed to be at issue are used to cooperate with requests for domestic information related to criminal and national security investigations. Under federal law, telecommunications and broadband companies must allow authorities to intercept electronic information pursuant to a court order. It couldn’t be determined if systems that support foreign intelligence surveillance were also vulnerable in the breach.

«

You asked for an example demonstrating what people mean when they say that backdoor systems won’t only be used by “the good guys.”
unique link to this extract

Google must crack open Android for third-party stores, rules Epic judge • The Verge

Sean Hollister:

»

Google’s Android app store is an illegal monopoly — and now it will have to change.

Today, Judge James Donato issued his final ruling in Epic v. Google, ordering Google to effectively open up the Google Play app store to competition for three whole years. Google will have to distribute rival third-party app stores within Google Play, and it must give rival third-party app stores access to the full catalog of Google Play apps, unless developers opt out individually.

These were Epic’s biggest asks, and they might change the Android app marketplace forever — if they aren’t immediately paused or blocked on appeal.

And they’re not all that Epic has won today.

Starting November 1st, 2024, and ending November 1st, 2027, Google must also:
• Stop requiring Google Play Billing for apps distributed on the Google Play Store (the jury found that Google had illegally tied its payment system to its app store)
• Let Android developers tell users about other ways to pay from within the Play Store
• Let Android developers link to ways to download their apps outside of the Play Store
• Let Android developers set their own prices for apps irrespective of Play Billing

Google also can’t:
• Share app revenue “with any person or entity that distributes Android apps” or plans to launch an app store or app platform
• Offer developers money or perks to launch their apps on the Play Store exclusively or first
• Offer developers money or perks not to launch their apps on rival stores
• Offer device makers or carriers money or perks to preinstall the Play Store
• Offer device makers or carriers money or perks not to preinstall rival stores

«

Google’s downfall in this case – unlike Apple’s when Epic brought the same complaint – is that it has done all sorts of deals favouring one developer or another, whereas Apple is just consistently brutal to everyone.

However this won’t be an overnight change: Google gets eight months to come up with the system to implement this.
unique link to this extract

What happened to Elon Musk? • The Atlantic

Lora Kelley talks to Charlie Warzel:

»

Lora: Why is Musk getting so involved in this presidential election, and with Trump (who apparently said he would give Musk a role leading a government-efficiency commission if he wins)? Is he making some kind of play to be a great man of history, or is he after power in a potential Trump administration?

Charlie: Elon Musk basically bought Donald Trump at the top. He endorsed him moments after the first assassination attempt, when Trump was riding a wave of positive attention, when Joe Biden was still in the race and it looked like Trump was probably going to dominate him. So much has changed since Musk endorsed Trump in July. If he were truly a savvy political operator, he would be hedging his bets right now, saying I can’t fully alienate myself from one political party, because I have all these government contracts and so many other interests that I need to be able to at least sit in a room with with Democrats.

I think the fact that he has effectively just become the in-house social-media team for Donald Trump speaks to the fact that he’s not just making a political calculation. He’s not playing a game of 3-D chess. It seems to me that he’s truly radicalized.

Here’s a guy who has, like, six jobs and has decided to spend most of his time tweeting propaganda for a political candidate and hosting him on his platform. Does he want another job? It’s entirely possible. But I really think what he wants more than anything else is to be that sort of Rupert Murdoch person for this political group. He seems to be trying to fit himself into the role of power broker.

Lora: In some ways, Musk’s turn feels surprising. But has he always sort of been like this?

Charlie: I started covering Musk in the 2010s. And there were signs of this stuff—picking the fight with the cave diver, the way he would dismiss claims around Tesla, irresponsibly tweeting in ways that had the power to move stock prices. He was a loose cannon and showed a lot of signs of his disregard for the rule of law and authority.

«

Plus, as Warzel points out, there’s been a certain amount of audience capture, which means Musk has begun to perform for those who celebrate his behaviour, reinforcing it.
unique link to this extract

Fake News! The Top 100 Community Noted Twitter accounts • MeidasTouch News

Ron Filipkowski:

»

I wanted to note some other trends with the Worst 100. Many of the accounts are foreign, and most of those are from Asia. Many of them were permanently banned under the old regime but were reinstated by Musk. Many are not political – they just post fake stuff for clicks and focus on pop culture, the entertainment industry, or post false information about the weather, science, the environment, etc..

49 of the Top 100 worst offending accounts are overtly political. 48 of the 49 most Community Noted political accounts are right-wing. Only one – ‘Blade of the Sun’ is from the Left. Ironically, many of these accounts complain in their bios that they are “anti-fake news” or “anti-woke” while posting one lie and fake video after another. Six of the accounts have been permanently suspended, but for reasons other than posting false info. Most of the accounts have well over 100K followers and many over 1 million. Elon Musk follows many of them and has retweeted them often – including the Community Noted posts. Elon Musk himself checks in at #55 of the list of worst offenders with 89 posts getting Community Noted. 

Another big problem is that Musk has actually created an incentive structure to post fake things. Sensationalized claims get amplified by his algorithm and lots of clicks. There is financial incentive to continue to do it. You don’t face any suspensions and false posts tend to draw more engagement than true ones. Forty of the Worst 100 are monetized subscription accounts – so Musk is actually paying them to post fake things.

«

Included in that top 100 is the New York Post (“the only major American media company to make the top 100”) and, scraping in at No.96, ex-PM Rishi Sunak (“the most prominent politician to make the list”, where by “prominent” he means “had an important job”, because there are some right nutters in there).

Basically, all of the top 10 Most Noted should be on your blocklist. Perhaps the next 10 too. Worth pointing out that Community Notes predated Musk; it was called “Birdwatch” and launched in 2021.
unique link to this extract

How long will life exist on Earth? • The Atlantic

Ross Andersen:

»

Wikipedia’s “Timeline of the Far Future” is one of my favorite webpages from the internet’s pre-slop era. A Londoner named Nick Webb created it on the morning of December 22, 2010. “Certain events in the future of the universe can be predicted with a comfortable level of accuracy,” he wrote at the top of the page. He then proposed a chronological list of 33 such events, beginning with the joining of Asia and Australia 40 million years from now. He noted that around this same time, Mars’s moon Phobos would complete its slow death spiral into the red planet’s surface. A community of 1,533 editors have since expanded the timeline to 160 events, including the heat death of the universe. I like to imagine these people on laptops in living rooms and cafés across the world, compiling obscure bits of speculative science into a secular Book of Revelation.

Like the best sci-fi world building, the Timeline of the Far Future can give you a key bump of the sublime. It reminds you that even the sturdiest-seeming features of our world are ephemeral, that in 1,100 years, Earth’s axis will point to a new North Star. In 250,000 years, an undersea volcano will pop up in the Pacific, adding an extra island to Hawaii. In the 1 million years that the Great Pyramid will take to erode, the sun will travel only about 1/200th of its orbit around the Milky Way, but in doing so, it will move into a new field of stars. Our current constellations will go all wobbly in the sky and then vanish.

Some aspects of the timeline are more certain than others. We know that most animals will look different 10 million years from now. We know that the continents will slowly drift together to form a new Pangaea. Africa will slam into Eurasia, sealing off the Mediterranean basin and raising a new Himalaya-like range across France, Italy, and Spain. In 400 million years, Saturn will have lost its rings. Earth will have replenished its fossil fuels.

«

Wait a minute – replenished its fossil fuels with what? Or should that be who?
unique link to this extract

Apple shares trailer for ‘Submerged’ immersive Vision Pro short film • MacRumors

Juli Clover:

»

Announced back in July, Submerged is a short film that’s set in World War II, and it follows a group of sailors that are struggling to survive a deadly torpedo attack. It was created by Austrian filmmaker Edward Berger, who directed 2022 movie All Quiet on the Western Front.

Apple has not provided details on the length of Submerged, but most Apple Immersive Video content is on the shorter side. Apple has been regularly adding Immersive Video to the Vision Pro since the device came out last February. The immersive content is in 3D, and is meant to make the viewer feel like they are part of the scene.

Apple Immersive Video content can be viewed in the Apple TV app in Australia, Canada, Hong Kong, France, Germany, Japan, Singapore, the U.K., and the U.S. Users in China can watch the content through the Migu Video and Tencent Video apps.

Submerged is set to premiere on the Vision Pro on Thursday, October 10.

«

There’s a one-minute YouTube promo for the film (which is 17 minutes long) in the article. It’s hard to know what it would really be like, because if you don’t have a Vision Pro, you won’t know what immersive video is like. (Seems like they decided to take “immersive” seriously, with all the water.)

But good to see that Apple is actually trying to create some content for the Vision Pro.
unique link to this extract

As bird flu spreads, two new cases diagnosed in California • The New York Times

Apoorva Mandavilli:

»

Two more people were diagnosed with bird flu this week, even as scientists in Missouri continued to investigate a possible cluster of infections in that state, federal health officials said at a news briefing on Friday.

In California, two farmworkers who were exposed to infected dairy cattle at different farms tested positive for the virus, called H5N1, state health officials said on Thursday. Those cases bring the total this year to 16, not including those under investigation.

The cases do not come as a surprise, because the number of infected herds in California has risen to 56 from 16 two weeks ago, said Dr. Nirav Shah, the principal deputy director at the Centers for Disease Control and Prevention.

“As there are more herds that test positive, there are more workers who are exposed, and where there are more workers who are exposed, the chances of human infection increase,” he said. The risk to the public remains low, he added.

Still, experts said that the appearance of H5N1 in multiple states was worrisome.

«

Watching brief. That’s all it is, nothing else. Not at all.
unique link to this extract

Regeneron head says weight-loss drugs could cause “more harm than good” • Financial Times

Oliver Barnes:

»

The co-founder of Regeneron has warned that blockbuster weight-loss drugs could cause “more harm than good” unless the rapid muscle loss associated with the treatments is solved, as the US biotech pushes ahead with trials of muscle-preserving medicines.

Clinical studies suggest that patients treated with the new class of weight-loss drugs, known as GLP-1s, lose muscle at far faster rates than people losing weight from diet or exercise, exposing them to health problems, said George Yancopoulos, who also serves as Regeneron’s chief scientific officer.

For the two in every five patients who discontinue the treatments within a year, according to a 2024 JAMA study, this means that they are likely to rebound to their original weight with less muscle and a higher body fat percentage, “adding insult to injury”, said Yancopoulos.

“I do think that the GLPs should be viewed with a lot of concern in terms of the way they’re actually being used in the real world,” said Yancopoulos. “They could be leading to successive changes in body composition that could be creating more harm than good in the long term.”

Regeneron is among a growing list of drugmakers researching experimental drugs to preserve lean muscle mass in combination with GLP-1 drugs as a route into a potentially $130bn-a-year market that is dominated by Ozempic and Wegovy maker Novo Nordisk and Eli Lilly, the company behind Mounjaro and Zepbound.

Regeneron, a $111bn biotech that specialises in antibody treatments, is testing a drug called trevogrumab, which blocks the hormone myostatin, which limits muscle growth, in combination with Wegovy in mid-stage trials.

«

Damn. And we thought we were doing so well with peak obesity.
unique link to this extract

Uber and Lyft drivers use Teslas as makeshift robotaxis, raising safety concerns • Reuters

Akash Sriram and Abhirup Roy:

»

A self-driving Tesla carrying a passenger for Uber rammed into an SUV at an intersection in suburban Las Vegas in April, an accident that sparked new concerns that a growing stable of self-styled “robotaxis” is exploiting a regulatory gray area in US cities, putting lives at risk.

Tesla CEO Elon Musk aims to show off plans for a robotaxi, or self-driving car used for ride-hailing services, on Oct. 10, and he has long contemplated a Tesla-run taxi network of autonomous vehicles owned by individuals.

Do-it-yourself versions, however, are already proliferating, according to 11 ride-hail drivers who use Tesla’s Full Self-Driving (FSD) software. Many say the software, which costs $99 per month, has limitations, but that they use it because it helps reduce drivers’ stress and therefore allows them to work longer hours and earn more money.

Reuters is first to report about the Las Vegas accident and a related inquiry by federal safety officials, and of the broad use by ride-hail drivers of Tesla autonomous software.

While test versions of self-driving cabs with human backup drivers from robotaxi operators such as Alphabet’s Waymo and General Motors’ Cruise are heavily regulated, state and federal authorities say Tesla drivers alone are responsible for their vehicles, whether or not they use driver-assist software. Waymo and Cruise use test versions of software categorized as fully autonomous while Tesla FSD is categorized as a level requiring driver oversight.

«

It’s just a common-or-garden shunt though isn’t it. Whether or not the driver had “FSD” (which isn’t) on or not, if they’re in the driving seat, they’re responsible. By contrast the legal situation for the actual robotaxis operated by Waymo and GM is different.

But it does show that being a taxi driver is tedious and tiring. Of course you’d use driver assistance software if you could.
unique link to this extract

Police seldom disclose use of facial recognition despite false arrests • The Washington Post

Douglas MacMillan, David Ovalle and Aaron Schaffer:

»

Police departments in 15 states provided The Post with rarely seen records documenting their use of facial recognition in more than 1,000 criminal investigations over the past four years. According to the arrest reports in those cases and interviews with people who were arrested, authorities routinely failed to inform defendants about their use of the software — denying them the opportunity to contest the results of an emerging technology that is prone to error, especially when identifying people of color.

In fact, the records show that officers often obscured their reliance on the software in public-facing reports, saying that they identified suspects “through investigative means” or that a human source such as a witness or police officer made the initial identification.

In Evansville, Ind., for example, police said they identified a man who beat up a stranger on the street from his tattooed arms, long hair and previous jail booking photos. And in Pflugerville, Tex., police said they learned the name of a man who helped steal $12,500 in merchandise from Ulta Beauty “by utilization of investigative databases.”

Both of these suspects were identified with the aid of facial recognition, according to internal police records — information that was never shared with the accused, according to them or their attorneys. A spokeswoman for Pflugerville declined to answer questions about this case. Evansville police did not respond to requests for comment.

«

The problem is that there have been multiple wrongful arrests, including one person who spent six days in jail for using credit cards to buy things in a state he had never visited. (Quite hard to prove you’ve never been somewhere.)
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

New data suggests that obesity rates have fallen in the US – perhaps due to the new generation of GLP-1 agonists. CC-licensed photo by Tony Alter on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 10 links for you. Use them wisely. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

How to fight back against a traffic-less web • SparkToro

Rand Fishkin:

»

My position on this is that zero click is taking over everything. Google is trying to answer searches without clicks. Facebook is trying to keep people on Facebook. LinkedIn wants to keep people on LinkedIn.

Here’s another post. This is from Tim Soulo over at Ahrefs. Oh, look: 96.55% of web pages get zero traffic from Google.

And this is this is not super shocking to anyone, but what scares me is that this report that we did at SparkToro which looked at zero click searches in US and the EU was done in May, and June of this year looking at the five previous months. And you’ll recall Google had rolled back AI overviews, which hadn’t rolled out very far. But today, if you search for, for example, “my smoke alarm randomly went off”. Well, here’s this AI overview.

Last night I was searching for something related to my Dungeons and Dragons game. The AI overview takes over the whole page. Today Geraldine was searching for something related to lighting installation: AI takes over the whole page. There, you might argue well but look you could get a click here, or maybe someone will click on Reynolds restoration services.

I’m sorry friends, I think this is taking a tremendous amount more traffic than even what we measured back in June of this year. My friend Adam just put together this fireball x y z site, which is actually probably the best website I have ever read, or you will ever read, about finding the best smoke alarms out there. I can’t recommend it enough. Adam walked me through it and was like, what what do you think? I replied, I think that you have done a superb job, that your personal, deep dive into smoke alarms is second to none on the Internet, and that it will get absolutely no freaking traffic.

And what instead you are gonna have to do here is try and influence what these LLMs and Google is telling people. And to do that, you’re essentially gonna have to be in all the places where Google is pulling information from, which is a lot of these websites that rank in the top ten and all across the rest of the web.

«

The original page has lots of repeated little phrases, which I thought was a tactic to spot copying. In fact it’s just a machine-generated transcript of a little video. But this is the new reality: your (new) site won’t get traffic from Google.
unique link to this extract

I taught for most of my career. I quit because of ChatGPT • TIME

Victoria Livingstone:

»

In my most recent job, I taught academic writing to doctoral students at a technical college. My graduate students, many of whom were computer scientists, understood the mechanisms of generative AI better than I do. They recognized LLMs as unreliable research tools that hallucinate and invent citations. They acknowledged the environmental impact and ethical problems of the technology. They knew that models are trained on existing data and therefore cannot produce novel research. However, that knowledge did not stop my students from relying heavily on generative AI. Several students admitted to drafting their research in note form and asking ChatGPT to write their articles.

As an experienced teacher, I am familiar with pedagogical best practices. I scaffolded assignments. I researched ways to incorporate generative AI in my lesson plans, and I designed activities to draw attention to its limitations. I reminded students that ChatGPT may alter the meaning of a text when prompted to revise, that it can yield biased and inaccurate information, that it does not generate stylistically strong writing and, for those grade-oriented students, that it does not result in A-level work. It did not matter. The students still used it.

In one activity, my students drafted a paragraph in class, fed their work to ChatGPT with a revision prompt, and then compared the output with their original writing. However, these types of comparative analyses failed because most of my students were not developed enough as writers to analyze the subtleties of meaning or evaluate style. “It makes my writing look fancy,” one PhD student protested when I pointed to weaknesses in AI-revised text.

…I found myself spending many hours grading writing that I knew was generated by AI. I noted where arguments were unsound. I pointed to weaknesses such as stylistic quirks that I knew to be common to ChatGPT (I noticed a sudden surge of phrases such as “delves into”). That is, I found myself spending more time giving feedback to AI than to my students.

So I quit.

«

unique link to this extract

How Gemini successfully picked out my next read • Pocket Lint

Eli Becht:

»

When I’m in a reading slump, I like to go online and get suggestions for my next book. This includes asking people on Reddit or browsing through lists on apps like Goodreads, but I tried something different this time. I recently made the switch from Google Assistant to Google Gemini on my Android, and I decided to let it figure out what my next read was. It’s useful in the workplace, so why not reading suggestions?

Instead of letting it pick blindly, I told Gemini my favorite book is The Hobbit, and asked for five suggestions for what to read next. I’m as wary as they come when it involves AI, but I came away impressed with the response. Instead of coming back with a generic answer, Gemini backed up its suggestions with tidbits about the books and helped sell me on why I should choose one of them as my next read.

With a famous book like The Hobbit, you can easily guess one of the suggestions. Other than The Lord of the Rings, Gemini recommended The Princess Bride by William Golding, The Earthsea Cycle by Ursula K. Le Guin, Watership Down by Richard Adams, and The Chronicles of Narnia by C.S. Lewis.

Gemini sort of cheated by recommending an entire series, like Earthsea and Narnia, but it’s hard to argue with the results.

In fact, Gemini understood me so well that I have already read four of the five suggestions, and they are all sitting on my shelf. The only book I haven’t read is Watership Down, so I’ll need to add that one to my cart.

«

That’s a pretty terrible starting point, so I tried: favourite book Ringworld but I have read all the Ringworld series; favourite author Philip K Dick but I have read all the PKD books. ChatGPT recommended The Mote in God’s Eye (not bad, though I’ve read it), The Three-Body Problem (excellent but ditto) and The Stars My Destination by Alfred Bester (great, ditto).

Next time round it offered Blindsight by Peter Watts (never heard of it/him), The Algebraist by Iain M Banks (know him ofc) and Eon by Greg Bear (think I’ve read it). So not bad if you push it.
unique link to this extract

We may have passed peak obesity • Financial Times

John Burn-Murdoch:

»

Around the world, obesity rates have been stubbornly climbing for decades, if anything accelerating in recent years. But now newly released data finds that the US adult obesity rate fell by around two percentage points between 2020 and 2023.

We have known for several years from clinical trials that Ozempic, Wegovy and the new generation of diabetes and weight loss drugs produce large and sustained reductions in body weight. Now with mass public usage taking off — one in eight US adults have used the drugs, with 6% being current users — the results may be showing up at the population level.

While we can’t be certain that the new generation of drugs are behind this reversal, it is highly likely. For one, the decline is steepest among college graduates, the group most likely to be using them.

Crucially, the US National Health and Nutrition Examination Survey, which reported the unprecedented decline in obesity levels, uses weight and height measurements taken by medical examiners, not self-reported values. This makes it far more reliable than other surveys. American waistlines really do seem to be shrinking.

What makes this all the more remarkable is the contrast in mechanisms behind the respective declines in smoking and obesity. The former was eventually achieved through decades of campaigning, public health warnings, tax incentives and bans. With obesity, a single pharmaceutical innovation has done what those same methods have repeatedly failed to do.

If you take a step back, this is an astonishing achievement. Weight gain has proved far harder to combat than almost any other public health issue in history. Obesity has been such a formidable foe because everything is stacked against those trying to lose weight.

«

This is amazing. The BBC’s Today programme did an entire segment on it on Saturday morning, and we’re only just getting our heads around the implications of this.
unique link to this extract

New molecule can mimic the effects of fasting and exercise • Aarhus University

»

It is well known that regular exercise and periodic fasting have a series of positive effects on the body. Exercise and skipping meals makes for a stronger heart and reduces fat levels in the blood. The explanation lies in the body’s natural reaction in which increasing levels of lactate (the salt of lactic acid) and ketones act as efficient fuel for cells which benefit the body’s organs.

A group of chemistry, metabolism and diabetes researchers from Aarhus University has now created a molecule that can induce the same metabolic effects, without physical exertion or fasting.

The study has just been published in the scientific journal Journal of Agricultural and Food Chemistry.

“We’ve developed a molecule that can mimic the body’s natural metabolic response to strenuous exercise and fasting. In practice, the molecule brings the body into a metabolic state corresponding to running 10 kilometers at high speed on an empty stomach,” explains Professor Thomas Poulsen from the Department of Chemistry at Aarhus University. He is one of the leading researchers behind the study.

“When lactate and ketone levels in the blood increase, the production of an appetite-suppressing hormone increases and the level of free fatty acids in the blood decreases. This has a number of health benefits, for example reducing the risk of developing metabolic syndrome.”

…”It can be difficult to maintain motivation to run many kilometers at high speed and go without food. For people with physical ailments such as a weak heart or general weakness, a nutritional supplement can be the key to better recovery,” explains Poulsen.

«

This is mindblowing. Can be taken by mouth (screw you, Ozempic!) and already in human clinical trials.
unique link to this extract

Matt Mullenweg: ‘WordPress.org just belongs to me’ • The Verge

Emma Roth:

»

“WordPress.org just belongs to me personally,” Mullenweg said during an interview with The Verge. WordPress.org exists outside the commercial realm of Automattic, as a standalone publishing platform that offers free access to its open-source code that people can use to create their own websites. But it’s not a neutral, independent arbiter of the ecosystem. “In my role as owning WordPress.org, I don’t want to promote a company, which is A: legally threatening me and B: using the WordPress trademark. That’s part of why we cut off access from the servers.”

Mullenweg’s feud with WP Engine fans out in a few different directions. He’s criticized WP Engine for not putting enough time and money into developing the open-source WordPress ecosystem, saying that if you gave $1 to the WordPress Foundation, “you’d be a bigger donor than WP Engine.” And Mullenweg has brought up the possibility that WP Engine “hacked” the Automatic-owned WooCommerce plug-in to collect commissions meant for Automattic, which WP Engine has denied. From those arguments, the fight appears to be one over what is and isn’t appropriate in the open-source software world.

But Mullenweg has since sidelined those arguments to make the case that WP Engine — and its “hacked up, bastardized simulacra” of the WordPress open-source code, as he describes it — is infringing on Automattic’s trademark: WordPress.

«

At this point lawsuits are being filed, and one has to think it’s not wise of Mullenweg to try to take on an organisation (WP-Engine) which is backed by a hedge fund worth billions.
unique link to this extract

Stop asking people “What do you do?” • WSJ

Joanne Lipman:

»

For people who have taken career breaks, four little words—“What do you do?”—can provoke dread. It seems to conceal a bundle of judgments: What’s your social status? What’s your income? What’s your education? Are you worth my time to talk to?

This has long been an issue for professional women who leave the workforce to raise kids. They describe feeling invisible and being ignored by people they meet. But the dreaded question is now affecting a wider swath of people: stay-at-home fathers, career-changing young people, gig workers, baby boomers forced into retirement and laid-off workers.

“It is truly the absolute worst question you can get when you’re out of work. Society wants to put you in an easy-to-digest box,” says Orlando-based Jen Kling, 40, a consumer brand marketer who has been laid off three times and is now an independent consultant.

It’s also a head-scratcher when trying to frame an answer. When New York entrepreneur James Reichert, 62, moved to Canada temporarily for his then-wife’s job, he printed up business cards that read “Trophy Husband.” When Ashley Scott, 35, a Philadelphia corporate sustainability manager, was laid off from a previous job, she took to telling people she was in grad school. She found that when she said “I’m looking for a job, or I just got laid off… People would look at you like you’re a loser.”

For many of us, work isn’t just a way to pay for our lives; it’s how we define ourselves—and others. We are what we do. Psychologists have a term for this: “enmeshment.” The concept was first coined to describe an unhealthy blurring of boundaries in personal relationships. But it applies with almost absurd accuracy to our relationship with work, when we are so closely linked to our careers that we have no idea who we are without them.

«

unique link to this extract

Messages via satellite provides lifeline to iPhone users in Hurricane Helene fallout • 9to5Mac

Ryan Christoffel:

»

Hurricane Helene has caused massive damage and taken over 100 lives across several US states. Many thousands of people are without power and/or cell service. But in the wake of the storm, reports have surfaced about a key iOS 18 feature that has been a lifeline for survivors: Messages via satellite.

Apple added Messages via satellite to millions of iPhones via its recent iOS 18 update. And now, according to reports on social media, it seems the feature arrived just in time.

Here are a few tweets highlighting how useful the feature has proven: Asheville resident;
father contacting son; North Carolina resident.

A common message across social media around the time Helene hit hardest was a call for users to update to iOS 18 so they’ll gain access to this feature. Apple notably shipped iOS 18 and iOS 17.7 simultaneously, leading many to stick with 17 for now.

«

Perhaps I wasn’t the target audience, but I hadn’t heard about iOS 18 offering Messages via satellite. Seems worthwhile. (I still haven’t upgraded to iOS 18; not sure what there is that’s compelling.)
unique link to this extract

Quantum advantage for NP approximation? For REAL this time? • Shtetl-Optimized

Scott Aaronson:

»

The other night I spoke at a quantum computing event and was asked—for the hundredth time? the thousandth?—whether I agreed that the quantum algorithm called QAOA was poised revolutionize industries by finding better solutions to NP-hard optimization problems. I replied that while serious, worthwhile research on that algorithm continues, alas, so far I have yet to see a single piece of evidence that QAOA outperforms the best classical heuristics on any problem that anyone cares about. I added I was sad to see the arXiv flooded with thousands of relentlessly upbeat QAOA papers that dodge the speedup question by simply never raising it at all. I said that, in my experience, these papers reliably led outsiders to conclude that surely there must be excellent known speedups from QAOA—since otherwise, why would so many people be writing papers about it?

Anyway, the person right after me talked about a “quantum dating app” (!) they were developing.

«

All the QAOA goes right over my head, but I’m very intrigued by the quantum dating app. Is the idea that you’re compatible with everyone until you meet them? Or (next two © Oliver Johnson) you swipe both right and left? You know where to meet your match but not how fast to go with them?

Your suggestions for how the quantum dating app works welcome.
unique link to this extract

Apple releases Depth Pro, an AI model that rewrites the rules of 3D vision • VentureBeat

Michael Nuñez:

»

Apple’s AI research team has developed a new model that could significantly advance how machines perceive depth, potentially transforming industries ranging from augmented reality to autonomous vehicles.

The system, called Depth Pro, is able to generate detailed 3D depth maps from single 2D images in a fraction of a second—without relying on the camera data traditionally needed to make such predictions.

The technology, detailed in a research paper titled “Depth Pro: Sharp Monocular Metric Depth in Less Than a Second,” is a major leap forward in the field of monocular depth estimation, a process that uses just one image to infer depth.

This could have far-reaching applications across sectors where real-time spatial awareness is key. The model’s creators, led by Aleksei Bochkovskii and Vladlen Koltun, describe Depth Pro as one of the fastest and most accurate systems of its kind.

Monocular depth estimation has long been a challenging task, requiring either multiple images or metadata like focal lengths to accurately gauge depth.

But Depth Pro bypasses these requirements, producing high-resolution depth maps in just 0.3 seconds on a standard GPU. The model can create 2.25-megapixel maps with exceptional sharpness, capturing even minute details like hair and vegetation that are often overlooked by other methods.

«

What is a “standard GPU”? Though 0.3 seconds doesn’t sound too shabby, it’s probably not quick enough for driving.. is it?
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Increasing the number of steps you take each day is definitely correlated with living longer, and if you raise them you benefit. CC-licensed photo by Timo Newton-Syms on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

It’s Friday, so there’s another post due at the Social Warming Substack at about 0845 UK time. It’s about hyperbole. Get excited!

A selection of 10 links for you. Stepping out. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Someone put facial recognition tech onto Meta’s smart glasses to instantly dox strangers • 404 Media

Joseph Cox:

»

A pair of students at Harvard have built what big tech companies refused to release publicly due to the overwhelming risks and danger involved: smart glasses with facial recognition technology that automatically looks up someone’s face and identifies them. The students have gone a step further too. Their customized glasses also pull other information about their subject from around the web, including their home address, phone number, and family members. 

The project is designed to raise awareness of what is possible with this technology, and the pair are not releasing their code, AnhPhu Nguyen, one of the creators, told 404 Media. But the experiment, tested in some cases on unsuspecting people in the real world according to a demo video, still shows the razor thin line between a world in which people can move around with relative anonymity, to one where your identity and personal information can be pulled up in an instant by strangers.

Nguyen and co-creator Caine Ardayfio call the project I-XRAY. It uses a pair of Meta’s commercially available Ray Ban smart glasses, and allows a user to “just go from face to name,” Nguyen said.

The demo video posted to X on Tuesday shows the pair using the tech against various people. In one of the first examples, Ardayfio walks towards the wearer. “To use it, you just put the glasses on, and then as you walk by people, the glasses will detect when somebody’s face is in frame,” the video says. “After a few seconds, their personal information pops up on your phone.”

«

It’s very impressive. Unexpectedly, the effect of being able to recognise people and get their context is that you seem to make a lot more friends, or get friendly reactions. People like being recognised and having their achievements mentioned. After all, who wouldn’t?

Give it time: this is going to get faster, more accurate, and the samizdat will become everyday.
unique link to this extract

No man’s hand: artificial intelligence does not improve police report writing speed • Journal of Experimental Criminology

Ian Adams, Matt Barter, Kyle McLean, Hunter Boehme and Irick Geary at various US universities:

»

Objectives: This study examines the potential of artificial intelligence (AI) to reduce the time police officers spend writing reports, a task that consumes a significant portion of their workday.

Methods: In a pre-registered randomized controlled trial, we test this claim within the patrol division of a medium-sized police department (n = 85) at the individual report level (n = 755). Analyses utilize mixed-effects regression accounting for the nested structure of report-writing.

Results: AI assistance did not significantly affect the duration of writing police reports. Alternative specifications beyond those specified in the pre-registration, including a difference-in-differences approach observing report duration over a full year (n = 6084), confirm the null findings are robust.

«

That’s going to be a disappointment for the police in Colorado who thought chatbots would be good for this job.
unique link to this extract

Why the US can’t impose its will over global trade in electric cars • Financial Times

Alan Beattie:

»

middle-income countries such as Turkey and Brazil wanting to increase domestic EV consumption are actively courting Chinese producers.

And even taking into account the protectionist motive, Joe Biden’s administration may well have a point about the security threats of EVs as “smartphones on wheels”, with manufacturers able to collect personal data and potentially control the cars remotely. But this is an unpropitious environment for the American sheriff to stick up “WANTED FOR DATA RUSTLING” posters around the place and try to run Chinese producers out of town.

The lure of US market access, via which Washington traditionally exerts control over other countries’ trade and tech policies, is weaker than it ought to be. American consumer preferences and the domination of the Detroit carmakers have left the US EV market pitifully under-developed. EVs in 2023 had a 10% share of total sales compared with 38% in China and 21% in the EU, and even the EV tax credits in Biden’s Inflation Reduction Act have so far had only limited effect.

EV prices relative to traditional vehicles in the US market are higher than in China and the EU, and Washington has reduced competitive pressure by walling off its market to Chinese exporters with 100% tariffs.

…Strict implementation might simply force carmakers to create a separate North American supply chain with non-Chinese software. In that case, Dunne says, the global car market could divide in two: a high-priced low-tech island comprising the US and Canada and a cheaper, more digitally connected market for the rest of the world. (Mexico, which is part of the US-Canada trade bloc but also exports cars outside it, would probably straddle the two.)

It’s somewhat against the historical grain for US companies to be behind on technology and its households weak on consumption. But that’s where we’ve ended up with EVs.

«

unique link to this extract

Israel could bomb Iran’s oil. Energy markets aren’t panicking • POLITICO

Ben Lefebrvre:

»

The risk of an escalating war between Israel and Iran is testing the global market’s faith that crude oil prices would be insulated from a widening of hostilities across the Middle East.

For decades, conflicts in the oil-rich region frequently spooked oil markets and weighed on the economy. But now, Middle East military skirmishes are causing more shrugs than drastic price spikes — a welcome development for the Biden administration, which has faced political criticism from Republicans over fuel prices and is trying to contain the fallout from Iran’s launch of nearly 200 missiles into Israel on Tuesday.

Increased oil production from the United States, Brazil and other places in the past two decades has diversified the global fuel supply, which means oil markets rely less on Middle East shipments that Tehran could disrupt, energy and security analysts told POLITICO.

“For those of us who spend our lives looking at the effects of a [Middle East] crisis on oil prices, obviously the past 10-plus years have been a complete washout,” said Michael Knights, an analyst at the think tank The Washington Institute for Near East Policy. “No matter how insane the thing is, it has a minimal impact on oil. The market has proven time and time again it can make up shortfalls.”

«

Whatever this is, it’s not 1973 all over again: we’re not going to see those queues at the fuel pumps.
unique link to this extract

The Before and After – Columbia Journalism Review

Lauren Watson on the effect of Facebook removing news links in Canada:

»

Meta’s retreat from journalism didn’t stop Canadians from seeking out news—but it did prevent them from finding information from legitimate sources. According to a report by the Media Ecosystem Observatory, a research collaboration between McGill University and the University of Toronto, in the year since the ban went into effect, Canadians have seen less reporting online, even as they continue to use Meta to read, watch, and listen to news: 70% of survey respondents do so on Facebook, 65% on Instagram.

Some of that can be explained by screenshots of articles, which tripled in frequency in the four months following the ban [by Facebook after a Canadian law demanding payment if it included news links]. But the researchers also found that only 22% of Canadians are aware that Meta has bailed on journalism. That has turned Canadian newsgathering on social media into a game of telephone—out-of-context photos and summaries absent links to the articles from which they’ve been sourced—that few even know is being played.

“It would be one thing if they made the absence clear, but they went from blocking the news to facilitating the bamboozling of the news,” David Beers, the founding editor of The Tyee, told me. “If you were an old-fashioned Orwellian dictator, you couldn’t come up with a more clever plan.”

When fall arrived, Meta’s news ban faced its first major test, in Canada’s worst wildfire season to date. In British Columbia, more than three hundred and eighty fires burned; some twenty thousand people were placed under evacuation orders. Canadian officials observed that public service announcements were failing to get around.

“I find it astonishing,” David Eby, the province’s premier, said in a press conference, “that we are at this stage of the crisis and the owners of Facebook and Instagram have not come forward and said, ‘Look, we’re trying to make a point with the federal government, but it’s more important that people are safe, it’s more important that they have access to basic information through our networks, and then we can deal with our concerns with the federal government and their new laws later.’” Justin Trudeau, Canada’s prime minister, complained that Meta’s actions were “inconceivable.”

«

unique link to this extract

Here’s an easy way our trains could usefully connect us • The Times

Tom Whipple:

»

Over the years, there has not been a lot about my daily commute that has been consistent. There have been times during engineering works when I have not been certain when my morning train will arrive. There have been times during strikes when I have not been certain when, or even whether, it will leave.

What has been consistent though, through Covid, floods and industrial action, has been the knowledge that if I’ve forgotten to download an important document beforehand, then the journey will be even more frustrating. Because, despite the jaunty promises of the train operator, I know that for much of the journey the supposed train Wi-Fi is unlikely to help much. And it won’t help at all if I happen to be going past Ascot.

Party conference season has come to an end. We have been told we must face hard choices if we are to achieve growth. We have seen depressing graphs of what our productivity rates have become because those hard choices have gone unmade.

Here is my idea for getting a bit of growth, without hard choices. It won’t involve scarring the Cotswolds countryside or spending Covid-style billions on bat surveys and newt tunnels so that we can finally build a new road. It just involves making the wretched train Wi-Fi work — like it does in other countries.

Each morning I look across a carriage full of open laptops. I see web pages failing to open, emails failing to send. I see people trying to use their mobile hotspots but doing little better. Most of all I see people — who often look like they are paid pretty high hourly rates — trying, and failing, to work.

This is not just my commute. There are many subjects on which journalists are ill-equipped to speak but there is one on which we are world experts: being sent to random parts of the country and trying to work on trains. So believe me when I say that this is a problem everywhere.

This year a report in The Sunday Times revealed that our train wifi network is so dilapidated that operators are being forced to ration access. It also reported that the system could be upgraded across the country for £200m.

«

Which really isn’t a lot of money in the scheme of things.
unique link to this extract

The tragedy of the commons is a false and dangerous myth • Aeon Essays

Michelle Nijhuis:

»

In December 1968, the ecologist and biologist Garrett Hardin had an essay published in the journal Science called ‘The Tragedy of the Commons’. His proposition was simple and unsparing: humans, when left to their own devices, compete with one another for resources until the resources run out. ‘Ruin is the destination toward which all men rush, each pursuing his own best interest,’ he wrote. ‘Freedom in a commons brings ruin to all.’ Hardin’s argument made intuitive sense, and provided a temptingly simple explanation for catastrophes of all kinds – traffic jams, dirty public toilets, species extinction. His essay, widely read and accepted, would become one of the most-cited scientific papers of all time.

Even before Hardin’s ‘The Tragedy of the Commons’ was published, however, the young political scientist Elinor Ostrom had proven him wrong. While Hardin speculated that the tragedy of the commons could be avoided only through total privatisation or total government control, Ostrom had witnessed groundwater users near her native Los Angeles hammer out a system for sharing their coveted resource. Over the next several decades, as a professor at Indiana University Bloomington, she studied collaborative management systems developed by cattle herders in Switzerland, forest dwellers in Japan, and irrigators in the Philippines. These communities had found ways of both preserving a shared resource – pasture, trees, water – and providing their members with a living. Some had been deftly avoiding the tragedy of the commons for centuries; Ostrom was simply one of the first scientists to pay close attention to their traditions, and analyse how and why they worked.

«

Except.. what Ostrom demonstrated was that if there are community safeguards and punishments for overuse, then the commons won’t be overexploited. Without that, we see the TOTC (paper available here) occur again and again: overfishing, pollution, even climate change. (The original paper itself is a somewhat Malthusian treatise on the limited planet; it reads oddly in the modern context.)
unique link to this extract

The iPhone content machine: a visual essay • On my Om

Om Malik:

»

Apple’s iPhone 16 launch event differed greatly from most of its past events. It was larger and more overwhelming. There were fewer familiar faces among the attendees, and there was also a new type of attendee — content creators. They were busy filing short bursts of information to their followers in vertical formats: videos, selfies at Apple Park and occasional comments about the products themselves.

I decided to become a fly on the wall and chronicle the spectacle unfolding in front of me. I focused on those who were there to create content about the devices, not the devices themselves. It was fun to just float among the crowds with my Nikon Zf and a 40mm lens.

It was a wonderful spectacle — just to bask in this new kind of raw media energy. Content for the sake of content. Events for the sake of content. Fog of content. It’s the new way of the world.

«

Malik decided just to take some photos, and very good photos they are. (The big closeup is John Gruber, of Daring Fireball.) One person – not Gruber – commented that the attendees this year seemed essentially clueless about Apple, and about the executives they were talking to who tend to have long histories at the company:

»

My feeling is that they “saw” the keynote but didn’t actually “watch” it. It’s the same difference as “hearing” music versus “listening” to it. They’re more focused on how to later take selfies next to the new products, not necessarily think deeply about why the products were created and what impact they could have.

«

unique link to this extract

Craft and creativity • The Bookseller

Nadim Sadek:

»

There are less perceivably creative people in our world. The architects who design perfect arches. Whoever invented the wheel. The master-distiller, blending liquids in casks of sherry and port to make that perfect single-malt. Or a nurse who finds a way to make an old woman comfortable by playing her songs from her childhood. These days, also the TikToker who produces a new meme, combining a societal insight with a memorable tune and perhaps a signature dance. 

Each human is creative. But not each human can craft, whether it’s with paintbrushes, words or filters on a social-media site.

AI solves this. It’s not a Stradivarius. It’s not a Porsche. It’s not squirrel-hair brush. But it is a new expresser, a means of fashioning an artefact from a creative impulse without having to master the craft of expression.

So long as you can articulate your notion, AI can make a decent stab at producing an artefact to represent your creativity. It’ll make music to your command. Write words. Produce an image. Whatever you’re trying to conceive and give birth to, AI disintermediates the historic imperative of “crafting”. It takes your ideas and makes them evident. Others can see what you intend. People can relate to what you wish to convey. 

…If you’re reading this, you’re likely either a crafter, or someone involved in the craft-trade, including book publishing. AI is challenging the status quo. And it’s a positive thing.

«

I’m really not sure that this is a view widely shared by those who have to decide what to publish.
unique link to this extract

Effect of daily steps and sedentary time on death and cardiovascular risk • Kudos

»

The study (in the British Journal of Sports Medicine, March 2024) suggested that an increase in the number of daily steps is associated with a lower risk of both death and cardiovascular disease (CVD). Here, the greatest benefit was observed between 9,000 and 10,500 steps per day. This optimal range for lowering health risks remains consistent regardless of whether a person has high or low levels of sedentary time. Specifically, individuals with high sedentary time (more than 10.5 hours per day) had a higher risk of death if they walked fewer steps, compared to those with lower sedentary time. Also, even a modest increase in daily steps (between 4,000 and 4,500 steps) can significantly lower the risk of death and CVD.

Increasing daily steps to around 9,000 to 10,500 can significantly lower the risk of death and CVD, independent of sedentary time. Even a small increase in daily steps can have a positive impact on health, and reducing sedentary time further improves these benefits.

«

OK, no excuse now! Though Strava data suggests that those who do 9,000 steps per day are in the top 3% of walkers. The fact that increasing steps is helps seems obvious – it’s exercise. But this has come from “device data” – hip-worn accelerometers worn over the course of three years (2013-2015) by 100,000 participants in the UK aged between 40 and 69.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

There’s a move away from touchscreens as a general user interface for specific tasks in cars and even phones. CC-licensed photo by JC on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 9 links for you. Untouchable. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

A network of AI ‘Nudify’ sites are a front for notorious Russian hackers • 404 Media

Joseph Cox:

»

Multiple sites which promise to use AI to ‘nudify’ any photos uploaded are actually designed to infect users with powerful credential stealing malware, according to new findings from a cybersecurity company which has analyzed the sites. The researchers also believe the sites are run by Fin7, a notorious Russian cybercrime group that has previously even set up fake penetration testing services to trick people into hacking real victims on their behalf.

The news indicates that services for producing AI-generated nonconsensual intimate content are becoming enticing enough that hackers feel it is worth the time and effort to build fake versions they can then use to hack people. The news also shows that Fin7 is alive despite the U.S. Department of Justice saying last year that “Fin7 as an entity is no more.”

Hostinger, the domain registrar for most of the fake nudify sites, blocked the domains after 404 Media sent it a list of questions earlier this week. 404 Media also found that one of the Fin7-run sites was included one of the web’s biggest porn site aggregators, potentially putting many people who stumbled across the site at risk.

“The deepfake AI software may have an audience of mostly men with a decent amount who use other AI software or have crypto accounts,” Zach Edwards, senior threat analyst at cybersecurity firm Silent Push, told 404 Media in an online chat.

«

Honeytraps: old ploy, modern method.
unique link to this extract

Staying a step ahead: mitigating the DPRK IT worker threat • Google Cloud Blog

Codi Starks, Michael Barnhart, Taylor Long, Mike Lombardi, Joseph Pisano, and Alice Revelli:

»

UNC5267 is not a traditional, centralized threat group. IT workers consist of individuals sent by the North Korean government to live primarily in China and Russia, with smaller numbers in Africa and Southeast Asia. Their mission is to secure lucrative jobs within Western companies, especially those in the US tech sector.

UNC5267 gains initial access through the use of stolen identities to apply for various positions or are brought in as a contractor. UNC5267 operators have primarily applied for positions that offer 100% remote work. Mandiant observed the operators engaging in work of varying complexity and difficulty spanning disparate fields and sectors. It is not uncommon for a DPRK IT worker to be working multiple jobs at once, pulling in multiple salaries on a monthly basis. One American facilitator working with the IT workers compromised more than 60 identities of US persons, impacted more than 300 US companies, and resulted in at least $6.8m of revenue to be generated for the overseas IT workers from in or around October 2020 until October 2023.

…Mandiant has identified a substantial number of DPRK IT worker resumes used to apply for remote positions. In one resume from a suspected IT worker, the email address—previously observed in IT worker-related activities—was also linked to a fabricated software engineer profile hosted on Netlify, a platform often used for quickly creating and deploying websites. The profile claimed proficiency in multiple programming languages and included fake testimonials with stolen images from high-ranking professionals, likely stolen from CEOs, directors, and other software engineers’ LinkedIn profiles.

…To accomplish their duties, UNC5267 often remotely accesses victim company laptops situated within a laptop farm. These laptop farms are typically staffed with a single facilitator who is paid monthly to host numerous devices in one location. Mandiant has identified evidence that these laptops are often connected to an IP-based Keyboard Video Mouse (KVM) device, although a recurring theme across these incidents is the installation of multiple remote management tools on victim corporate laptops immediately following shipment to the farm. These indicate that the individual is connecting to their corporate system remotely via the internet, and may not be geographically located in the city, state, or even country in which they report to reside.

«

They’re also not very good at the programming jobs. Which shouldn’t surprise you. But able to work multiple jobs at once? Impressive!
unique link to this extract

Touch screens are over. Even Apple is bringing back buttons • WSJ

Christopher Mims:

»

The tyranny of touch screens may be coming to an end.

Companies have spent nearly two decades cramming ever more functions onto tappable, swipeable displays. Now buttons, knobs, sliders and other physical controls are making a comeback in vehicles, appliances and personal electronics.

In cars, the widely emulated ultra-minimalism of Tesla’s touch-screen-centric control panels is giving way to actual buttons, knobs and toggles in new models from Kia, BMW’s Mini, and Volkswagen, among others. This trend is delighting reviewers and making the display-focused interiors of Tesla and its imitators feel passé.

Similar re-buttonization is occurring in everything from e-readers to induction stoves.

Perhaps the most prominent exponent of this button boom is the company that set us lurching toward touch screens in the first place. Apple added a third button it calls the “action button” to its full slate of new iPhone 16s unveiled this month, after introducing the feature on its upscale Apple Watch Ultra and Pro-model iPhones over the past couple of years. It also added a button-like “camera control” input on the iPhone’s side.

As Apple shows, companies aren’t just rediscovering buttons, they’re reconceiving them. The camera control includes touch features, and the company has also developed the “force sensor” that enables its AirPods to respond when you squeeze their stems.

«

Pendulum swings are probably more common in technology than we realise. Unbundle! Then: bundle again! Make things modular! Then… make things integrated! Replace buttons with touchscreens! Then.. actually, buttons work better while we still have fingers.
unique link to this extract

BBC confronts neo-Nazi who gave UK rioters arson tips • BBC News

Ed Thomas:

»

The BBC has confronted a neo-Nazi in Finland who shared online instructions on how to commit arson with UK rioters during the summer.

The 20-year-old was an administrator in the Southport Wake Up group on the Telegram messaging app, where he was known as “Mr AG”. He posted the arson manual, which was pinned to the top of the group chat.

In late July and early August, the group was key in helping to organise and provoke protests that turned to violence in England and Northern Ireland.

We tracked Mr AG – whose real name is Charles-Emmanuel Mikko Rasanen – to an apartment on the outskirts of the Finnish capital, Helsinki. It was from here, more than 1,000 miles away from Southport, that the neo-Nazi took a prominent online role during the UK riots.

On 29 July, within hours of the killings of three young girls at a Taylor Swift-themed dance class in Southport, the Southport Wake Up group was created. Within days it had grown to more than 14,000 members. Mr Rasanen – or Mr AG as he was known online – helped to run the group chat.

The group organised the very first protest in the UK, on St Luke’s Road in Southport, the day after the killings. That protest later turned into a riot. Before the group was taken down by Telegram, a series of other protest locations were advertised, as well as a list of dozens of refugee centres, suggested as potential targets.

…The BBC travelled to Finland to confront Mr Rasanen – we had previously emailed him. He refused to answer any of our questions, but did not deny sending the posts or being an administrator of the Southport Wake Up group.

Before we left him, he also accused the BBC of harassment and rang the police.

«

Totally weird: he’s a mixed-race Finn who celebrates Hitler. Neat work tracking him down, which seems to have been done by Finnish investigative journalists: very Girl With The Dragon Tattoo of them.
unique link to this extract

More than 9,000 scam Facebook pages deleted after Australians lose $43.4m to celebrity deepfakes • The Guardian

Josh Taylor:

»

Australians could see fewer deepfake images of celebrities being hauled off in handcuffs, or promoting a fraudulent cryptocurrency investment on Facebook, after Meta launched a new one-stop shop for banks to share information on scams that has blocked 8,000 pages and 9,000 celebrity scams in its first six months of operation.

From January to August 2024, Australians reported $43.4m in losses from scams on social media to Scamwatch, with close to $30m relating to fake investment scams.

Meta, the parent company of Facebook and Instagram, has faced pressure from politicians and regulators in the past few years to tackle the plague of scams featuring deepfake images of public figures such as David Koch, Gina Rinehart, Anthony Albanese, Larry Emdur, Guy Sebastian and others which are used to promote investment scams.

The company is being sued by the mining magnate Andrew Forrest over the company’s alleged failure to tackle scams using his image.

Meta announced on Wednesday it had partnered with the Australian Financial Crimes Exchange (AFCX) to launch the Fraud Intelligence Reciprocal Exchange (Fire) that provides a dedicated reporting channel for scams between Meta and financial providers of the victims of the scams.

…Since launching a pilot in April, there have been 102 reports, resulting in Meta removing more than 9,000 scam pages, and 8,000 AI-generated celebrity investment scams on Facebook and Instagram.

«

Those celebrity fake ads are all over Twitter, but I doubt that they’re going to be taken down with anything like the same alacrity. It’s hardly worth celebrities suing Twitter, since it will just tie them up in court, and Musk has more money than they do. No obvious solution if the platform doesn’t see it as important.
unique link to this extract

The coolest thing about smart glasses is not the AR. It’s the AI • MIT Technology Review

Mat Honan:

»

when I tried Snap’s new Spectacles a couple of weeks ago, I was less taken by the ability to simulate a golf green in the living room than I was with the way I could look out on the horizon, ask Snap’s AI agent about the tall ship I saw in the distance, and have it not only identify it but give me a brief description of it. Similarly, in The Verge Alex Heath notes that the most impressive part of Meta’s Orion demo was when he looked at a set of ingredients and the glasses told him what they were and how to make a smoothie out of them.

The killer feature of Orion or other glasses won’t be AR Ping-Pong games—batting an invisible ball around with the palm of your hand is just goofy. But the ability to use multimodal AI to better understand, interact with, and just get more out of the world around you without getting sucked into a screen? That’s amazing.

And really, that’s always been the appeal. At least to me. Back in 2013, when I was writing about Google Glass, what was most revolutionary about that extremely nascent face computer was its ability to offer up relevant,  contextual information using Google Now (at the time the company’s answer to Apple’s Siri) in a way that bypassed my phone.

While I had mixed feelings about Glass overall, I argued, “You are so going to love Google Now for your face.” I still think that’s true.

Assistants that help you accomplish things in the world, without having to be given complicated instructions or making you interface with a screen at all, are going to usher in a new wave of computing. While Google’s Project Astra demo, a still unreleased AI agent that it showed off this summer, was wild on a phone, it was not until Astra ran on a pair of smart glasses that things really fired up.

«

That’s always been the obvious use of smart glasses. Especially now we have AI to integrate to it.
unique link to this extract

How Hurricane Helene became a monster storm – The Verge

Justine Calma:

»

It made landfall with winds reaching 140 miles per hour, making it a major storm and a Category 4 out of 5 on the Saffir-Simpson hurricane scale.

Helene packed a punch with water, too. When it hit Florida’s Big Bend region, it brought a massive storm surge, inundating the coastline with up to 15 feet of seawater. The underwater topography off Florida’s west coast, with a more gradual incline, acted like a ramp, making it easier for the storm to bring a taller wall of water with it. The sheer size of the hurricane also meant that the storm surge flooded a wider area.

Heavy rainfall dropped more water onto communities, leading to historic flooding in western North Carolina. Close to 14 inches of rain were recorded at the Asheville airport over three days between September 25th and 27th. The highest preliminary total was more than 31 inches of rain, recorded in Busick, North Carolina.

…Climate change is altering the calculus for storms like Helene. Rising global temperatures create conditions conducive to more intense storms that can gain strength quickly and stay more powerful onshore. Helene developed amid soaring sea surface temperatures in the Caribbean Sea and the Gulf of Mexico. Waters along the storm’s early path got as high as 31ºC (87.8ºf), providing ample fuel. The atmosphere’s ability to hold moisture is increasing because of greenhouse gas emissions from fossil fuels, allowing for more severe downpours.

To know how big of a role climate change played with Helene specifically, scientists will have to conduct more research. But Balaguru likens the effect of climate change to the world having a weakened immune system. “It doesn’t mean that you will become sick. It just increases your tendency to become sick,” Balaguru says.

Altogether, the pieces were in place for the perfect storm with Helene. “The storm started big, which was bad, it went over hot water, which was bad, it hit a place that is prone to high storm surge, and then it accelerated and went into populated areas and took wind and rainwater to those populated areas,” Knox says. “You don’t want to see much worse.”

«

unique link to this extract

Oura nears $500m in annual revenue and readies new ring • Bloomberg via MSN

Mark Gurman and Evan Gorelick:

»

Oura Health Oy, a Finnish health technology company known for its fitness-tracking rings, will see annual sales double this year to roughly $500m and expects “healthy” growth in 2025.

Chief executive officer Tom Hale, speaking in an interview, said that Oura is building a loyal following after selling more than 2.5 million rings. Still, the company isn’t yet at the stage of planning an initial public offering, he said.

The 11-year-old business, which pioneered the concept of finger-worn activity trackers, makes its money by selling rings for $299 and subscriptions priced at $6 a month. It’s more of a niche market than smartwatches or earbuds, but the field is getting more crowded. Samsung Electronics Co. recently launched a $400 product called the Galaxy Ring.

Hale is upbeat about expanding the business. The company’s profit margins are closer to that of a software company than a hardware maker, he said, and Oura’s subscribers have been sticking with the product.

“Retention is better than any other subscription model I’ve seen,” Hale said. “To double this business, we don’t have to do that much.” He said that the company’s roughly half a billion dollars in revenue for calendar 2024 would be twice what it recorded in 2023.

Though Oura declined to discuss future products, people with knowledge of its plans say the company is introducing a fourth-generation ring in October. The device will have a thinner design and better battery life, as well as more accurate activity tracking, they said. It’s set to be the company’s biggest product overhaul in three years.

In addition to tracking fitness, Oura rings assess the quality of a user’s sleep and provide a “readiness score.” About 80% of Oura’s revenue comes from hardware, with the rest provided by software subscriptions, Hale said

«

Neil Cybart (via whom this comes) reckons half of those 2.5 million rings will be sold this year. I just can’t imagine it as a mass market thing.
unique link to this extract

Windows MR headsets no longer work in Windows 11 24H2 • UploadVR

David Heaney:

»

Microsoft has removed Windows Mixed Reality from Windows 11.

With Windows 11 24H2, the latest major version of Microsoft’s PC operating system, you can no longer use a Windows MR headset in any way – not even on Steam.

This includes all the Windows MR headsets from Acer, Asus, Dell, HP, Lenovo, and Samsung, including HP’s Reverb G2, released in 2020.

UploadVR tested Windows 11 24H2 with a Reverb G2 and found the above notice. Microsoft confirmed to UploadVR that this is an intentional removal when it originally announced the move back in December.

In August 3.49% of SteamVR users were using a Windows MR headset, roughly 80,000 people. If they install Windows 11 24H2, their VR headset will effectively become a paperweight.

Steam said: “Existing Windows Mixed Reality devices will continue to work with Steam through November 2026, if users remain on their current released version of Windows 11 (version 23H2) and do not upgrade to this year’s annual feature update for Windows 11 (version 24H2).”

The death of Windows MR headsets comes on the same week Microsoft revealed that HoloLens 2 production has ended, and that software support for the AR headset will end after 2027.

«

I think that’s what’s known as a signal. HoloLens going is significant: it seems that Microsoft has decided that VR, at least its form, isn’t the thing.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Some university students in America now quail at reading a long book in a week. Or any books. CC-licensed photo by vickysandoval22 on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.

A selection of 10 links for you. Novel. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Google paid $2.7bn to bring back an AI genius who quit in frustration • WSJ

Miles Kruppa and Lauren Thomas:

»

At a time when tech companies are paying eye-popping sums to hire the best minds in artificial intelligence, Google’s deal to rehire Noam Shazeer has left others in the dust. 

A co-author of a seminal research paper that kicked off the AI boom, Shazeer quit Google in 2021 to start his own company after the search giant refused to release a chatbot he developed. When that startup, Character.AI, began to flounder, his old employer swooped in.

Google wrote Character a check for around $2.7bn, according to people with knowledge of the deal. The official reason for the payment was to license Character’s technology. But the deal included another component: Shazeer agreed to work for Google again.

Within Google, Shazeer’s return is widely viewed as the primary reason the company agreed to pay the multibillion-dollar licensing fee.

The arrangement has thrust him into the middle of a debate in Silicon Valley about whether tech giants are overspending in the race to develop cutting-edge AI, which some believe will define the future of computing. 

“Noam is clearly a great person in that space,” said Christopher Manning, director of the Stanford Artificial Intelligence Laboratory. “Is he 20 times as good as other people?” 

«

The fact that Google has this amount of money to spend on a single individual (come on, they don’t need the technology) is just mindboggling. Heading off a rival? Is that the true profit that he will add to the company? It’s hard to think anyone since a few of the original hires at Google has managed to be that valuable.
unique link to this extract

Signal’s Meredith Whittaker: ‘I see AI as born out of surveillance’ • Financial Times

Madhumita Murgia:

»

Until 2017, Whittaker had thought she could successfully mobilise change from inside the machine, building up ethical AI research and development programmes at Google in collaboration with academics at universities and companies such as Microsoft. But in the autumn of that year, a colleague contacted her about a project they were working on. They had learnt it was part of a Department of Defense pilot contract, codenamed Project Maven, that used AI to analyse video imagery and eventually improve drone strikes. “I was basically just a . . . dissent court jester,” she says, still visibly disappointed.

She drafted an open letter to Google’s chief executive, Sundar Pichai, that received more than 3,000 employee signatures, urging the company to pull out of the contract. “We believe that Google should not be in the business of war,” the letter said.

“The Maven letter was sort of like, I can’t make my name as an ethical actor redounding to Google’s benefit,” she says. “You’re talking about Google becoming a military contractor. It’s still shocking, although it’s become normalised for us, but this is a centralised surveillance company with more kompromat than anyone could ever dream of, and now they’re partnering with the world’s most lethal military, as they call themselves.

“Yeah, that was the end of my rope.”

Whittaker went on to help organise employee protests and walkouts, in which more than 20,000 Google workers participated, to protest against the company’s handling of other ethical matters such as sexual harassment allegations against high-profile executives. At the time, Google’s management opted not to renew the Pentagon contract once it expired. But Whittaker left Google in 2019, after the company presented her with a set of options that she says gave her no choice but to quit. “It was like, you can go be an administrator, doing spreadsheets and budgets for the open source office [and] stop all the shit I had been building forever.”

«

Fascinating interview – there’s much more (the link should pass the paywall) and it’s all engrossing.
unique link to this extract

CNN puts a paywall on its website as TV revenues decline • SF Gate

Stephen Battaglio:

»

CNN has long had one of the most visited news websites in the world. Starting Tuesday, users are going to have to pay for it.

The Warner Bros. Discovery-owned news operation is putting a paywall on CNN.com, requiring U.S. users to pay $3.99 for access or a discounted rate of $29.99 a year. The subscription will allow unlimited usage of the site, which is visited by 150 million people globally each month.

Users will be asked to subscribe after accessing a number of free stories, according to an internal memo from Alex MacCallum, executive vice president of digital products and services for CNN.

CNN’s reason for the move is rooted in the problems that plague all of traditional television. Consumers are spending more time with online video and canceling their traditional pay-TV subscriptions. Revenues from cable and satellite subscribers are declining as cord-cutting continues at a steady pace each year. The trend, along with a decline in ratings, has put pressure on CNN’s profit margins.

Whether consumers will pay for a product they have used for free over the years remains to be seen. Mark Thompson, who took over as CNN’s chairman last year, turned the New York Times into a successful digital subscription site during his tenure at that company.

MacCallum’s memo said subscribers “will receive benefits like exclusive election features, original documentaries, a curated daily selection of our most distinctive journalism, and fewer digital ads.” CNN is currently developing video content with some of its talent designed to be behind the paywall on the site, according to people familiar with the plans.

«

That’s not a lot of money to ask. And yet it’s going to mean a lot of people not going on the site: any amount of friction will do that. Passwords, usernames, different devices, it’s going to be the usual mess.
unique link to this extract

The elite college students who can’t read books • The Atlantic

Rose Horowitch:

»

Nicholas dames has taught Literature Humanities, Columbia University’s required great-books course, since 1998. He loves the job, but it has changed. Over the past decade, students have become overwhelmed by the reading. College kids have never read everything they’re assigned, of course, but this feels different. Dames’s students now seem bewildered by the thought of finishing multiple books a semester. His colleagues have noticed the same problem. Many students no longer arrive at college—even at highly selective, elite colleges—prepared to read books.

This development puzzled Dames until one day during the fall 2022 semester, when a first-year student came to his office hours to share how challenging she had found the early assignments. Lit Hum often requires students to read a book, sometimes a very long and dense one, in just a week or two. But the student told Dames that, at her public high school, she had never been required to read an entire book. She had been assigned excerpts, poetry, and news articles, but not a single book cover to cover.

“My jaw dropped,” Dames told me. The anecdote helped explain the change he was seeing in his students: It’s not that they don’t want to do the reading. It’s that they don’t know how. Middle and high schools have stopped asking them to.

In 1979, Martha Maxwell, an influential literacy scholar, wrote, “Every generation, at some point, discovers that students cannot read as well as they would like or as well as professors expect.” Dames, who studies the history of the novel, acknowledged the longevity of the complaint. “Part of me is always tempted to be very skeptical about the idea that this is something new,” he said.

And yet. “I think there is a phenomenon that we’re noticing that I’m also hesitant to ignore.” Twenty years ago, Dames’s classes had no problem engaging in sophisticated discussions of Pride and Prejudice one week and Crime and Punishment the next. Now his students tell him up front that the reading load feels impossible.

«

Pride & Prejudice is 108,500 words (or so). Crime & Punishment is 107,500 words (or so). Think of all the TikToks you could watch in the time it takes you to read them!
unique link to this extract

How Germany outfitted half a million balconies with solar panels • Grist

Akielly Hu:

»

Matthias Weyland loves having people ask about his balcony. A pair of solar panels hang from the railing, casting a sheen of dark blue against the red brick of his apartment building. They’re connected to a microinverter plugged into a wall outlet and feed electricity directly into his home. On a sunny day, he’ll produce enough power to supply up to half of his family’s daily needs.

Weyland is one of hundreds of thousands of people across Germany who have embraced balkonkraftwerk, or balcony solar. Unlike rooftop photovoltaics, the technology doesn’t require users to own their home, and anyone capable of plugging in an appliance can set it up. Most people buy the simple hardware online or at the supermarket for about $550 (500 euros.)

The ease of installation and a potent mix of government policies to encourage adoption has made the wee arrays hugely popular. More than 550,000 of them dot cities and towns nationwide, half of which were installed in 2023. During the first half of this year, Germany added 200 megawatts of balcony solar. Regulations limit each system to just 800 watts, enough to power a small fridge or charge a laptop, but the cumulative effect is nudging the country toward its clean energy goals while giving apartment dwellers, who make up more than half of the population, an easy way to save money and address the climate crisis.

“I love the feeling of charging the bike when the sun is shining, or having the washing machine run when the sun is shining, and to know that it comes directly from the sun,” Weyland said. “It’s a small step you can take as a tenant” and an act of “self-efficacy, to not just sit and wait until the climate crisis gets worse.”

«

Guess they could also not shut down their nuclear power stations and stop using coal? Just a thought.
unique link to this extract

Fraud. So much fraud • Science

Derek Lowe:

»

Charles Piller and the team here at Science dropped a big story on Thursday morning, and if you haven’t read it yet, you should. It’s about Eliezer Masliah, who since 2016 has been the head of the Division of Neuroscience in the National Institute on Aging (NIA), and whose scientific publication record over at least the past 25 years shows multiple, widespread, blatant instances of fraud. There it is in about as few words as possible.

As is so often the case, image manipulation is at the heart of the scandal. Readers here will be all too familiar with the techniques of cutting and pasting Western blots in order to make them tell the story the authors want told, and of re-using images and parts of images over and over even when they’re supposed to be produced from different experiments at different times. That’s what we’re seeing here, and a 300-page dossier has been assembled with examples of it.

Splicing, cloning, overlaying, copy-and-pasting, duplication of the same image with different captions about different research in different journals: a great deal of effort seems to have gone into carefully doctoring, cleaning, beautifying, and spicing up these papers digitally. After looking over examples, I find the evidence convincing and impossible to explain (at least in my mind) as anything other than sustained, deliberate acts of deception lasting for decades.

Hundreds of them. Again and again. The dossier references 132 papers with apparent problems. Unfortunately, these include many highly cited papers on mechanisms of synaptic damage (Masliah specialized in Alzheimer’s and Parkinson’s mechanisms, particularly around the alpha-synuclein protein).

As the article details, this all has some direct drug discovery implications, particularly for an antibody called prasinezumab which targets alpha-synuclein. All four of the fundamental papers about prasinezumab (as cited on the web site of its developer, Prothena) are full of manipulated images, unfortunately.

«

As the blogpost points out, this is disastrous for Alzheimer’s research. (Via Benedict Evans.)
unique link to this extract

AI crawlers are hammering sites and nearly taking them offline • Fast Company

Chris Stokel-Walker:

»

A number of websites have begun to take action to fend off crawlers, seeking to avoid the negative impact of being bombarded with requests. An increasing number of websites are putting restrictions on AI crawlers, according to a recent analysis by the Data Provenance Initiative (DPI), a group of AI researchers. In the DPI’s analysis, around one in four tokens from the most critical web domains called upon by crawlers have put up restrictions. And social media is buzzing with complaints about the increasing instances of web crawlers pushing up traffic on websites.

Edd Coates is one of those who has raised concerns online. He runs Game UI Database, a database of details taken from games designed to be used as a reference tool. The website was relaunched in early August, gaining large volumes of visitors keen to check it out. But then a few weeks later, the website’s performance declined dramatically, slowing to a crawl. “I thought that was weird, because we had about a quarter of the people visiting the website that did at the relaunch,” says Coates. “And it’s somehow running slower.”

Coates and his web developer checked the website’s server logs, which turned up the cause of the problem: a crawler by OpenAI was pouncing on the website. “They were hitting the site so hard,” he says. “It was, like, 200 times a second.” OpenAI doesn’t dispute its GPTBot crawler visited Game UI Database, but does dispute the scale of how frequently their crawler was hitting the website, showing evidence that suggested the number of queries per second was only around three.

An OpenAI spokesperson told Fast Company: “We enable publishers to use industry-standard tools to express preferences about access to their websites. By using robots.txt publishers can set time delays and reduce load on their systems, choose to allow access to only certain pages or directories, or opt out entirely. We stopped accessing this website as soon as they updated their robots.txt directions for our bot, as our systems recognized and respected this.”

«

All these years it’s existed and robots.txt still doesn’t have a “don’t hit this site more than X times per second/minute/hour” setting.
unique link to this extract

Study finds zero cases of cervical cancer among women vaccinated for HPV before age 14 • STAT

Annalisa Merelli:

»

A historic new study out of Scotland shows the real-world impact of vaccines against the human papillomavirus: no cases of cervical cancer were detected in women born between 1988-1996 who were fully vaccinated against HPV between the ages of 12 and 13.

Many previous studies have shown that HPV vaccines are extremely effective in preventing cervical cancer. But the study, published in the Journal of the National Cancer Institute, is the first to monitor a national cohort of women over such a long time period and find no occurrence of cervical cancer.

“The study is super exciting. It shows that the vaccine is extremely effective,” said Kathleen Schmeler, a professor of gynecologic oncology at the University of Texas MD Anderson Cancer Center, who was not involved in the research. “It’s obviously early. We’re just starting to see the first data of the impact of the vaccine because it takes so long from the time of the vaccine to the effects.”

The results underscore the importance of working to increase uptake of the HPV vaccine in the US, said Schmeler. Scotland, for example, introduced routine immunization in schools in 2008, and close to 90% of students in their fourth year of secondary school (equivalent to 10th grade in the US.) in the 2022-2023 school year had received at least one dose of the vaccine. In the US, where HPV vaccines are not administered in school, uptake among adolescents ages 13 to 17 is a little over 60%.

«

In case you didn’t know, HPV is identified as a key cause of cervical cancer. Even one dose given before girls become sexually active seems to be effective. In the age cohort, the expectation was for 15 to 17 cases. There could be other HPV strains, and this isn’t the end of the story. But it’s a big bookmark in the story.
unique link to this extract

Stem cells reverse woman’s diabetes — a world first • Nature

Smriti Mallapaty:

»

A 25-year-old woman with type 1 diabetes started producing her own insulin less than three months after receiving a transplant of reprogrammed stem cells. She is the first person with the disease to be treated using cells that were extracted from her own body.

“I can eat sugar now,” said the woman, who lives in Tianjin, China, on a call with Nature. It has been more than a year since the transplant, and, she says, “I enjoy eating everything — especially hotpot.” The woman asked to remain anonymous to protect her privacy.

James Shapiro, a transplant surgeon and researcher at the University of Alberta in Edmonton, Canada, says the results of the surgery are stunning. “They’ve completely reversed diabetes in the patient, who was requiring substantial amounts of insulin beforehand.”

The study, published in Cell, follows results from a separate group in Shanghai, China, who reported in April that they had successfully transplanted insulin-producing islets into the liver of a 59-year-old man with type 2 diabetes. The islets were also derived from reprogrammed stem cells taken from the man’s own body, and he has since stopped taking insulin.

The studies are among a handful of pioneering trials using stem cells to treat diabetes, which affects close to half a billion people worldwide. Most of them have type 2 diabetes, in which the body doesn’t produce enough insulin or its ability to use the hormone diminishes. In type 1 diabetes, the immune system attacks islet cells in the pancreas.

«

So far the woman’s transplanted cells have generated insulin for a year; other researchers want to see them work for five years before they consider her “cured”. But it’s a big breakthrough.
unique link to this extract

Did you solve it? The box problem that baffled the boffins • The Guardian

Alex Bellos:

»

The 15 boxes problem.

Andrew and Barbara are playing a game, in which fifteen boxes are arranged in a 3-row 5-column grid as shown below:

Prizes are put in two randomly-chosen boxes. Andrew will search the boxes row by row, so his search order is ABCDEFGHIJKLMNO. Barbara will search column by column, so her order is AFKBGLCHMDINEJO.

If Andrew and Barbara open their boxes together each turn, that is, on the first turn, they both open A, on the second, Andrew opens B and Barbara opens F, on the third Andrew opens C, and Barbara opens K, and so on, who is more likely to find a prize first?

a) Andrew.
b) Barbara.
c) Both equally likely.

«

Think a bit about this one. I got the correct answer despite misreading the question (it matters that there are *two* prizes hidden.) Bellos says it intrigues mathematicians because they struggle to find an intuitive explanation for why the correct answer is correct. The obvious next step to find that intuitive explanation, I think, is to consider whether the result would be different if the grid were square, and if it were taller than it is wide but both players used the same strategy. (And would adding more prizes change anything?) Bellos has a new book out – Think Twice – which is full of intriguing puzzles like this. For those that like that sort of thing, that is the thing they like.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified