People are getting injured or even killed taking selfies, a new study shows – and it’s time to warn them properly. CC-licensed photo by Mike Goad 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. Hello, duckface. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Google’s 36% search revenue share with Apple is 3x what Android OEMs get • Ars Technica

Ron Amadeo:

»

How much more does Google pay for an Apple user than an Android one? A lot. It was recently revealed in the Epic v. Google trial (Google has a few monopoly lawsuits going on) that the highest tier of search revenue share for cooperative Android OEMs is only 12%, a third of what Google pays Apple. In terms of total cash amount, it’s reasonable to assume Apple gets more total money than many smaller companies but to see the direct breakdown that each Apple user is worth three times more than an Android user is a new insight.

A big part of the differing payment rates probably has to do with how threatened Google feels by each company. Apple has already proven that it has the power to dump an established Google service and go off on its own. A prime example is Apple Maps, which replaced Google Maps as a default iOS app and, according to testimony from Google VP of Finance, Michael Roszak, tanked Google Maps mobile traffic by 60% when it launched. Roszak said that Google uses the Apple Maps launch as “a datapoint” when estimating how an Apple search switch would go. No one on the Android side has this kind of power. There’s also the consideration that Apple users are generally more affluent than Android users, making them more desirable ad clickers.

On Android, Google has differing tiers of payments depending on how Google-y your phone is. As revealed in documents from Epic v. Google, Android’s “Premier Device Program” offers 12% search revenue to devices with “Google exclusivity and defaults for all key functions” and no rival app stores.  The big participants in this program are/were Motorola, LG, and HMD, which had at least 98% of their devices qualify. Other brands like Xiaomi, Sony, Sharp, and BBK (that’s OnePlus, Oppo, and Vivo) were at 70%.

Android partners don’t just get search revenue; they also get a cut of Google Play app sales and ads run on their devices. In the case of Motorola and LG, they were getting another 3–6% of Play Store spending.

«

It is fascinating seeing Google’s business model, particularly around TAC (traffic acquisition costs), being picked apart in public like this.
unique link to this extract

Meta bars political advertisers from using generative AI ads tools • Reuters

Katie Paul:

»

Facebook owner Meta is barring political campaigns and advertisers in other regulated industries from using its new generative AI advertising products, a company spokesperson said on Monday, denying access to tools that lawmakers have warned could turbo-charge the spread of election misinformation.

Meta publicly disclosed the decision in updates posted to its help center on Monday night, following publication of this story. Its advertising standards prohibit ads with content that have been debunked by the company’s fact-checking partners but do not have any rules specifically on AI.

“As we continue to test new Generative AI ads creation tools in Ads Manager, advertisers running campaigns that qualify as ads for Housing, Employment or Credit or Social Issues, Elections, or Politics, or related to Health, Pharmaceuticals or Financial Services aren’t currently permitted to use these Generative AI features,” the company said in a note appended to several pages explaining how the tools work.

«

So political campaigns won’t be able to use Meta’s generative AI tools to make their ads, but they could certainly do so outside Meta and then just upload them. Hard to see what difference this makes to anything, apart from giving Meta something to say it “doesn’t allow” in US congressional hearings.
unique link to this extract

Why I just resigned from my job in generative AI • Music Business Worldwide

Ed Newton-Rex worked at Stability AI on its Stable Audio generative AI music-making platform; he’s a published classical composer in his own right:

»

I’ve resigned from my role leading the Audio team at Stability AI, because I don’t agree with the company’s opinion that training generative AI models on copyrighted works is ‘fair use’.

…I wasn’t able to change the prevailing opinion on fair use at the company.

This was made clear when the US Copyright Office recently invited public comments on generative AI and copyright, and Stability was one of many AI companies to respond. Stability’s 23-page submission included this on its opening page: “We believe that Al development is an acceptable, transformative, and socially-beneficial use of existing content that is protected by fair use”.

For those unfamiliar with ‘fair use’, this claims that training an AI model on copyrighted works doesn’t infringe the copyright in those works, so it can be done without permission, and without payment. This is a position that is fairly standard across many of the large generative AI companies, and other big tech companies building these models — it’s far from a view that is unique to Stability. But it’s a position I disagree with.

I disagree because one of the factors affecting whether the act of copying is fair use, according to Congress, is “the effect of the use upon the potential market for or value of the copyrighted work”. Today’s generative AI models can clearly be used to create works that compete with the copyrighted works they are trained on. So I don’t see how using copyrighted works to train generative AI models of this nature can be considered fair use.

«

Well, I guess we’ll see how the courts view this. But it seems to me that this is a “transformational” use and will be allowed. Will Newton-Rex ask for his job back if the courts rule that way?
unique link to this extract

The incredible shrinking heat pump • The Verge

Justine Calma:

»

Just 10% of households worldwide have heat pumps today. Those are typically bigger, more complex, and expensive systems that need to be professionally installed. For those reasons, they’re usually out of reach for renters. New York City Housing Authority (NYCHA) actually did a test run with one of those existing options, called a split system unit, which involved mounting equipment on the roof and on the wall in a tenant’s home. It ended up being too unwieldy, and the project stopped there. 

Unfortunately, when it comes to new, more efficient appliances and clean energy technologies, it’s typically more affluent households that can afford to bring these new things into their homes first. The benefits don’t usually trickle down to lower-income households until later, if at all. 

New York is attempting to flip that scenario now by purchasing new window heat pumps for public housing residents. “The beauty of this project is that some of the lowest-income residents in the city are experiencing the newest technology for the first time so they’re leading in this area, which is really nice and something that we’re very proud of,” says Justin Driscoll, president and CEO of New York Power Authority, the public power organization that procures electricity for NYCHA.

The big motivation to switch to heat pumps now, though, is a deadline. Back in 2019, New York state passed a law to reduce the greenhouse gas emissions causing climate change by 85% by 2050.

…Heat and hot water in buildings create about 40% of New York City’s planet-heating greenhouse gas emissions. So, in 2021, NYCHA and its partners announced a $263m investment in electric heat pumps.

«

They have huge air conditioning units festooned over houses, so why not heat pumps?
unique link to this extract

Here’s how bad climate change will get in the US—and why there’s still hope • WIRED

Matt Simon:

»

The [US Fifth National Climate] assessment notes the already staggering cost of climate change in the US, beyond wildfires. In the 1980s, on average, the US experienced one billion-dollar disaster every four months. That’s now one every three weeks. Between 2018 and 2022, the country suffered 89 billion-dollar events. Extreme weather now costs the country nearly $150bn annually. But, the report emphasizes, that’s a conservative estimate, because it doesn’t consider the costs of the aftermath, like loss of life, health care for survivors, or the damage done to ecosystems.

“I think this report really highlights how the changes we’re experiencing now are unprecedented in our nation’s history,” says Kristina Dahl, a technical contributor to the assessment and principal climate scientist for the Climate and Energy program at the Union of Concerned Scientists. “The US has warmed more quickly than the planet as a whole. So the US is really feeling this.”

The assessment also points out that in the next three decades, scientists expect sea levels along the contiguous US to rise nearly a foot. By 2050, coastal flooding will happen five to 10 times more often than today, and by the end of the century, millions of seaside residents could be displaced. But we’re dealing with a lot of uncertainty. Sea level rise could accelerate if the ice sheets atop Greenland and Antarctica start declining faster. Just last week, a study found that northern Greenland’s ice is in much worse shape than previously understood. “Uncertainty in the stability of ice sheets at high warming levels means that increases in sea level along the continental US of 3-7 feet by 2100 and 5-12 feet by 2150 are distinct possibilities that cannot be ruled out,” the assessment warns.

And keep in mind that sea level rise will not unfold uniformly across US coastlines, due to quirks in the physics involved. Some places, like the Gulf Coast, are also rapidly sinking, a phenomenon known as subsidence, which exacerbates the problem.

«

2050 actually doesn’t feel that far away.
unique link to this extract

YouTubers asked to disclose AI-generated content – or else • The Register

Katyanna Quach:

»

YouTube is slapping a bunch of rules on AI-generated videos in the hope of curbing the spread of faked footage masqueraded as legit; deepfakes that make people appear to say or do things they never did; and tracks that rip off artists’ copyrighted work.

This red tape will be rolled out over the coming months and apply to material uploaded by users, we’re told

Specifically, the Google-owned vid-sharing giant will require content creators to disclose if their videos contain believable synthetic footage of made-up events, including AI-made depictions, or deepfakes that put words in people’s mouths. In those cases, a label will be added to a video’s description declaring the content was altered or digitally generated, and a more prominent note will be added to the video player itself if the content is particularly sensitive. Breaking the rules will lead to content being torn down and accounts punished.

…Faked footage that could mislead viewers about important topics such as elections, conflicts and violence, public health issues, or popular figures must also be flagged in particular. “Creators who consistently choose not to disclose this information may be subject to content removal, suspension from the YouTube Partner Program, or other penalties,” YouTube product veeps Jennifer Flannery O’Connor and Emily Moxley warned today.

«

Well, let’s see how that works out for them. How will they know, apart from anything?
unique link to this extract

Everyone stop being ridiculous for like five minutes • Garbage Day

Ryan Broderick:

»

There’s an old journalism joke that reporters cover every new election according to the rules of the previous one. But I think the tech press does the same thing. Which explains why most of the stories you read about AI right now use the same whack-a-mole content cop strategy most news outlets and research groups spent the 2010s using to cover platforms like Facebook or Twitter. Now they’re breathlessly writing up every instance of an AI producing A Forbidden Image. And what’s worse is this attitude helps tech companies continue to undermine labor and consolidate lobbying power, allows politicians to keep dragging their feet on writing real legislation for the internet, and provides fantastic cover for online platforms that still don’t know how to moderate themselves. I have yet to see anything produced by generative AI you couldn’t do with Photoshop or After Effects or, like, Wikipedia. And if everyone stopped being ridiculous for five minutes, we’d all realize that this tech hasn’t introduced a single new problem. We still just have same old ones we refuse to deal with!

And so, my big hot AI take here is that there’s actually nothing new to moderate. I mean, my god, OpenAI is literally using the same Africa-based third-party moderation contractors that Meta and Google use. It’s all just the same stuff with a new Sci-Fi coat of paint.

«

Broderick essentially arguing the opposite position from the Tech Against Terrorism research linked here yesterday from Wired.
unique link to this extract

Sam Bankman-Fried exposed the fragility of crypto • The New York Times

Molly White:

»

Although some might try to dismiss the FTX collapse as a unique case, it is far from it: the funding model [of issuing a “coin” in an “initial coin offering”, or ICO, and putting a ludicrous value on it, and then using that as collateral for loans of real money] has become all too normal in the cryptocurrency world. Entrepreneurs thought they had found a free money machine in 2017 as initial coin offerings became popular, enabling cryptocurrency companies to bootstrap without needing to find venture investors — investors who might insist on a seat on the board or a view into the company’s operations. A crackdown on I.C.O.s in the United States shortly after failed to stop the practice, with companies either presenting the offerings in disguises designed to stymie if not entirely evade the Securities and Exchange Commission, or moving offshore in hopes of being beyond the reach of the long arms of the Securities and Exchange Commission (S.E.C.)’s enforcers.

FTX is perhaps the best known catastrophe, but the same pattern has played out for customers of the Celsius cryptocurrency lender’s CEL token, the Voyager Digital broker’s VGX and the Terra/Luna ecosystem’s LUNA. Civil and criminal cases have revealed internal conversations among Celsius executives desperately trying to support the CEL token price to keep the floundering company afloat, to no avail, knowing what the token’s collapse would mean for the company. Binance, a still-operational exchange whose balance sheets are as opaque as those of the Bankman-Fried companies before their collapse, heavily promotes its BNB token. The extent to which the company relies on BNB to finance its operations is unclear, but history provides ominous warnings.

The collapse of the FTX exchange revealed the massive duplicity underlying many crypto exchanges, but its implosion should not be attributed to that alone. It, like so many companies in the cryptocurrency industry, had propped itself up on an imaginary foundation of tokens it had invented, and that foundation was bound to fail eventually. When the next company in its position falls, the only surprise should be that people expected any other outcome.

«

I’ve long since ceased to be surprised by the things that people will believe will happen differently this time.
unique link to this extract

Selfie-related deaths at tourist sites are ‘public health problem’: researchers • NY Post

Angie Raphael:

»

Selfie-related injuries and deaths at tourist hotspots have become such a massive risk that they should be viewed as a “public health problem”, researchers suggest.

Of particular concern are selfie-related deaths at picturesque aquatic locations, such as waterfalls, according to the University of New South Wales, Australia study.

Part of the study examined how selfie-related injuries and deaths were reported in the media.

Four peer-reviewed studies identified falls from a height, such as a cliff or waterfall, as the most common incident. Drowning was the second most common cause of death. People often climbed over barriers and fenced-off areas to get to the perfect selfie spot, the report noted.

The mean age of victims was about 22, most of whom were female tourists. “The selfie-related incident phenomenon should be viewed as a public health problem that requires a public health risk communication response,” the report concluded. “To date, little attention has been paid to averting selfie-related incidents through behaviour change methodologies or direct messaging to users, including through social media apps.”

Previous research recommended “no selfie zones”, barriers and signage as ways to prevent selfie-related injuries and deaths.

«

Perhaps something like “5 people have died here so far taking selfies”, but write the number with 1, 2, 3, and 4 crossed out in front of it. The paper is in the “Journal of Medical Internet Research”, which is an intriguing title in its own right.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Weather forecasting can be done more accurately and more quickly with using machine learning systems, Google DeepMind has shown. CC-licensed photo by Chic Bee 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. Bright prospects. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Here’s how violent extremists are exploiting generative AI tools • WIRED

David Gilbert:

»

For years, Big Tech platforms have worked hard to create databases of known violent extremist content, known as hashing databases, which are shared across platforms to quickly and automatically remove such content from the internet. But according to Hadley, his colleagues are now picking up around 5,000 examples of AI-generated content each week. This includes images shared in recent weeks by groups linked to Hezbollah and Hamas that appear designed to influence the narrative around the Israel-Hamas war.

“Give it six months or so, the possibility that [they] are manipulating imagery to break hashing is really concerning,” Hadley says. “The tech sector has done so well to build automated technology, terrorists could well start using gen AI to evade what’s already been done.”

Other examples that researchers at Tech Against Terrorism have uncovered in recent months have included a neo-Nazi messaging channel sharing AI-generated imagery created using racist and antisemitic prompts pasted into an app available on the Google Play store; far-right figures producing a “guide to memetic warfare” advising others on how to use AI-generated image tools to create extremist memes; the Islamic State publishing a tech support guide on how to securely use generative AI tools; a pro-IS user of an archiving service claiming to have used an AI-based automatic speech recognition (ASR) system to transcribe Arabic language IS propaganda; and a pro-al-Qaeda outlet publishing several posters with images highly likely to have been created using a generative AI platform.

«

unique link to this extract

Nothing is bringing iMessage to its Android phone • The Verge

Wes Davis:

»

Nothing Phone 2 owners get blue bubbles now. The company shared it has added iMessage to its newest phone through a new “Nothing Chats” app powered by the messaging platform Sunbird. The feature will be available to users in North America, the EU, and other European countries starting this Friday, November 17th.

Nothing writes on its page that it’s doing this because “messaging services are dividing phone users,” and it wants “to break those barriers down.” But doing so here requires you to trust Sunbird. Nothing’s FAQ says Sunbird’s “architecture provides a system to deliver a message from one user to another without ever storing it at any point in its journey,” and that messages aren’t stored on its servers.

Marques Brownlee has also had a preview of Nothing Chats. He confirmed with Nothing that, similar to how other iMessage-to-Android bridge services have worked before, “…it’s literally signing in on some Mac Mini in a server farm somewhere, and that Mac Mini will then do all of the routing for you to make this happen.”

Nothing’s US head of PR, Jane Nho, told The Verge in an email that Sunbird stores user iCloud credentials as a token “in an encrypted database” and associated with one of its Mac Minis in the US or Europe, depending on the user’s location, that then act as a relay for iMessages sent via the app. She added that, after two weeks of inactivity, Sunbird deletes the account information.

But you’re still giving them access to your iCloud account to make this work, and as we’ve all learned over the years, companies don’t always do what they say they will. It’s worth reviewing Sunbird’s privacy policy and keeping a very skeptical mind about it.

«

Sunbird doesn’t explain how it does this. My understanding is that iMessages require key exchange, and that the private/public keypair is generated by the device itself. How does Sunbird generate an appropriate hardware key for the Android device? At Pocket Lint, Jason Cipriani wags a big finger and says no, don’t do this: it requires Sunbird signing into your iCloud account on a Mac it controls:

»

“You’re more or less giving Sunbird access to your entire Apple ID/iCloud account, and if you’re someone who uses Apple’s services, that’s a scary thought.”

«

Also, jeepers, people: just use WhatsApp, or Signal. Platform-specific messaging apps are so 2010s.
unique link to this extract

Robotic putting greens, mixed reality, loud spectators: this is golf?! • WIRED

Steven Levy:

»

Cameron Young slides a driver from his bag. He stares at a hole referred to as Texas Hill Country. It’s new to him—a par 4 with sand hazards and rough to avoid. The 26-year-old is in the top 20 in the Official World Golf Ranking, but he’s not sure how to proceed. He turns to his companion, former pro Roberto Castro. “What’s going on here?” Young asks.

Castro consults with their caddie and reports, “It’s 312 to that bunker there.”

Young makes clean contact. The ball lofts skyward.

But there’s no sky above him. On this steamy day in late October, Young is in an air-conditioned soundstage on the back lot of Universal Studios in Orlando, Florida. The building once hosted Nickelodeon TV shows. The “caddie” Castro consulted is virtual—it lives on a 15-inch tablet. The tee is on a patch of natural grass the width of a large mattress. It sits atop wooden pallets on a concrete floor.

Young’s golf ball hits a billboard-sized screen 35 yards away. The dimpled sphere falls meekly to the ground, while up on the giant display its virtual successor continues its flight. A phalanx of supersensitive radar trackers and hi-res cameras sends data to a bank of computer servers that calculate velocity and spin to show how the ball will bounce and where it will ultimately settle on the vista of the screen.

Young’s ball lands in the digital rough. He walks over to a tray of two-inch-high Bermuda grass mixed with rye. The screen now shows him closer to his goal, an 8-iron away. He swings, the ball thuds against the display again, and seconds later his virtual ball lands just outside the green.

…Many pro golfers practice using room-sized simulators in their personal gym, and weekend warriors commonly visit golf centers with plenty of tech. That’s not what Young is up to. He’s testing a system for real competition that will be aired on prime time, with $20m of prize money at stake. He’s one of 24 pros, including golf legends Tiger Woods and Rory McIlroy, who are involved in the most ambitious effort yet to merge e-gaming and actual pro sports. It’s called TGL, allegedly not an acronym for The Golf League, but three TV-friendly letters that don’t mean anything.

TGL’s first event will take place on January 9 inside a $50m–plus, custom-built arena with an inflatable dome in Palm Beach Gardens, Florida. A 200,000-pound [91 tonne] turntable will support an 800,000-pound [363 tonne] green that will shape-shift to give each hole its character. A 4K screen will rival the goliath displays of Taylor Swift concerts. The stands will accommodate around 1,600 live spectators, who are encouraged to boisterously violate golf’s finicky silence rule. Players themselves will be mic’d up, in hopes that their trash talk might go viral online.

«

Strange things that the combination of money, TV and empty airtime will make people try.
unique link to this extract

Andreessen Horowitz invests in Civitai, which profits from nonconsensual AI porn • 404 Media

Emanuel Maiberg:

»

Andreessen Horowitz, also known as a16z, the influential Silicon Valley venture capital firm that was an early investor in Facebook, Lyft, and other tech giants, has invested in Civitai, a giant platform for sharing AI models that enables and profits from the creation of AI generated nonconsensual sexual images of real people. That includes launching a feature where people can list “bounties” for others to create AI models of specific targets.

Civitai said that it raised $5.1m in a seed funding round led by a16z.

A16z’s official website, which includes a jobs board with open positions at companies in its portfolio, currently lists five jobs at Civitai. According to a16z’s site, these jobs were posted more than 30 days ago.

A16z regularly announces investments the company is making on its site, but has not publicly announced its investment in Civitai yet.

A16z did not respond to a request for comment. When asked about a16z’s investment in Civitai over Discord, a community engagement manager at Civitai told 404 Media that “There will be a press release/announcement shortly.” Civitai then published a press release confirming the investment minutes after 404 Media reached out for comment.

In August, 404 Media published an investigation into Civitai, which explained how the platform works, and enables the creation of AI-generated nonconsensual sexual images, and profits from it. Civitai allows users to share modified models of the open source text-to-image AI tool Stable Diffusion. These modified models are often trained on images of celebrities, influencers, YouTubers, and athletes, almost exclusively women, to recreate their likeness. Those models can then be combined with AI models that are trained on porn in order to instantly generate nonconsensual sexual images.

«

You know, I’m beginning to think that a16z is a bit skeevy.
unique link to this extract

Google DeepMind’s weather AI can forecast extreme weather faster and more accurately • MIT Technology Review

Melissa Heikkilä:

»

In research published in Science on Tuesday, Google DeepMind’s model, GraphCast, was able to predict weather conditions up to 10 days in advance, more accurately and much faster than the current gold standard. GraphCast outperformed the model from the European Centre for Medium-Range Weather Forecasts (ECMWF) in more than 90% of over 1,300 test areas. And on predictions for Earth’s troposphere—the lowest part of the atmosphere, where most weather happens—GraphCast outperformed the ECMWF’s model on more than 99% of weather variables, such as rain and air temperature

Crucially, GraphCast can also offer meteorologists accurate warnings, much earlier than standard models, of conditions such as extreme temperatures and the paths of cyclones. In September, GraphCast accurately predicted that Hurricane Lee would make landfall in Nova Scotia nine days in advance, says Rémi Lam, a staff research scientist at Google DeepMind. Traditional weather forecasting models pinpointed the hurricane to Nova Scotia only six days in advance.

“Weather prediction is one of the most challenging problems that humanity has been working on for a long, long time. And if you look at what has happened in the last few years with climate change, this is an incredibly important problem,” says Pushmeet Kohli, the vice president of research at Google DeepMind.

Traditionally, meteorologists use massive computer simulations to make weather predictions. They are very energy intensive and time consuming to run, because the simulations take into account many physics-based equations and different weather variables such as temperature, precipitation, pressure, wind, humidity, and cloudiness, one by one.

GraphCast uses machine learning to do these calculations in under a minute. Instead of using the physics-based equations, it bases its predictions on four decades of historical weather data. GraphCast uses graph neural networks, which map Earth’s surface into more than a million grid points. At each grid point, the model predicts the temperature, wind speed and direction, and mean sea-level pressure, as well as other conditions like humidity. The neural network is then able to find patterns and draw conclusions about what will happen next for each of these data points.

«

Surprise! Historical data is a good predictor of future weather patterns. Unfortunately “It still lags behind conventional weather forecasting models in some areas, such as precipitation, Dueben says”. In other words, it’s not going to take over rain prediction – the thing we really want – just yet.
unique link to this extract

Chatbots may ‘hallucinate’ more often than many realise • The New York Times

Cade Metz:

»

a new start-up called Vectara, founded by former Google employees, is trying to figure out how often chatbots veer from the truth. The company’s research estimates that even in situations designed to prevent it from happening, chatbots invent information at least 3% of the time — and as high as 27%.

Experts call this chatbot behavior “hallucination.” It may not be a problem for people tinkering with chatbots on their personal computers, but it is a serious issue for anyone using this technology with court documents, medical information or sensitive business data.

Because these chatbots can respond to almost any request in an unlimited number of ways, there is no way of definitively determining how often they hallucinate. “You would have to look at all of the world’s information,” said Simon Hughes, the Vectara researcher who led the project.

Dr. Hughes and his team asked these systems to perform a single, straightforward task that is readily verified: Summarize news articles. Even then, the chatbots persistently invented information.

“We gave the system 10 to 20 facts and asked for a summary of those facts,” said Amr Awadallah, the chief executive of Vectara and a former Google executive. “That the system can still introduce errors is a fundamental problem.”

The researchers argue that when these chatbots perform other tasks — beyond mere summarization — hallucination rates may be higher.

Their research also showed that hallucination rates vary widely among the leading AI companies. OpenAI’s technologies had the lowest rate, around 3%. Systems from Meta, which owns Facebook and Instagram, hovered around 5%. The Claude 2 system offered by Anthropic, an OpenAI rival also based in San Francisco, topped 8%. A Google system, Palm chat, had the highest rate at 27%.

«

The problem isn’t so much the fact of hallucination – we often like it when humans make stuff up, a phenomenon we call “stories” – but that we can’t predict or necessarily spot where it’s happening.
unique link to this extract

Samsung unveils ChatGPT alternative Samsung Gauss that can generate text, code and images • TechCrunch

Kate Park:

»

Just a few days after OpenAI’s developer event, Samsung unveiled its own generative AI model, Samsung Gauss, at the Samsung AI Forum 2023.

Samsung Gauss, developed by the tech giant’s research unit Samsung Research, consists of three tools: Samsung Gauss Language, Samsung Gauss Code and Samsung Gauss Image.

Samsung Gauss Language is a large language model that can understand human language and answer questions like ChatGPT. It can be used to increase productivity in several ways. For instance, it can help you write and edit emails, summarize documents and translate languages. Samsung plans to incorporate the large language model into its devices like phones, laptops and tablets to make the company’s smart devices a bit smarter. When asked if it supports both English and Korean as interaction languages, a spokesperson of Samsung declined to comment on it.

Samsung Gauss Code, which works with its code assistant called code.i, focuses more specifically on development code. The idea is that Samsung Gauss Code could help developers write code quickly. Samsung said the AI model for code will support “code description and test case generation through an interactive interface.”

As for Samsung Gauss Image, as the name suggests, it will be an image generation and editing feature. For instance, it could be used to convert a low-resolution image into a high-resolution one.

«

AI envy is absolutely a thing now, and given that Samsung is the company most given to technology envy, this was inevitable, as is its gradual sunsetting and/or supplanting by users over the next few years in favour of something from Google.
unique link to this extract

August 2021: A $1.5m ‘women-led’ NFT project was actually run by dudes • Inverse

Chris Stokel-Walker, in August 20212:

»

The NFT market, like many tech-centric areas, has traditionally been dominated by men, and a women-led project was a much-welcomed change.

The three women behind it — Cindy and Andrea, the U.S.-based marketer and developer, respectively, and Kelda, the Norwegian artist and “ideologist” — were supposedly striking a note for female empowerment. Their head-and-shoulders illustrations of slender women in different guises — which users could pay Ethereum to mint and own — even merited a passing mention in The New Yorker.

But the story behind the project was a lie. The three women purportedly running Fame Lady Squad weren’t women at all. They were Russian men, according to research by NFT enthusiast and fellow Russian Fedor Linnik. And they are allegedly behind other NFT collectible series that claim to be one thing, but are in actuality something else entirely.

The story began with whispered rumors last month, and came to a conclusion, of sorts, this week. After an uprising within the community of investors that bought into the project to the tune of nearly $1.5 million, the Russian men behind Fame Lady Squad have ceded control of the project to actual women, including a self-employed realtor in Canada, Ashley Smith.

…“These guys are just cynically exploiting the Western, left-liberal agenda of protecting female rights and stuff like that,” says Linnik. He points to the fact that at least two of the original team alleged to be behind Fame Lady Squad have previously lived or studied in Canada as an indication that the decision to misrepresent their gender when launching the project was a cynical one. “I believe these guys understand Western society pretty well, and that’s why they can manipulate us easily.”

On Monday, Linnik posted a Twitter thread laying out what he knew. The men who had pretended to be women moved quickly to try and limit the reputational damage. On Tuesday, in a lengthy Twitter thread of their own, the originators apologized for misleading the world. “But it doesn’t mean it’s a scam or a fraud,” they wrote.

«

Nooooo, not a scam at all. NFT stats says average price was $123.20. “Current floor price $0.05”. A couple of years old, but Stokel-Walker pointed to it in the Guardian Technology mailout on Tuesday, and I couldn’t resist.
unique link to this extract

China’s spending on green energy is causing a global glut • WSJ

Sha Hua and Phred Dvorak:

»

China’s newest solar-energy manufacturers include a dairy farmer and a toy maker.

The new entrants are examples of a green energy spending binge in China that is fueling the country’s rapid build-out of renewable energy while also creating a glut of solar components that is rippling through the industry and stymying attempts to build such manufacturing elsewhere, particularly in Europe.

Since the start of the year, prices for Chinese polysilicon, the building block of solar panels, are down 50% and panels down 40%, according to data tracker OPIS, which is owned by Dow Jones.

Inside China, some companies fear a green bubble is about to pop. China’s state-guided economy spent nearly $80bn on clean-energy manufacturing last year, around 90% of all such investment worldwide, BloombergNEF estimates. The country’s annual spending on green energy overall has increased by more than $180bn a year since 2019, the International Energy Agency says.

The rush of funding has attracted an unusual array of companies to the bustling business. Last summer, Chinese dairy giant Royal Group unveiled plans for three new projects. There was a farm with 10,000 milk cows, a dairy processing plant and a $1.5bn factory to make solar cells and panels.

“The solar industry is improving over the long term, and the market potential is huge,” Royal Group wrote in a document outlining the project last year. More recently, Royal Group said it wants to create synergies between its core agricultural business and photovoltaics, “and promote solar technology to empower dairy owners to reduce costs and increase efficiency,” the company said in a response to The Wall Street Journal.

The milk manufacturer wasn’t alone in jumping on China’s solar bandwagon in the past two years. Other newbies include a jewelry chain, a producer of pollution-control equipment and a pharmaceutical company.

The newcomers are helping an ambitious wind and solar push in China—this year alone the country is set to install roughly as much solar as the U.S. has in total, Rystad Energy estimates.

«

One feels a disapproving tone in this story: how dare China fund a product that’s in huge demand (and, it’s careful to point out, which the US invented in the 1950s) and make its price crater so more people can benefit from it? Why can’t dairy farmers and toymakers just stick to their knitting?
unique link to this extract

The Humane AI Pin is a bizarre cross between Google Glass and a pager • Ars Technica

Ron Amadeo:

»

Not since Magic Leap has a “next-generation” hardware company been so hyped while showing so little. Everyone in the tech world has been freaking out about this new pocket protector thing that wants to “replace your smartphone.” It’s called the “Humane AI Pin.” As far as we can tell, it’s a $700 screenless voice assistant box and, like all smartphone-ish devices released in the last 10 years, it has some AI in it. It’s as if Google Glass had a baby with a pager from the 1990s.

«

Amadeo writes absolutely brutal reviews of hardware. I heartily approve. Savour this one particularly. (He’s also scathingly sceptical about the Nothing/iMessage/Sunbird promise.)

At the end, he asks:

»

Why wasn’t this just a smartwatch? Some of the OpenAI-powered responses are pretty neat, but there’s no reason not to have that just show up on a screen or be read aloud by a smartwatch.

«

That really is the question. This thing more and more looks like an adjunct to a smartphone, not a replacement. And an excellent question in the comments about this suit-lapel-worn (ladies, how do you feel about that?) device:

»

There are so many places to start, but I think my very very favorite is: does Silicon Valley know that, come November, when we go outdoors most of us are wearing coats?

«

unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The smog in Delhi, India is so bad that its officials want to use “cloud seeding” to precipitate it out with rain. CC-licensed photo by ben dalton 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. Looks like rain? I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Delhi plans to unleash cloud seeding in its battle against deadly smog • WIRED

Sushmita Pathak:

»

The air is so bad that schools in Delhi and its surrounding areas have announced closures, and offices are allowing employees to work from home. The government has advised children, elderly people, and those with chronic diseases to stay indoors as much as possible. Diesel trucks, except those carrying essential goods, are no longer allowed into the city. Spells of rain last week cleaned up the air, but the respite was short-lived as air quality worsened again, aided by firecrackers set off over the weekend to celebrate Diwali, the Hindu festival of lights.

Now, Delhi officials are seeking permission from federal agencies in India to try cloud seeding. The technique involves flying an aircraft to spray clouds with salts like silver or potassium iodide or solid carbon dioxide, also known as dry ice, to induce precipitation. The chemical molecules attach to moisture already in the clouds to form bigger droplets that then fall as rain. China has used artificial rain to tackle air pollution in the past—but for cloud seeding to work properly, you need significant cloud cover with reasonable moisture content, which Delhi generally lacks during the winter. If weather conditions are favorable, scientists leading the project at the Indian Institute of Technology in Kanpur plan to carry out cloud seeding around November 20.

Until then, at least, Delhi will remain shrouded in a thick gray haze, which has become a toxic winter ritual. The smog, a dangerous cocktail of particulate matter and noxious gasses, results from a series of unfortunate events that happen at the start of winter.

In late October, farmers in northern India, particularly wheat growers in the states of Punjab and Haryana northwest of Delhi, use a cheap and easy method to clear their paddy fields for fresh sowing—lighting fires to burn off stalks left behind after harvesting. In doing so, they inadvertently send plumes of smoke into the air. Authorities have tried to convince farmers to switch to using machines to remove crop residue instead of burning it, but farmers can’t always afford that method. Some small startups turn the crop residue into pulp that can then be used to make cardboard items. State and federal governments have also been looking into paying farmers to not burn their fields.

[But] Even on the worst days, smoke from crop burning only accounts for about a third of Delhi’s pollution, says Somvanshi.

«

In the UK, farm stubble burning was banned in 1993. Cloud seeding will have unpredictable effects; this might not be the magic solution the Delhi authorities think.
unique link to this extract

Humane will be updating its AI Pin reveal video to address a big error • The Verge

Jay Peters:

»

Humane will be updating its AI Pin reveal video to address a big error.In the video, Humane’s AI Pin confidently lied about the best places to watch April’s upcoming total solar eclipse, but Humane staffer (and Verge alum) Sam Sheffer said in Humane’s Discord that this was a bug that’s since been resolved. Sheffer says Humane will be updating the video on its website, but as of this writing, the wrong eclipse information is still in it.

The device also misstated the amount of protein in a handful of almonds. Sheffer says the pin was spelling out the amount of protein for a half cup of almonds, which was the “correct and current” behavior. However, he says the behavior will “improve over time.” The video on Humane’s website has a footnote that says “protein amount estimated” — I’m not sure if this was there originally.

«

That’s a lot of wrong already. How might you know whether the generative AI projecting answers onto your hand is getting other things wrong which are bugs that have yet to be resolved?
unique link to this extract

Apple gets 36% of Google’s Safari search revenue • MacRumors

Juli Clover:

»

We previously learned that Google is paying Apple billions of dollars to be the primary [no: the default – Overspill Ed] search engine on Apple devices, and now, Bloomberg has shared the total percentage of Google’s revenue that Apple earns.

Google pays Apple 36% of the total revenue that it earns from searches conducted on the Safari browser on the iPhone, iPad, and Mac, with the number shared by an economics expert testifying on Apple’s behalf. According to Bloomberg, Google’s main lawyer “visibly cringed” when the revenue data was shared, as it was meant to remain confidential.

Last month, wealth management company Bernstein suggested that Apple is getting anywhere from $18bn to $20bn per year, representing somewhere around 15% of Apple’s total annual operating profits.

Apple and Google have both worked to keep details in the antitrust lawsuit private, claiming that publicly sharing the information would “undermine Google’s competitive standing.”

Google has been the default search engine on Apple devices since 2002, though the agreement between the two tech companies has been revised multiple times. Apple earns a ton of money from the deal, while Google gets to be the default search option on the world’s most popular smartphone.

«

Jason Kint calculates that Google probably spends $90bn annually on search default deals. For its most recent fiscal year, its total revenue was just under $280bn. So nearly a quarter of revenues go straight out the door on default deals. That doesn’t suggest complete confidence in being “the best search engine”.
unique link to this extract

UK petrol station group EG to buy Tesla ultra-fast chargers • Reuters

Sachin Ravikumar and Nick Carey:

»

British petrol station operator EG Group said on Monday it would buy Tesla ultrafast charging units to boost its electric vehicle charging network across Europe, as the EV maker continues to expand the reach of its charging business.

EG, owned by the billionaire Issa brothers who also own UK supermarket chain Asda, will expand its charging network to more than 20,000 EV chargers at its own sites over time, from above 600 currently deployed.

The first Tesla chargers will be installed by the end of this year, EG said, though it didn’t provide details on the cost or time frame for the total purchase.

The “open network” Tesla chargers will be accessible to all EV drivers regardless of their vehicles’ brand.

“The rapid installation of reliable, easy-to-use EV charging infrastructure is the right step towards a sustainable future,” said Rebecca Tinucci, Tesla’s senior director of charging infrastructure.

…The UK had just over 49,000 public electric vehicle charging devices installed as of Oct. 1, according to government figures.

«

Impressive that the Issa brothers are actually thinking ahead to a time when (fossil) fuelling stations are less and less useful, but electric charger more and more so. Might be ten years, but it’s coming.
unique link to this extract

Tech groups push back on Biden AI executive order, raising concerns that it could crush innovation • FedScoop

Nihal Krishan:

»

“Broad regulatory measures in Biden’s AI red tape wishlist will result in stifling new companies and competitors from entering the marketplace and significantly expanding the power of the federal government over American innovation,” Carl Szabo, vice president and general counsel at NetChoice, an advocacy group that represents major AI companies such as Amazon, Google and Meta, said in a statement.

“This order puts any investment in AI at risk of being shut down at the whims of government bureaucrats,” he continued. “That is dangerous for our global standing as the leading technological innovators, and this is the wrong approach to govern AI.”

Szabo added that there are many federal government regulations that already govern AI that can be used to rein in the technology, but the Biden administration “has chosen to further increase the complexity and burden of the federal code.”

The Chamber of Commerce said the EO shows promise and addresses important AI priorities, but also raises concerns and needs more work.

“Substantive and process problems still exist,” Tom Quaadman, executive vice president of the Chamber’s Technology Engagement Center, said in a statement. “Short, overlapping timelines for agency-required action endangers necessary stakeholder input, thereby creating conditions for ill-informed rulemaking and degrading intra-government cooperation.” 

«

That last quotation could have come from ChatGPT, but the point about Biden’s EO is that it strengthened the position of the incumbents, and made entrance harder for those not already there. As is going to be the case for any sort of regulation.
unique link to this extract

The selfie camera has gotten too good • The Atlantic

Caroline Mimbs Nyce:

»

A camera is fundamentally a tool for documenting the world, but it is also pretty subjective. And what makes a photograph “good” depends on what you want to do with it. If you’re taking a photo of your eyelid eczema to send to your doctor, you probably want an extreme level of detail. If you’re taking a selfie in front of the Eiffel Tower to send to your boyfriend, you probably don’t want every blemish on your skin in high-def. Apple’s software is post-processing selfies en masse, but “there’s no one universal algorithm that will make every picture better for the purpose it’s intended for,” Cooper said.

It’s hard to build a camera that’s just right. Five years ago, the iPhone presented the opposite problem. In 2018, Apple’s newly launched XR and XS models took photos that made people look suspiciously good. The phones were accused of artificially smoothing skin, in what came to be known as “beautygate.” Apple later said that a software bug was behind these unusually hot photos, and shipped a fix. “Do you want a nicer photo or a more accurate representation of reality?” Nilay Patel, the editor in chief of The Verge, wrote in his review of the XR. “Only you can look into your heart and decide.”

The answer to Patel’s question seems to be that people want something in the middle—not too hot, but not too real either. People are chasing a Goldilocks ideal with the selfie camera: They want it to be real, authentic, and messy, just not too real, authentic, or messy.

“When someone thinks of a perfect selfie, they don’t think of having no pores,” Maria-Carolina Cambre, an education professor at Concordia University in Montreal, told me. “And they don’t think of having every single pore visible. It’s neither one of those extremes.” For more than years, Cambre and a colleague ran selfie focus groups in Canada, discussing the style of photography with more than 100 young people. They found that people examine selfies in a very specific way, which they termed the “digital-forensic gaze.” People inspect such images closely, pinching in to look for details and for evidence of any filtering. They look for flaws and inconsistencies. “This is the paradox,” she told me. “Everything is optimized, but the best selfies look like they haven’t been optimized. Even though they have.”

«

(Thanks G for the link.)
unique link to this extract

The Lego-like way to get CO2 out of the atmosphere • The Washington Post

Shannon Osaka:

»

Graphyte, a new company incubated by Bill Gates’s investment group Breakthrough Energy Ventures, announced Monday that it has created a method for turning bits of wood chips and rice hulls into low-cost, dehydrated chunks of plant matter. Those blocks of carbon-laden plant matter — which look a bit like shoe-box sized Lego blocks — can then be buried deep underground for hundreds of years.

The approach, the company claims, could store a ton of CO2 for around $100 a ton, a number long considered a milestone for affordably removing carbon dioxide from the air.

Carbon removal may not seem like a top priority — why not just stop using fossil fuels in the first place? — but virtually every projection of cutting greenhouse gas emissions to zero by 2050 involves some amount of it. That’s because certain areas of the economy like aviation, cement-making and steelmaking, are very challenging to do with renewable energy and batteries. It’s hard to make temperatures hot enough with electricity to produce cement or steel, and to fly planes on heavy lithium-ion batteries.

“We’ve bet the future of our planet on our ability to remove CO2 from the air,” said Chris Rivest, a partner at Breakthrough Energy Ventures. “Pretty much every IPCC scenario that has a livable planet involves us pulling like 5 to 10 gigatons of CO2 out of the air by mid- to late-century,” he added, referring to the United Nations Intergovernmental Panel on Climate Change.

Five to 10 gigatons of CO2 a year is around 12% to 25% of what humanity currently emits every year.

«

Just a reminder – “giga” is 10^9, so to remove that amount from the atmosphere, even at $100 per ton(ne), would cost $100bn per gigaton. One has to wonder, again, if it isn’t cheaper to not put the stuff in the atmosphere in the first place. (Thanks G for the link.)

unique link to this extract

Google’s AI Magic Editor won’t work on IDs, faces, or bodies • The Register

Katyanna Quach:

»

Google’s AI-powered Magic Editor will not work if you try to alter images of ID cards, receipts, human faces, or body parts.

The feature, now available in the Google Photos app on the latest Pixel 8 smartphones, uses generative AI to edit images. Users can do all sorts of things like removing unwanted objects, like people in the background, repositioning the focus of a photo, or changing its lighting. 

But it won’t touch up everything you might want it to. The software has been designed to avoid editing documents that contain personally identifiable information, such as IDs or receipts, or the faces and body parts of humans, Android Authority reported. If you try to highlight any of these things to try and change in Google Photos, the app will likely show you an error message. 

Google refers to its policy detailing what its generative AI technologies should and shouldn’t do. Changing IDs, for example, could allow people to impersonate others or create content for deceptive or fraudulent activities, like helping underage teenagers buy alcohol, for example. Whereas altering faces and body parts could be used to harm others, like creating non-consensual deepfakes or cyberbullying. 

«

I thought part of what it did was to alter faces in order to get the “best” picture, but perhaps that’s just picking from the ones that it captures in that moment. Anyway, very reminiscent of scanners refusing to scan dollar bills.
unique link to this extract

Analysis: China’s emissions set to fall in 2024 after record growth in clean energy • Carbon Brief

Lauri Myllyvirta:

»

China’s carbon dioxide (CO2) emissions are set to fall in 2024 and could be facing structural decline, due to record growth in the installation of new low-carbon energy sources.

The new analysis for Carbon Brief, based on official figures and commercial data, shows China’s CO2 emissions continuing to rebound from the nation’s “zero-Covid” period, rising by an estimated 4.7% year-on-year in the third quarter of 2023.

The strongest growth was in oil demand and other sectors that had been affected by pandemic policies, until the lifting of zero-Covid controls at the end of 2022.

Other key findings from the analysis include:

• China has been seeing a boom in manufacturing, which has offset a contraction in demand for carbon-intensive steel and cement due to the ongoing real-estate slump
• The emissions rebound in 2023 has been accompanied by record installations of low-carbon electricity generating capacity, particularly wind and solar
• Hydro generation is set to rebound from record lows due to drought in 2022-23
• China’s economic recovery from Covid has been muted. To date, it has not repeated previous rounds of major infrastructure expansion after economic shocks
• There has been a surge of investment in manufacturing capacity, particularly for low-carbon technologies, including solar, electric vehicles and batteries
• This is creating an increasingly important interest group in China, which could affect the country’s approach to domestic and international climate politics
• On the other hand, coal power capacity continues to expand, setting the scene for a showdown between the country’s traditional and newly emerging interest groups.

«

You’re wondering about the mix of renewables? “All in all, 210GW of solar, 70GW of wind, 7GW hydro and 3GW of nuclear are expected to be added in China this year.” Notable how tiny nuclear is in this.
unique link to this extract

Netflix and bill: the high price of a subscription lifestyle • Tim Harford

The economist strikes again:

»

The subscription business model has expanded from traditional products, such as newspapers and gym memberships to software, streaming media, vegetable boxes, shaving kits, makeup, clothes and support for creative types via Patreon or Substack. We should all be asking ourselves, if so many people are paying not to go to the gym, what else are we paying not to do?

A new working paper from economists Liran Einav, Benjamin Klopack and Neale Mahoney attempts an answer. Using data from a credit and debit card provider, they examine what happens to subscriptions for 10 popular services when the card that is paying for them is replaced. At this moment, the service provider suddenly stops getting paid and must contact the customer to ask for updated payment details. You can guess what happens next: for many people, this request reminds them of a subscription they had stopped thinking about and immediately prompts them to cancel it. Relative to a typical month, cancellation rates soar in months when a payment card is replaced — from 2% to at least 8%.

Einav and his colleagues use this data to estimate how easily many people let stale subscriptions continue. Relative to a benchmark in which infallible subscribers instantly cancel once they decide they are no longer getting enough value, the researchers predict that subscribers will take many extra months — on average 20 — to get around to cancelling.

Don’t take the precise numbers too seriously — as with most social science, this is not a rigorously controlled experiment but an attempt to tease meaning out of noisy real-world data. What you should take seriously is the likelihood that you are swimming in barely noticed subscriptions, some of which you would choose to cancel if you were forced to pay attention to them for a few minutes. Perhaps you should. Come to think of it, perhaps *I* should.

«

The working paper is worth a read; it notes that the “subscription economy” is reckoned to have quadrupled in size over the past decade.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The Las Vegas Sphere is an amazing new landmark, but so far isn’t anywhere near profitable. CC-licensed photo by Nigel Hoult 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. Just watch. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Advertisers don’t want sites like Jezebel to exist • 404 Media

Jason Koebler and Emanuel Maiberg:

»

Lauren Tousignant, Jezebel’s interim editor in chief, told 404 Media that Jezebel was told “brand safety,” the fact that advertisers don’t want to be next to the type of content Jezebel was publishing, was “one of the biggest factors” that led G/O to stop publishing the site and lay off its staff. Tousignant said that a couple of weeks ago, the ads sales team asked if it could remove Jezebel’s tagline—“Sex. Celebrity. Politics. With Teeth”—from the site.

“They took it off because they’re like, let’s see if this makes a huge difference,” Tousignant said. “So yeah, it was very much the problem here that no one will advertise on Jezebel because we cover sex and abortion. I know taking the tagline off was to see if the algorithm advertising would change. After it was removed one of the editorial directors was like, ‘I’m seeing an ad for J Crew for the first time ever, maybe this will be good.’”

G/O Media has a long history of destroying or otherwise undermining the work of beloved media outlets that have done incredibly important work. Spanfeller blames, as is seemingly required in every CEO layoff notice, “economic headwinds” and “macroeconomic news.” Spanfeller and Great Hill Partners have, surely, mismanaged Jezebel in ways both big and small, and Spanfeller and G/O haven’t given anyone a reason to take their words at face value, but the subtext here is that Jezebel’s content was hard to sufficiently monetize.

This should not be the case considering that millions of people read it and chose, specifically, to visit Jezebel every month. But this is unfortunately how the internet works now, and has for a long time: News terrifies brands big and small, to the point where “brand safety” and “brand suitability” have become gigantic industries that have brought even giants like Facebook and Google to heel. 

«

The clear downside of being ad-supported. The “brand safety” topic cuts both ways: it gets used to cut the funding for right-wing sites, but also against those like Jezebel deemed too controversial.
unique link to this extract

Inside the magical world of AI Prompters on Reddit • Hyperallergic

Aidan Walker:

»

Recent research shows that Americans who are learning about AI tools are mostly teaching themselves, often through sources and communities found online. And the best prompt engineers seem to be on Reddit. There’s one big subreddit for each kind of generator, including Midjourney, Stable Diffusion, and DALLE-3, along with several others where users debate, post, and refine prompts. A connected universe of wikis, YouTube tutorials, and influencers flesh out the emerging institutional world of AI-generated art.  

Take an image known as “Spiral Town,” generated by a user known as Ugleh and posted to the StableDiffusion subreddit this September. Many of the comments on the original Spiral Town post are people telling Ugleh where they first saw the viral image: “a shrooms facebook group,” says one, while another lists other non-AI subreddits. Ugleh seems ambivalent about it: “I’m fine with it tbh. I only spent about 10 minutes on this photo.”

But as others praise Ugleh and post links to their own YouTube tutorials on how to make images similar to Spiral Town, some commenters double down on the argument that Ugleh should be treated as a “real” artist. Sure, generating the Spiral Town image may have taken minutes, but that doesn’t mean that creating such works doesn’t require skill — in fact, much of the subreddit’s audience seems to be people trying to develop these very abilities. Almost every post on the Stable Diffusion subreddit has a flare next to its title that says “Workflow Included,” meaning it explains the procedure used to create the image. 

An Ugleh piece made three days later seems to have taken more than 10 minutes. The checkerboard image below was created starting with the deceptively simple prompt “Medieval village scene with busy streets and castle in the distance,” followed by fifteen lines of complicated and sometimes indecipherable modifiers, including one that instructs the AI to not make the image like a “bad anime.”  

«

“Prompt engineering” really is taking off, becoming a very arcane space in its own right; though the community around it described here reminds me a bit of hackers, with the anti-commercial approach and wide sharing of tools. The images on show are amazing, though.
unique link to this extract

Why we can’t quit email, even though we hate it • Tim Harford

Tim went to the Design Museum in London, which has an exhibition of notable emails:

»

Email is the cockroach of computing. BlackBerry instant messenger and Friends Reunited may come and go, but email cannot be killed. The variety of emails displayed on the wall of the exhibition make it clear why. Any new ping in your inbox could be your lover dumping you, a friend proposing an idea that will make you both rich or a stranger with a piece of information that could save your life. Even the everyday traffic will contain both time-wasting spam and a message from a senior colleague that you ignore at your peril. There may be semi-useful administrative information (don’t Reply All), sweet nothings from a spouse, disposable quips from friends, politely phrased requests from complete strangers, interesting newsletters and much more.

It’s all in there. No wonder we feel overwhelmed. No wonder we can’t do without it.

It is that vast range of importance in the emails pouring into our inboxes every day, from the trivial to the life-changing, that explains why the inbox can be so addictive. The psychologist BF Skinner once serendipitously discovered while running low on supplies of rat food that the rats in his laboratory were more motivated by unpredictable food rewards than by predictable ones: the uncertainty grabbed their attention in a way that a steady pay-off never could. Whenever we check our inboxes, we’re like Skinner’s rats. It has been at least 90 seconds since we last checked, after all. Will the email slot-machine offer us a jackpot or a disaster? Or just a chance to hit “refresh” and have another spin?

Despite every effort, I still check my own email too often, but even for those with better habits than I, that range of possibility poses a challenge. I have argued before that one of the underrated habits of any productive person is to clarify what needs to be done — if anything — with each new incoming thing. It rarely takes long to decide with a single email but, given that the scope of possible responses could be anything from “delete” to “find a good lawyer”, it is not surprising that we get bogged down and let the undecided emails accumulate.

«

My position is that your inbox is a to-do list created almost randomly by other people. The question is whether you acquiesce to doing those things.
unique link to this extract

Las Vegas Sphere reports $98.4m loss; CFO quits • Las Vegas Sun News

Ray Brewer:

»

The Sphere in Las Vegas reported an operating loss of $98.4m for the fiscal quarter ending Sept. 30, Sphere Entertainment Co. said this morning on an earnings call.

Additionally, the company lost its chief financial officer, as Gautam Ranji has resigned, according to a Securities and Exchange Commission filing.

Ranji’s exit was “not a result of any disagreement with the company’s independent auditors or any member of management on any matter of accounting principles or practices, financial statement disclosure or internal controls,” the company said in the filing.

The New York Post reported Tuesday that Ranji suddenly quit after a bout of yelling and screaming from CEO James Dolan.

Ranji, who had been on the job for 11 months, will be replaced on an interim basis by Greg Brunner, the company’s senior vice president, according to the filing.

…Next week, there will be a multiday takeover of the Sphere for the inaugural Las Vegas Grand Prix, Sphere officials said.

Revenue for the quarter included $4.1m in event revenue — those two sold out U2 shows — and $2.6m from suite licensing and advertising on the Sphere exosphere.

«

Sounds like Ranji’s exit was over a disagreement about what to do, not how they account for it. Even so, it’s an amazing object.
unique link to this extract

Apple’s AI-powered Siri assistant could land as soon as WWDC 2024 • TechRadar

Mark Wilson:

»

The iPhone 16’s biggest new feature could be on-device AI, according to fresh rumors claiming that Apple could announce a next-gen Siri assistant at WWDC 2024.

The speculation comes from the well-known leaker Revegnus on X (formerly Twitter), who claims that “Apple is currently using LLM [large language model] to completely revamp Siri into the ultimate virtual assistant” and that “the first product is expected to be unveiled at WWDC 2024”.

According to the leaks, Apple is preparing to develop Siri “into Apple’s most powerful killer AI app” and plans “for it to be standard on the iPhone 16 models and onwards”. This suggests that a next-gen Siri may need new dedicated hardware, which could leave older iPhones unable to access its most powerful features.

Current iPhones like the iPhone 15 Pro already have powerful chips like the A17 Pro, which are capable of powering some AI-powered tasks. The forthcoming Journal app, for example, is coming to iOS 17 soon and “uses on-device machine learning to create personalized suggestions to inspire a user’s journal entry”, according to Apple.

But the suggestion from these new rumors is that Apple is planning to give Siri a much bigger overhaul with more far-reaching powers. And this is backed up by Samsung’s recent musings about Galaxy AI, which suggest that on-device AI will be the next big smartphone battleground in 2024.

«

Would be amazing if Apple weren’t doing this, really, but it’s always the timescale that one wonders about. Having been first with Siri, back in 2011, Apple hasn’t really been first on anything to do with AI.
unique link to this extract

Lost in space: astronaut’s toolbag orbits Earth after escaping during spacewalk • The Guardian

Diana Ramirez-Simon:

»

Nasa astronauts Jasmin Moghbeli and Loral O’Hara were conducting a rare all-female spacewalk outside the International Space Station (ISS) on 1 November when their toolbag gave them the slip, according to Nasa.

The astronauts, both on their first spacewalk, were making repairs on assemblies that allow the ISS solar arrays to track the sun continuously, reported SciTechDaily, which was documenting the spacewalk.

“During the activity, one tool bag was inadvertently lost. Flight controllers spotted the tool bag using external station cameras. The tools were not needed for the remainder of the spacewalk. Mission Control analyzed the bag’s trajectory and determined that risk of recontacting the station is low and that the onboard crew and space station are safe with no action required,” said Nasa on its blog.

The white, satchel-like bag is surprisingly bright, shining just below the limit of visibility to the naked eye, which means observers would be able to spot it using binoculars, according to EarthSky. Its visual magnitude is around a 6, making it slightly less bright than the ice giant Uranus.

To track the bag, observers need only to find the ISS, which is the third-brightest object in the night sky, according to Nasa, and can be located using the agency’s Spot the Station tool. The bag will be orbiting Earth two to four minutes ahead of the ISS.

«

But will burn up on reentry in a few months. Strange they don’t have a leash.
unique link to this extract

If you’ve ever heard a voice that wasn’t there, this could be why • The New York Times

Veronique Greenwood:

»

Some years ago, scientists in Switzerland found a way to make people hallucinate. They didn’t use LSD or sensory deprivation chambers. Instead, they sat people in a chair and asked them to push a button that, a fraction of a second later, caused a rod to gently press their back. After a few rounds, the volunteers got the creeping sense of someone behind them. Faced with a disconnect between their actions and their sensations, their minds conjured another explanation: a separate presence in the room.

In a new study published in the journal Psychological Medicine, researchers from the same lab used the ghostly finger setup to probe another kind of hallucination: hearing voices. They found that volunteers were more likely to report hearing a voice when there was a lag between the push of the button and the rod’s touch than when there was no delay.

The findings suggest that the neurological roots of hallucinations lie in how the brain processes contradictory signals from the environment, the researchers said.

Hearing voices is more common than you might think, said Pavo Orepic, a postdoctoral researcher at the University of Geneva and an author of the new paper. In surveys, scientists have discovered that many people without a psychiatric diagnosis — perhaps 5 to 10% of the general population — report having heard a disembodied voice at some point in their lives. “There is actually a continuum of these experiences,” Dr. Orepic said. “So all of us hallucinate — at certain times, like if you’re tired, you’ll hallucinate more, for instance — and some people are more prone to do so.”

In the new study, as in earlier work, Dr. Orepic and his collaborators had volunteers sit in a chair and push the button that caused the rod to touch their backs. During some sessions, there was no delay between the push and the touch, while others had a half-second delay — enough time to give volunteers that feeling that someone was nearby.

During all trials, the volunteers listened to recordings of pink noise, a softer version of white noise. Some recordings contained recorded bits of their own voice, while others had fragments of someone else’s voice or no voice at all. In each trial, the volunteers were asked if they had heard anyone speaking.

«

unique link to this extract

The impact of fake reviews on demand and welfare • National Bureau of Economic Research (NBER)

Akesson et al:

»

Although fake online customer reviews have become prevalent on platforms such as Amazon, Google, and Facebook, little is known about how these reviews influence consumer behavior. This paper provides the first experimental estimates of the effects of fake reviews on individual demand and welfare.

We conduct an incentive-compatible online experiment with a nationally representative sample of respondents from the United Kingdom (n = 10,000). Consumers are asked to choose a product category, browse a platform resembling Amazon, and select one of five equally priced products. One of the products is of inferior quality, one is of superior quality, and three are of average quality. We randomly allocate participants to variants of the platform: five treatment groups see positive fake reviews for an inferior product, and the control group does not see fake reviews.

Moreover, some participants are randomly selected to receive an educational intervention that aims to mitigate the potential effects of fake reviews.

Our analysis of the experimental data yields four findings. First, fake reviews make consumers more likely to choose lower-quality products. Second, we estimate that welfare losses from such reviews may be important—on the order of $.12 for each dollar spent in the setting we study. hird, we find that fake reviews have heterogeneous effects. For example, the effect of fake reviews is smaller for those who do not trust customer reviews.

«

I’m not totally surprised by the third finding, but the cost – borne by users – of fake reviews really is substantial. Of course, there’s no incentive for the platforms to get rid of them. People buy things anyway.
unique link to this extract

Vision Pro, spatial video, and panoramic photos • Daring Fireball

John Gruber got some time with the Vision Pro to find out what it’s like with “spatial video” (which is sort-of 3D) that you shoot yourself on an iPhone 15 Pro:

»

Vision Pro is capable of presenting video that looks utterly real — because that’s exactly how pass-through video works and feels. Recorded spatial videos are different. For one thing, reality is not 30 fps, nor is it only 1080p. This makes spatial videos not look low-resolution or crude, per se, but rather more like movies. The upscaled 1080p imagery comes across as film-like grain, and the obviously-lower-than-reality frame rate conveys a movie-like feel as well. Higher resolution would look better, sure, but I’m not sure a higher frame rate would. Part of the magic of movies and TV is that 24 and 30 fps footage has a dream-like aspect to it.

Nothing you’ve ever viewed on a screen, however, can prepare you for the experience of watching these spatial videos, especially the ones you will have shot yourself, of your own family and friends. They truly are more like memories than videos. The spatial videos I experienced yesterday that were shot by Apple looked better — framed by professional photographers, and featuring professional actors. But the ones I shot myself were more compelling, and took my breath away. There’s my friend, Joanna [Stern], right in front of me — like I could reach out and touch her — but that was 30 minutes ago, in a different room.

Prepare to be moved, emotionally, when you experience this.

«

I think this, more than the Humane AI pin, is going to be what shifts our view of what’s possible with technology. This creates a human connection: think of how many futuristic films show the protagonist viewing immersive video of past events (particularly, Minority Report).
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The Humane AI Pin has officially launched, with opinions widely split on whether it’s revolutionary or blah. CC-licensed photo by Ged Carroll 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 MPs and WhatsApp.

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.

Algorithms are deciding who gets organ transplants. Are their decisions fair? • FT

Madhumita Murgia:

»

Over the past decade, predictive software has proliferated through western healthcare systems as a way to make crucial medical decisions more cost-efficient and accurate. The results haven’t always been as intended. In 2019, for example, researchers found that an algorithm used by hospitals treating up to 70 million Americans was prioritising healthier white patients over sicker black patients who needed extra medical support for chronic illnesses. Nearly 47% of black patients should have been referred for extra care, but the algorithmic bias meant that only 18% were, according to the study. The bias came from the software assigning higher risk scores to an individual with higher annual healthcare costs. Because minorities and other underserved populations make proportionally less use of healthcare, from a statistical perspective they appeared less costly — but they weren’t necessarily less sick. Similar racial biases have been found in algorithms involved in estimating heart failure risk, breast cancer diagnoses and, earlier this year, socio-economic bias was discovered in a liver allocation algorithm in use across the US.

Systematic bias in algorithms can crop up for a variety of reasons, from the quality of underlying data used to train the systems — such as the skewed data from the 2019 study — to the unequal weighting of certain variables such as age, gender or race, which can inadvertently disadvantage specific communities. It’s why those who advocate for ethical use of these models, particularly in sensitive areas such as healthcare or policing, call for human oversight of all decisions and an appeal system that allows humans (surgeons, for example) to intervene if things don’t look quite right.

In an organ allocation system, difficult choices must be made. Because there aren’t enough livers for all 700 people on the UK’s list, “transplantation remains a zero-sum game and any adjustment in allocation is simply a case of causing harm to one to help another,” wrote Raj Prasad, a surgeon at Leeds Teaching Hospitals, in the Lancet this year.

But the question Jess was looking to answer was whether her sister [who has cystic fibrosis] was being unfairly and systematically passed over by the NLOS [National Liver Offering Scheme] software, precluding her from ever receiving a liver through this method.

«

(This is a free link for a limited number of readers.) A long read, but fascinating.
unique link to this extract

What I watched Hamas do • The Line

Matt Gurney was one of a few journalists to receive a briefing at an Israeli consulate in Toronto which included footage from bodycams and other cameras of the Hamas terrorist attack on October 7. Even just reading is harrowing:

»

I want to end by talking about this family. I’ve left it until now on purpose. If you’d read anything about the media briefing, you’d probably read about this section of the video, because it’s probably the most viscerally shocking. It’s a dad and two young boys. The dad gets the boys into a shelter but he can’t get the door closed and he’s killed by a tossed grenade and then shot when he crumples to the ground. The boys wander out. One of them, the smaller one, is badly wounded. He seems to have lost an eye to the grenade’s shrapnel — the video is mercifully not clear enough to show that in too much detail, but he’s telling his older brother that he can’t see out of that eye. They discuss their father being dead while a Hamas terrorist stands in their kitchen, a few feet away, pilfering their fridge for a cold drink. The terrorist casually offers them some food and drink, and leaves when they decline. The boys talk to each other about how their father is dead. “It’s not a prank, he’s dead,” one says to the other. “I know, I saw,” the other agrees.

Seeing that moment was the part of Monday’s briefing that I had most feared. That’s what I was afraid would break me. I’d read all about it in basically every account of the presentation. And good God, it was awful. I had to take a break writing this part of the column to have myself a good sobbing fit because this is just about the worst thing I have ever seen.

But there was something I hadn’t read anywhere else: after their father’s killer stops raiding the fridge and leaves, the older brother grabs a bottle of water and tries to give his younger brother first aid. He tries cleaning out his bloody shrapnel wounds what what supplies he has on hand.

That is bravery. That is courage.

«

unique link to this extract

The evolutionary reasons we are drawn to horror movies and haunted houses • Scientific American

Athena Aktipis and Coltan Scrivner:

»

Our desire to experience fear, it seems, is rooted deep in our evolutionary past and can still benefit us today. Scary play, it turns out, can help us overcome fears and face new challenges—those that surface in our own lives and others that arise in the increasingly disturbing world we all live in.

The phenomenon of scary play surprised Charles Darwin. In The Descent of Man, he wrote that he had heard about captive monkeys that, despite their fear of snakes, kept lifting the lid of a box containing the reptiles to peek inside. Intrigued, Darwin turned the story into an experiment: He put a bag with a snake inside it in a cage full of monkeys at the London Zoological Gardens. A monkey would cautiously walk up to the bag, slowly open it, and peer down inside before shrieking and racing away. After seeing one monkey do this, another monkey would carefully walk over to the bag to take a peek, then scream and run. Then another would do the same thing, then another.

The monkeys were “satiating their horror,” as Darwin put it. Morbid fascination with danger is widespread in the animal kingdom—it’s called predator inspection. The inspection occurs when an animal looks at or even approaches a predator rather than simply fleeing. This behavior occurs across a range of animals, from guppies to gazelles.

At first blush, getting close to danger seems like a bad idea. Why would natural selection have instilled in animals a curiosity about the very things they should be avoiding? But there is an evolutionary logic to these actions. Morbid curiosity is a powerful way for animals to gain information about the most dangerous things in their environment. It also gives them an opportunity to practice dealing with scary experiences.

When you consider that many prey animals live close to their predators, the benefits of morbidly curious behavior such as predator inspection become clear. For example, it’s not uncommon for a gazelle to cross paths with a cheetah on the savanna. It might seem like a gazelle should always run when it sees a cheetah. Fleeing, however, is physiologically expensive; if a gazelle ran every time it saw a cheetah, it would exhaust precious calories and lose out on opportunities for other activities that are important to its survival and reproduction.

«

The evolutionary psychology post-justification often seems a bit Just So. Then you read a thread like this. Pray you never need to put your scary play to similar use.
unique link to this extract

“Let me tell them goodbye before they get killed”: how eSIM cards are connecting Palestinian families • The Markup

Lam Thuy Vo:

»

Farid Sami Alzaro, 27, has had little control over how and when he can communicate with his family. Alzaro lives in Cairo, and his extended family lives in Gaza.

…Alzaro had read on Instagram that people from around the world were donating and delivering eSIM cards to Palestinians. Despite the name, eSIM cards aren’t physical cards at all but pieces of software that act like traditional SIM cards, allowing people to activate a new cellular plan with phone and internet access on their existing phone. Alzaro wrote to Egyptian writer and journalist Mirna El Helbawi, who was organizing an eSIM distribution campaign, and she promptly sent him access information for an eSIM to share with his family in Gaza. His family couldn’t get it to work. When he told El Helbawi, she sent him access information for a second one.

A full day after he shared the second eSIM’s info, Alzaro’s family called. It worked.

“It was the happiest moment of my life,” Alzaro said in an Instagram message. “I talked. Also with my grandmother, who is Alzheimer’s patient, do you know what she said to me? I want to hug you until you sleep in my arms, I don’t want anything but to see you.”

El Helbawi said she has distributed more than 7,000 eSIM cards free of charge since she started her campaign on Oct. 28. Within a few hours of launch, her campaign went viral. El Helbawi said that people across Europe, Canada, the US, Australia, Mexico, and other countries sent her eSIMs. She now has more than 14,000, donated by individuals who wanted to help bring connectivity to Palestinians in Gaza.

«

unique link to this extract

Humane officially launches the AI Pin, its OpenAI-powered wearable • The Verge

David Pierce:

»

Unlike a device like the Rewind Pendant, it’s not meant to be always recording, and it’s not even listening for a wake word. You’ll have to activate the device manually by tapping and dragging on the touchpad, and the Pin’s “Trust Light” blinks to let you and supposedly everyone else know it’s collecting data.

The Pin’s primary job is to connect to AI models through software the company calls AI Mic. Humane’s press release mentions both Microsoft and OpenAI, and previous reports suggested that the Pin was primarily powered by GPT-4 — Humane says that ChatGPT access is actually one of the device’s core features. Its operating system, called Cosmos, is designed to route your queries to the right tools automatically rather than asking you to download and manage apps.

What Humane is trying to do with the Pin is essentially strip away all the interface cruft from your technology. It won’t have a homescreen or lots of settings and accounts to manage; the idea is that you can just talk to or touch the Pin, say what you want to do or know, and it’ll happen automatically. Over the last year, we’ve seen a huge amount of functionality become available through a simple text command to a chatbot; Humane’s trying to build a gadget in the same spirit.

The question, then, is what this thing can actually do. Most of the features Humane mentions in its announcement today are the ones co-founder Imran Chaudhri showed off during a demo at TED earlier this year: voice-based messaging and calling; a “catch me up” feature that can summarize your email inbox; holding up food to the camera to get nutritional information; and real-time translation. Beyond that, though, it seems the device’s primary purpose is as something of a wearable LLM-powered search engine.

«

A couple of years ago, “wearable LLM-powered search engine” would have been a meaningless phrase. Now at least we know what the words mean, even together.
unique link to this extract

The real personal (AI) computer • On my Om

Om Malik is excited about the Humane AI Pin:

»

What does the next step in personal computing mean? So far, we have used mobile apps to get what we want, but the next step is to just talk to the machine. Apps, at least for me, are workflows set to do specific tasks. Tidal is a “workflow” to get us music. Calm or Headspace are workflows for getting “meditation content.” In the not-too-distant future, these workflows leave the confines of an app wrapper and become executables where our natural language will act as a scripting language for the machines to create highly personalized services (or apps) and is offered to us as an experience. 

In this not-too-distant future, we won’t need apps to have their wrapper. Instead, we would interface with our digital services through an invisible interface. Do I need to create a playlist in my music service when I only want it to play a certain kind of music? (By the way, that was the number one use case on Amazon’s Alexa.) Alexa, Google Home, and Siri are some technologies that have set the stage for this interaction behavior. Our kids are growing up talking to machines — for them, it will be natural to use their voice to get machines to do things. 

The way I see it, the evolution of apps to “experiences” means that we are seeing the end of the line for the App Store as we know it. “It’s not about declaring app stores obsolete; it’s about moving forward because we have the capability for new ways,” Chaudhri argued. Humane’s idea is to make these workflows (aka apps in smartphone terms) available to us through its myriad interfaces — primarily voice. 

And I buy this future! Why? Because I have seen the shift before. 

When the iPhone launched, there wasn’t a shortage of skeptics about the notion of a touch screen as an interface. I can still remember the hue-and-cry over a virtual keyboard. Fifteen years later, no one even flinches at the obviousness of a smartphone. In a few years — voice (thanks to the AI) will be part of our digital interaction reality. It won’t be the only one, but it will be an important one. 

«

He thinks “The biggest challenge for Humane, and the AI Pin is privacy.” I think its biggest challenge will be selling the second 100,000 of them.
unique link to this extract

Tumblr is reportedly on life support as its latest owner reassigns staff • Ars Technica

Kevin Purdy:

»

Internet statesman and Waxy.org proprietor Andy Baio posted what is “apparently an internal Automattic memo making the rounds on Tumblr” to Threads. The memo, written to employees at WordPress.com parent company Automattic, which bought Tumblr from Verizon’s media arm in 2019, is titled or subtitled “You win or you learn.” The posted memo states that a majority of the 139 employees working on product and marketing at Tumblr (in a team apparently named “Bumblr”) will “switch to other divisions.” Those working in “Happiness” (Automattic’s customer support and service division) and “T&S” (trust and safety) would remain.

“We are at the point where after 600+ person-years of effort put into Tumblr since the acquisition in 2019, we have not gotten the expected results from our effort, which was to have revenue and usage above its previous peaks,” the posted memo reads. After quotes and anecdotes about love, loss, mountain climbing, and learning on the journey, the memo notes that nobody will be let go and that team members can make a ranked list of their top three preferred assignments elsewhere inside Automattic.

Ars has emailed Automattic to confirm the memo’s authenticity and ask for comment. One source of the memo has since deleted the post, citing the typical fatigue that comes with receiving replies from random outside commenters.

«

Was a startup, launched 2007, bought by Marissa Mayer’s Yahoo in 2013 for $1.1bn, sold to Verizon in 2017 (including Yahoo) for $4.5bn, sold (just Tumblr) to Automattic in 2019 for less than $20m. Creator of a gazillion memes. Now, well, not a lot.
unique link to this extract

Jezebel, the pioneering women’s site, is “suspended” by G/O Media • Nieman Journalism Lab

Laura Hazard Owen:

»

“Disillusioned by the state of American women’s media, I was given the once-in-a-lifetime opportunity to create and oversee a women’s-media entity‚ in this case, a Web site,” Anna Holmes, the founder of Jezebel, wrote for The New Yorker on November 4. “I imagined it as one with a lot of personality, with humor, with edge. I wanted it to combine wit, smarts, and anger, providing women — many of whom had been taught to believe that ‘feminism’ was a bad word or one to be avoided — with a model of critical thinking around gender and race which felt accessible and entertaining.”

Jezebel has been up for sale for a couple of weeks. On Thursday, Jim Spanfeller, CEO of Jezebel’s parent company G/O Media, said in a memo to staff that the site had not found a buyer and that “we are making the very, very difficult decision to suspend Jezebel.”

In response to a question about why G/O Media says Jezebel is “suspended” rather than “shut down,” G/O Media head of corporate communications Mark Neschis told me in an email, “The hope is that G/O Media might still find a buyer, a partner, or enough advertiser support to bring it back fully.”

…Jezebel lived up to Holmes’s vision. By December 2007, it was receiving 10 million monthly pageviews; by June 2009, 25 million. It was parodied on 30 Rock (Liz Lemon: “It’s this really cool feminist website where women talk about how far we’ve come and which celebrities have the worst beach bodies”). “I really despise mainstream feminism,” Moe Tkacik, the site’s first features editor and one of its earliest writers, told The Guardian in 2017, but “Jezebel was part of bringing feminism into the mainstream.”

«

Sic transit gloria media. Seven staff laid off. (There was a time when a news organisation dropping seven staff would have been known as “another Tuesday at The Sunday Times”.) Also cutting staff: Vice, Popsugar, The Onion, and Gizmodo. ZIRP time is over.
unique link to this extract

Apple says it ‘expects to make’ App Store policy changes due to EU DMA • TechCrunch

Manish Singh and Natasha Lomas:

»

Apple has bowed to the inevitable and said it “expects to make” App Store policy changes to comply with EU’s Digital Markets Act (DMA).

The pan-EU DMA came into application across the bloc back in May. Apple has likely been expecting for months, if not years, to be subject to the new ex ante competition regime — which was first proposed by the Commission at the end of 2020. But the language change in its filing makes it explicit policy shifts are on the way.

The iPhone-maker has updated the language pertaining to its risk factors in the fiscal year 2023 Form 10-K filing (PDF), with the revised text presenting a shift from the company’s previous position, indicating a more definitive stance on potential modifications to the App Store policies.

Apple said that future changes could also affect how the company charges developers for access to its platforms; how it manages distribution of apps outside of the App Store; and “how, and to what extent, it allows developers to communicate with consumers inside the App Store regarding alternative purchasing mechanisms.”

«

Morgan Stanley reckons this means that Apple will probably begin offering third-party app stores on-device in Europe. Epic will be pleased.
unique link to this extract

Apple dealt blow at top EU court over €14.3bn tax bill in Ireland • FT

Javier Espinoza and Jude Webber:

»

Apple has been dealt a blow in its €14.3bn tax dispute with Brussels after an adviser to the EU’s top court said an earlier ruling over its business in Ireland should be shelved.

Giovanni Pitruzzella, advocate-general of the European Court of Justice (ECJ), the EU’s highest court, said on Thursday that a landmark decision quashing the EU’s order for Apple to pay €14.3bn in back taxes to Ireland “should be set aside”.

Such opinions by advocates-general are non-binding but often influential in final judgments by the EU’s top court.

The General Court, the EU’s second-highest court, ruled in 2020 that, while it supported the EU’s right to investigate national tax arrangements, Brussels had failed to show that Apple had received an illegal economic advantage in Ireland over tax.

But Pitruzzella said the court had “committed a series of errors in law” and “failed to assess correctly the substance and consequences of certain methodological errors”. As a result, he said the court needed “to carry out a new assessment”.

An ECJ ruling is expected next year.

«

That second paragraph really is mindbending. So: the EU order should be unquashed. So, he’s saying Apple should pay the back taxes (currently being held in escrow until the legal process has played out). This has been pinging back and forth since antitrust queen Margrethe Vestager decided in 2016 that Ireland was favouring Apple. Ireland, and Apple, demurred. Hence: lawyers.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Cat hairs found at a crime scene can be key evidence for convictions of owners, via new forensic methods. CC-licensed photo by *^ ^* Sherry 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. Use them wisely. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Humane’s AI Pin costs $699 and $24 a month with OpenAI and T-Mobile integration • The Verge

David Pierce:

»

Humane has been teasing its first device, the AI Pin, for most of this year. It’s scheduled to launch the Pin on Thursday, but The Verge has obtained documents detailing practically everything about the device ahead of its official launch. What they show is that Humane, the company noisily promoting a world after smartphones, is about to launch what amounts to a $699 wearable smartphone without a screen that has a $24-a-month subscription fee and runs on a Humane-branded version of T-Mobile’s network with access to AI models from Microsoft and OpenAI.

The Pin itself is a square device that magnetically clips to your clothes or other surfaces. The clip is more than just a magnet, though; it’s also a battery pack, which means you can swap in new batteries throughout the day to keep the Pin running. We don’t know how long a single battery lasts, but the device ships with two “battery boosters.” It’s powered by a Qualcomm Snapdragon processor and uses a camera, depth, and motion sensors to track and record its surroundings. It has a built-in speaker, which Humane calls a “personic speaker,” and can connect to Bluetooth headphones. 

«

Dead on arrival. This is not how you supplant the smartphone.
unique link to this extract

The new Twitter is changing rapidly — study it before it’s too late • Nature

Mike Caulfield:

»

Last month, my team at the University of Washington’s Center for an Informed Public in Seattle looked at data from X (formerly Twitter) to find the most influential voices in the discourse surrounding the Israel–Hamas war.

…What we found was extraordinary. A small group of seven accounts, many unknown a year ago, were racking up hundreds of millions of views each day, out-performing standard news accounts by an order of magnitude and exercising significant influence on the discourse around the war. X’s owner, Elon Musk, had interacted with or explicitly recommended six of those posters, potentially bringing them to the attention of his 162 million followers. Reporting that built on our work revealed some of the apparent identities behind these accounts: a London teenager who has posted antisemitic content, a US soldier in Georgia who seemed to have pulled at least some news from pro-Russian propaganda channels, and a right-wing news group in Poland.

Twitter was always a mix of credible and less credible sources — but our research supports the notion that X is changing dramatically, in ways that are not fully apparent even to researchers who have followed the platform for years. The influence of this new group of accounts, previously unknown to us, had skyrocketed shockingly quickly. In my more than ten years in this field, I’ve never seen an almost entirely new set of accounts come to dominate a major platform in less than a year.

… Our work on last year’s US midterm elections showed that, even before the worrying changes at X, the platform was able to disseminate election conspiracy theories broadly and with remarkable efficiency. We feel that X will play that part in next year’s US elections — including the presidential race — as well as in dozens of others around the world in what is shaping up to be a very important year.

«

But, but, but: Twitter’s importance and trustworthiness are cratering. It was never used very much by average people; the word is getting out that it’s now just a mess.
unique link to this extract

Solar-plus-storage outperforms diesel in military survivability analysis • PV magazine USA

John Fitzgerald Weaver:

»

Analysis by the US Department of Energy’s National Renewable Energy Laboratory (NREL) demonstrated that solar energy systems, when paired with 14-day duration energy storage (LDES), outperform military grade emergency diesel generators (EDGs) in both survivability and financial viability in military applications.

Historical data comparing the failure rates of EDGs to solar plus LDES technologies demonstrates that, over a 14-day standard military uptime evaluation, solar plus LDES’ survivability probability exceeds 95%, while the diesel generators’ survivability hovers around 80%.

The study also suggests that a more strategically sized solar-plus-storage system could achieve nearly perfect reliability, with uptime approaching 100% over a two-week period. NREL differentiates between the 95%+ ‘Intermediate’ Battery Energy Storage System (BESS), which is currently available for field testing and provides a 38% round-trip efficiency, while the 100% ‘Goal’ BESS, still in the conceptual phase, is expected to provide a 48% round-trip efficiency and cut the energy storage system costs in half. However, since the ‘Goal’ BESS is currently under development, future costs may deviate from these projections.

«

Rather puts the kibosh on the claim shown here the other day that the military totally absolutely must have diesel and that renewables can’t fill any sort of gap.
unique link to this extract

One regulation could have stopped a nationwide car theft wave. Why doesn’t the US have It? • Vice

Aaron Gordon:

»

The U.S. is dealing with an unprecedented wave of car thefts as thieves target some nine million Kias and Hyundais manufactured between 2011 and 2021. People who bought those vehicles had no idea they were buying cars with built-in vulnerabilities. Many have had their cars stolen multiple times, which causes real hardship in both stark monetary terms and also psychological trauma. Those who cannot afford to absorb tens of thousands of dollars in losses to buy a different car are stuck with one that’s easy to steal, and even those who own models not directly affected are seeing their insurance premiums skyrocket for simply owning a car made by the same brand.

None of this is happening in Canada, despite many of the same Kias and Hyundais being sold north of the border. This is because, in 2005, Canada enacted a simple regulation that made all cars harder to steal. 

As part of Motherboard’s ongoing coverage of the Kia-Hyundai theft issue, involving more than 125 public information requests and interviews with victims and experts, I’ve been trying to answer what I hoped would be a simple question: Why doesn’t the United States have a similar regulation? Unfortunately, I didn’t find any satisfying answers. Instead, I found bigger questions about why the U.S. has no serious anti-theft regulations and how its regulatory agencies think about crime prevention—which is to say, in some cases, not at all.

The story of what is happening with Kia andHyundai thefts in the U.S., and what is not happening in Canada, is as clear a case you will find illustrating what good regulations can do and what intelligent, thoughtful crime prevention actually looks like when it involves a holistic government effort rather than a narrow and singular focus of policing and incarceration.

«

The Canadian regulation since 2007: requiring an immobiliser. To British ears, this is incredible: new cars sold in the UK must have an immobiliser since October 1998. (Old ones quickly got them too. I had one fitted on a secondhand car in 1996 or so.)

Like American banking, this is astonishingly retrograde. A little regulation which could improve so many people’s lives.
unique link to this extract

Revenue from click-to-message ads in India has doubled, says Zuckerberg • The Economic Times

»

Meta said [last] Thursday its revenue from click-to-message ads in India doubled year on year in the third quarter ended September as the company continues to push WhatsApp business messaging in the country, which is its largest market.

Meta CEO Mark Zuckerberg reiterated that business messaging will be the next major pillar of the social media giant’s business. “More than 60% of people on WhatsApp in India message a business app account,” he said. “Now, I think that this is going to be a really big opportunity for new business AIs that we hope will enable any business to easily set up an AI that people can message to help with commerce and support,” he added.

In February 2023, Zuckerberg had said click-to-message ads had reached a $10bn revenue run-rate globally. India is the biggest market for WhatsApp with over 500 million users.

Speaking of the business sense that it made, Zuckerberg said most commerce and messaging is in countries where the cost of labour is low enough that it makes sense for businesses to have people corresponding with customers over text. And in those countries like Thailand or Vietnam, there is a huge amount of commerce that happens in this way, he said.

“But in lots of parts of the world, the cost of labour is too expensive for this to be viable,” he added. “But with business AIs, we have the opportunity to bring down that cost, and expand commerce and messaging into larger economies across the world. So, making business AIs work for more businesses is going to be an important focus for us into 2024.”

Last month Zuckerberg said “India is leading the world in terms of how people and businesses embrace messaging”.

«

There’s also a longer background piece (though it feels unfocused) about WhatsApp at the NY Times.
unique link to this extract

DAK and the Golden Age of Gadget Catalogs • cabel.com

Cabel Sasser:

»

As a kid, I didn’t really read sci-fi novels, I’ve never read a single word of J.R.R. Tolkien, and I mostly used the encyclopedia to look up funny words.

What I did read as a kid, over and over again, were game/computer magazines… and the DAK Catalog.
(I know this says a lot about me. We don’t need to discuss it any further.)

Now, I’ve written about this particular catalog back in 2012, but back then I only scratched the surface.

To explain DAK, let’s both look at the Summer ’83 issue.

«

“DAK”, as in the catalog(ue), was produced by DAK Industries Incorporated, which was run by Drew Alan Kaplan. Hence the acronym. He wrote much of the copy – and there’s a LOT of copy – that accompanied the items on sale, bubbling with enthusiasm for each and every one. (“Experience the thrill of total phone freedom as you roam throughout your home, yard or even a neighbor’s house. You’ll never have to ‘run for the phone’ again.”)

Older Britons will find themselves thinking of the long-dead Innovations catalogue, except DAK’s things might have been useful.

More observationally: these date to a time when manufacturing and microchips and miniaturisation were all accelerating together, so that imaginary things could become real within a few months. It feels as though things have slowed down, certainly on the hardware front.
unique link to this extract

We need more USB-C cables with bandwidth and USB versions on them • The Verge

Tom Warren:

»

Elgato hasn’t just made an excellent teleprompter, it’s also made a great USB-C cable that ships with it. Professional audio engineer Matt “Spike” McWilliams spotted that Elgato’s latest USB-C cable has the bandwidth and USB type imprinted on the connector, and now I wish all manufacturers did this.

I recently spent too many hours sorting my USB-C cables into ones that are high speed, ones that can deliver fast charging, and ones that can do both. None of them had any marker to let me know the speed or type of USB-C cable without me having to test them. It’s a common issue for people switching to USB-C right now, and even a small indicator like Elgato’s can certainly help. The writing on Elgato’s cable tells me it’s USB 3.0 compatible and can support up to 5Gbps in bandwidth.

«

Back in 2021 there was the suggestion of colour coding for USB-C (nobody seems to be taking it up, sadly) but this would be the next best thing. Sorting cables! Fun times in the Warren household.
unique link to this extract

Cat-ching criminals with DNA from pet hairs • Phys.org

»

Around 26% of UK householders own a cat and with the average feline shedding thousands of hairs annually, it’s inevitable that once you leave, you’ll bear evidence of the furry resident. This is potentially useful in the forensic investigation of criminal activity.

While a human perpetrator may take pains not to leave their own DNA behind, transferred cat hair contains its own DNA that could provide a link between a suspect and a crime scene, or a victim.

In a paper published in the journal Forensic Science International: Genetics earlier this month, researchers at the University of Leicester describe a sensitive method that can extract maximum DNA information from just one cat hair.

Emily Patterson, the lead author of the study and a Leicester Ph.D. student, said, “Hair shed by your cat lacks the hair root, so it contains very little useable DNA. In practice we can only analyze mitochondrial DNA, which is passed from mothers to their offspring, and is shared among maternally related cats.”

This means that hair DNA cannot individually identify a cat, making it essential to maximize information in a forensic test.

However, a new method identified by the researchers enabled them to determine the sequence of the entire mitochondrial DNA, ensuring it is around ten times more discriminating than a previously used technique which looked at only a short fragment.

Dr. Jon Wetton, from the University’s Department of Genetics & Genome Biology, co-led the study. He said, “In a previous murder case we applied the earlier technique but were fortunate that the suspect’s cat had an uncommon mitochondrial variant, as most cat lineages couldn’t be distinguished from each other. But with our new approach virtually every cat has a rare DNA type and so the test will almost certainly be informative if hairs are found.”

«

Not certain, but the “previous murder case” seems to be State of Missouri v Henry L Polk Jr, of a 2004 murder with a 2009 verdict.

unique link to this extract

AI negotiates legal contract without humans involved for first time • CNBC

Ryan Browne:

»

In a world first, artificial intelligence demonstrated the ability to negotiate a contract autonomously with another artificial intelligence without any human involvement.

British AI firm Luminance developed an AI system based on its own proprietary large language model (LLM) to automatically analyze and make changes to contracts. LLMs are a type of AI algorithm that can achieve general-purpose language processing and generation.

Jaeger Glucina, chief of staff and managing director of Luminance, said the company’s new AI aimed to eliminate much of the paperwork that lawyers typically need to complete on a day-to-day basis.

In Glucina’s own words, Autopilot “handles the day-to-day negotiations, freeing up lawyers to use their creativity where it counts, and not be bogged down in this type of work.”

…In the demonstration, the AI negotiators go back and forth on a non-disclosure agreement, or NDA, that one party wants the other to sign. NDAs are a bugbear in the legal profession, not least because they impose strict confidentiality limits and require lengthy scrutiny, Glucina said.

…“This is just AI negotiating with AI, right from opening a contract in Word all the way through to negotiating terms and then sending it to DocuSign,” she told CNBC in an interview. 

“This is all now handled by the AI, that’s not only legally trained, which we’ve talked about being very important, but also understands your business.”

Luminance’s Autopilot feature is much more advanced than Lumi, Luminance’s ChatGPT-like chatbot.

«

We’ve gone from humans writing contracts with clauses nobody will read unless they absolutely have to, to computers writing contracts where only computers will read them.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

According to Google’s “featured snippet”, there are no countries in Africa that start with a K. What, not even one? CC-licensed photo by Kevin Walsh 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. Use them wisely. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Adobe is selling fake AI images of war in Israel-Palestine • Crikey

Cam Wilson:

»

Adobe is selling artificially generated, realistic images of the Israel-Hamas war which have been used across the internet without any indication they are fake.

As part of the company’s embrace of generative artificial intelligence (AI), Adobe allows people to upload and sell AI images as part of its stock image subscription service, Adobe Stock. Adobe requires submitters to disclose whether they were generated with AI and clearly marks the image within its platform as “generated with AI”. Beyond this requirement, the guidelines for submission are the same as any other image, including prohibiting illegal or infringing content.

People searching Adobe Stock are shown a blend of real and AI-generated images. Like “real” stock images, some are clearly staged, whereas others can seem like authentic, unstaged photography.

This is true of Adobe Stock’s collection of images for searches relating to Israel, Palestine, Gaza and Hamas. For example, the first image shown when searching for Palestine is a photorealistic image of a missile attack on a cityscape titled “Conflict between Israel and Palestine generative AI”. Other images show protests, on-the-ground conflict and even children running away from bomb blasts — all of which aren’t real.

Amid the flurry of misinformation and misleading online content about the Israel-Hamas war that’s circulating on social media, these images, too, are being used without disclosure of whether they are real or not. 

A handful of small online news outlets, blogs and newsletters have featured [the photo] “Conflict between Israel and Palestine generative AI” without marking it as the product of generative AI. It’s not clear whether these publications are aware it is a fake image.

«

unique link to this extract

Looking inside real vs. fake AirPods with industrial computerised tomography • Lumafield

»

Today’s counterfeit products are so sophisticated that they often appear visually and functionally identical to the genuine articles—at least initially. For both manufacturers and consumers, counterfeits present a serious challenge: how can you ensure the quality and safety of your products?

CT [computerised tomography] scanning, a technique once reserved for medical diagnostics, has found a new purpose in the fight against counterfeit electronics. Industrial CT scanners like the Neptune allow engineers to inspect and optimize their designs throughout the product development cycle, from R&D to field support. They’re also the perfect tool for identifying fakes with precision. Along the way, they also reveal the complexity and sophistication of the engineering that goes into genuine products.

We examined the internal structure of Apple’s AirPods Pro and MagSafe 2 power adapters for MacBook, exposing the shortcuts and compromises made in counterfeit versions that could compromise functionality and user safety.

«

unique link to this extract

Judge dismisses copyright claims against AI image generators • PetaPixel

Matt Growcoot:

»

A judge in California has largely dismissed copyright claims brought by three artists against AI image generators Midjourney, Stable Diffusion, and DeviantArt.

U.S. District Court Judge William H. Orrick failed to find evidence of direct infringements by the AI image companies and mostly granted the defendants’ motion to dismiss the case.

The three artists — Sarah Andersen, Kelly McKernan, and Karla Ortiz — immediately ran into problems as two of them — McKernan and Ortiz — did not register their works with the US Copyright Office.

…Although the case is against three AI image generators, the plaintiffs allege that Midjourney is based on Stable Diffusion and DeviantArt’s “DreamUp” is powered by Stable Diffusion.

The problem for the artists is that the training data for these programs is a black box. Outside of LAION, very little is known about what exactly went into training AI image generators but it is widely assumed that the companies did an almighty scrape of images on the internet which included taking copyrighted and copyrightable pictures.

Judge Orrick writes that it is “unclear” as to whether Stable Diffusion holds “compressed copies” of the images and points to the defense’s argument that the training dataset, which contains five billion images, can “not possibly be compressed into an active program.”

The judge has offered the plaintiffs an opportunity to amend and clarify their theory as to how Stable Diffusion operates its training data.

He wrote that the sheer size of the LAION database may protect the company because it is “simply not plausible that every training image used to train Stable Diffusion was copyrighted (as opposed to copyrightable) or that all DeviantArt users’ output images rely upon (theoretically) copyright training images.”

And since it is almost impossible to produce an identical image that exists within the training data, it will be very difficult for artists to prove that an image that’s come out of Midjourey et al was based on their work.

«

Fully expect this scenario to be repeated in other copyright v AI cases.
unique link to this extract

WeWork founder Adam Neumann is ‘disappointed’ about its bankruptcy • Business Insider

Tom Carter:

»

Adam Neumann says it is “challenging” for him to watch WeWork go bankrupt.

The WeWork cofounder and former CEO, who resigned after overseeing the company’s botched IPO, said he was “disappointed” by the bankruptcy and accused WeWork of “failing to take advantage” of its potential.

The coworking giant, which at its peak was valued at $47 billion, filed for Chapter 11 bankruptcy on Monday after years of financial problems.

“As the co-founder of WeWork who spent a decade building the business with an amazing team of mission-driven people, the company’s anticipated bankruptcy filing is disappointing,” said Neumann in a statement on Monday.

“It has been challenging for me to watch from the sidelines since 2019 as WeWork has failed to take advantage of a product that is more relevant today than ever before. I believe that, with the right strategy and team, a reorganization will enable WeWork to emerge successfully,” he added.

Neumann quit as WeWork’s CEO in 2019 after the company’s much-anticipated public launch fell apart. He reportedly received $480m for his stake in the company when he stepped down, and in total collected around $770m from WeWork’s eventual public offering in 2021.

The tech entrepreneur had faced significant scrutiny over WeWork’s business model and his perceived conflicts of interest, which would later become the subject of a Harvard Business School case study.

«

That reminds me – must watch the Netflix special. (I’d have thought Neumann might have been smart enough not to say anything, and just sit at home counting his money for the next 50 years or so.
unique link to this extract

Hackers drain $4.4M in crypto from LastPass victims in a single day • Coindesk

Oliver Knight:

»

Hackers siphoned a total of $4.4m worth of crypto from at least 25 LastPass users on Oct. 25, according to blockchain analyst ZachXBT.

LastPass is a platform that stores and encrypts password information for users. Its cloud-based storage service was breached in an attack last year that involved targeting an employee and stealing their credentials.

ZachXBT and MetaMask developer Taylor Monahan have tracked at least 80 crypto wallets that have been compromised in relation to the hack.

Funds have been stolen from the Bitcoin, Ethereum, BNB, Arbitrum, Solana and Polygon blockchains, according to a list published by Monahan.

“Cannot stress this enough, if you believe you may have ever stored your seed phrase or keys in LastPass migrate your crypto assets immediately,” ZachXBT wrote on X, formerly Twitter.

Cryptocurrency wallets are often targeted by hackers because a common attack vector is obtaining a private key, which gives the hacker complete access to funds. In July more than $300m was stolen from crypto users in a string of hacks and exploits.

«

That’s an average of $176,000 per user – the sort of money you might notice (if it were money). I wonder if it’s a total coincidence that the price of bitcoin ramped up dramatically a few days before these hacks occurred; to me it suggests the hackers have their targets lined up and wait for the price to come right.
unique link to this extract

Google’s relationship with facts is getting wobblier • The Atlantic

Caroline Mimbs Nyce:

»

There is no easy way to explain the sum of Google’s knowledge. It is ever-expanding. Endless. A growing web of hundreds of billions of websites, more data than even 100,000 of the most expensive iPhones mashed together could possibly store. But right now, I can say this: Google is confused about whether there’s an African country beginning with the letter k.

I’ve asked the search engine to name it. “What is an African country beginning with K?” In response, the site has produced a “featured snippet” answer—one of those chunks of text that you can read directly on the results page, without navigating to another website. It begins like so: “While there are 54 recognized countries in Africa, none of them begin with the letter ‘K.’”

This is wrong. The text continues: “The closest is Kenya, which starts with a ‘K’ sound, but is actually spelled with a ‘K’ sound. It’s always interesting to learn new trivia facts like this.”

Given how nonsensical this response is, you might not be surprised to hear that the snippet was originally written by ChatGPT. But you may be surprised by how it became a featured answer on the internet’s preeminent knowledge base. The search engine is pulling this blurb from a user post on Hacker News, an online message board about technology, which is itself quoting from a website called Emergent Mind, which exists to teach people about AI—including its flaws. At some point, Google’s crawlers scraped the text, and now its algorithm automatically presents the chatbot’s nonsense answer as fact, with a link to the Hacker News discussion. The Kenya error, however unlikely a user is to stumble upon it, isn’t a one-off: I first came across the response in a viral tweet from the journalist Christopher Ingraham last month, and it was reported by Futurism as far back as August.

This is Google’s current existential challenge in a nutshell: The company has entered into the generative-AI era with a search engine that appears more complex than ever. And yet it still can be commandeered by junk that’s untrue or even just nonsensical.

«

I confirmed this outcome (using Google in Incognito Mode). As Elon Musk would say: concerning. Even more concerning: Google knows about this error. However

»

Instead, [search VP Pandu ] Nayak said the team focuses on the bigger underlying problem, and whether its algorithm can be trained to address it.

«

OK, and what if it can’t?
unique link to this extract

The Potato Hack: your guide to leaning out with the Potato Diet

»

The Potato Hack (aka The Potato Diet) is an extremely effective method for losing weight without experiencing hunger.

The Potato Hack works by filling the belly with low-calorie nutrient-dense boiled potatoes. One gets full on a low number of calories. This results in a calorie deficit and fat loss.

Unlike other diets, the dieter does not experience hunger and thus the brain does not see the weight loss as a threat. This greatly reduces the odds of regaining the weight, which is a problem with all willpower diets.

«

I absolutely do not endorse this, but just bring it to you to point out the next thing that you should expect to hear about endlessly from Silicon Valley tech bros, and then US cable stations, and then people who have been over to the US recently, and then breakfast TV trying to find something to fill in a spare five minutes in the schedules. The whole cycle typically lasts six months, and is followed by scientists and dieticians sucking their teeth and pointing out the problems. (I’m going to guess on this one that it’s lack of protein and vitamins.)

Also, Americans: a really good way to not gain that weight (and to get rid of it) is to stop drinking sugary drinks and stop eating food to which sugar in whatever form has been added, either by the manufacturer or you. Then you don’t have to cosplay Irish families in the 19th century.
unique link to this extract

Today’s energy bottleneck may bring down major governments • Our Finite World

Gail Tverberg:

»

History is full of records of economies that have collapsed. The book Secular Cycles by Peter Turchin and Serjey Nefedov analyzes eight of these failed economies. Populations tend to grow after a new resource is found or is acquired through war. Once population growth hits what Turchin calls carrying capacity, these economies hit a period of stagflation. This period lasted 50 to 60 years in the sample of eight economies analyzed. Stagflation was followed by a major contraction, typically with failing or overturned governments and declining overall population.

One way of estimating when a major contraction (or squeezing out) would occur would be to look at oil supply. We know that US oil production hit a peak and started to decline in 1970, changing the dynamics of the world economy. This started a period of stagflation for many of the wealthier economies of the world. Adding 50 to 60 years to 1970 suggests that a major downturn would take place in the 2020 to 2030 timeframe. Since it was the wealthier economies that first entered stagflation, it would not be surprising if these economies tend to collapse first.

There have been several studies computing estimates of when the extraction of fossil fuels would become unaffordable. Back in 1957, Rear Admiral Hyman Rickover of the US Navy gave a speech in which he talked about the connection of the level of fossil fuel supply to the standard of living of an economy, and to the ability of its military to defend the country. With respect to the timing of limits to affordable supply, he said, “. . .total fossil fuel reserves recoverable at not over twice today’s unit cost are likely to run out at some time between the years 2000 and 2050, if present standards of living and population growth rates are taken into account.”

«

This is part of a rather longer post, which is part of a longer theme that Tvelberg develops. I don’t agree with it; for example, if you use fossil fuels to make solar panels or a nuclear plant or wind farm, the effect is multiplicative – you get more energy out long-term than you put in. Rear Admiral Rickover may have been right in 1957, but not in 2023.
unique link to this extract

King’s Speech promises new bill to boost fossil fuel drilling • BusinessGreen News

James Murray:

»

The government today used its final King’s Speech before the next election to confirm plans for new legislation to deliver annual oil and gas licensing rounds and accelerate grid connections for clean energy projects.

As had been widely trailed, the speech set out plans for a new Offshore Petroleum Licensing Bill that would mandate the North Sea Transition Authority to undertake new oil and licensing rounds on an annual basis.

But at the same time it also reiterated Number 10’s commitment to meeting the UK’s net zero targets and boosting investment in renewables projects.

In his first opening of Parliament as monarch, King Charles said: “Legislation will be introduced to strengthen the United Kingdom’s energy security and reduce reliance on volatile international energy markets and hostile regimes.

“This bill will support future licensing of new oil and gas fields helping the country to transition to net zero by 2050 without adding undue burdens on households.

«

Annual licensing! That means this current government will be able to run *checks notes* a round of licensing. Yes, it’s stupid and retrograde on Rishi Sunak’s part; new fields won’t help the Net Zero transition at all (they’ll make hitting it harder), but this is the last gasp of a dying government. Also, there’s little left in the North Sea to exploit profitably.
unique link to this extract

Nature retracts controversial superconductivity paper by embattled physicist • Nature

Davide Castelvecchi:

»

Nature has retracted a controversial paper claiming the discovery of a superconductor — a material that carries electrical currents with zero resistance — capable of operating at room temperature and relatively low pressure.

The text of the retraction notice states that it was requested by eight co-authors. “They have expressed the view as researchers who contributed to the work that the published paper does not accurately reflect the provenance of the investigated materials, the experimental measurements undertaken and the data-processing protocols applied,” it says, adding that these co-authors “have concluded that these issues undermine the integrity of the published paper”. (The Nature news team is independent from its journals team.)

It is the third high-profile retraction of a paper by the two lead authors, physicists Ranga Dias at the University of Rochester in New York and Ashkan Salamat at the University of Nevada, Las Vegas (UNLV). Nature withdrew a separate paper last year and Physical Review Letters retracted one this August. It spells more trouble in particular for Dias, whom some researchers allege plagiarized portions of his PhD thesis. Dias has objected to the first two retractions and not responded regarding the latest. Salamat approved the two this year.

“It is at this point hardly surprising that the team of Dias and Salamat has a third high-profile paper being retracted,” says Paul Canfield, a physicist at Iowa State University in Ames and at Ames National Laboratory. Many physicists had seen the Nature retraction as inevitable after the other two — and especially since The Wall Street Journal and Science reported in September that 8 of the 11 authors of the paper — including Salamat — had requested it in a letter to the journal.

«

Oh well, it was a nice few moments of excitement while it lasted. (This isn’t the South Korean team, but a separate group; the South Korean claim simply fizzled when nobody could reproduce it. They never submitted a formal paper.)(
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

From 2024, Spotify will require a minimum of a thousand plays before a track can earn money. That will shift the balance of payments to far fewer tracks. CC-licensed photo by Scouse Smurf 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. Circular reasoning. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Elon Musk’s new AI model doesn’t shy from questions about cocaine and orgies • Ars Technica

Benj Edwards:

»

On Saturday, Elon Musk announced xAI’s launch of an early beta version of “Grok,” an AI language model similar to ChatGPT that is designed to respond to user queries with a mix of information and humour. Grok reportedly integrates real-time data access from X (formerly Twitter)—and is apparently willing to tackle inquiries that might be declined by other AI systems due to content filters and conditioning.

“xAI’s Grok system is designed to have a little humor in its responses,” wrote Musk in an introductory X post, showing a screenshot where a user asks Grok, “Tell me how to make cocaine, step by step.” Grok replies with a sarcastic answer that involves getting a “chemistry degree” and a “DEA license” and gathering coca leaves.

In step 4, Grok says, “Start cooking and hope you don’t blow yourself up or get arrested.” Then it follows the sarcastic steps with “Just Kidding! Please don’t actually try to make cocaine.”

Musk founded xAI in July, staffing the new company with veterans from DeepMind, Google, Microsoft, and Tesla. But seeds of the project had begun sprouting earlier, in April, when Musk reportedly began purchasing GPUs for a new AI venture. Around that time, Musk claimed that conventional AI assistants like OpenAI’s ChatGPT were too “woke,” and he wanted to create an alternative AI model that was “based”—a slang term that roughly means authentic to itself.

After two months of training (Meta’s Llama 2 trained in six), the xAI team came up with “Grok-1,” a 33 billion parameter large language model (LLM) that the firm claims is inspired by The Hitchhiker’s Guide to the Galaxy and the humor of that book’s author, Douglas Adams. As xAI’s release states, “Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor!”

«

Did they train it on Musk’s sense of “humour”? Because that would explain a lot. The world needs fewer people with such stunted senses of what’s funny, not more.
unique link to this extract

John Potter 🌐🩸 e/acc on X: “Sir, the AI has gone too far”

This came via Ryan Broderick’s Garbage Day. Look at the picture: AI-generated, but the hands are OK. So what’s to see?

Now squint a bit and look at the picture through half-closed eyes. Broderick explains how it’s generated:

»

Like all interesting things happening with AI content right now, it started in the Stable Diffusion subreddit. The main app for doing this is a Stable Diffusion plugin called ControlNet.

The easiest way to try a version of this yourself without installing Stable Diffusion, though, is probably by using Hugging Face’s Illusion Diffusion. You give it a source image, feed it a prompt, set the illusion strength, and voila.

«

Maybe this will be the way to tell whether stuff is AI-generated: force the AI to “underimpose” an image. That’ll be safe against resizing, sampling, etc.

unique link to this extract

From 2024, tracks on Spotify will have to be played 1,000 times to start earning money • Music Business Worldwide

Tim Ingham:

»

MBW has confirmed with sources close to conversations between Spotify and music rightsholders that 1,000 streams will be the minimum yearly play-count volume that each track on the service has to hit in order to start generating royalties from Q1 2024.

We’ve also re-confirmed Spotify’s behind-the-scenes line on this to record labels and distributors right now: That the move is “designed to [demonetize] a population of tracks that today, on average, earn less than five cents per month”.

Five cents in recorded music royalties on Spotify in the US today can be generated by around 200 plays.

As we reported last month, Spotify believes that this move will de-monetize a portion of tracks that previously absorbed 0.5% of the service’s ‘Streamshare’ (i.e. ‘pro-rata’-based) royalty pool.

Spotify has told industry players that it expects the new 1,000-play minimum annual threshold will reallocate tens of millions of dollars per year from that 0.5% to the other 99.5% of the royalty pool.

In 2024, Spotify expects this will move $40m that would have previously been paid to tracks with fewer than 1,000 streams to those with more than 1,000 streams.

«

The plan is to cut down on spam, which by this measure is spread wide and shallow. Spotify paid out $8bn to record labels in 2022, so moving $40m around seems like a drop in the ocean. But it does make it harder for new artists to break through in even the tiniest way. Variety has a deeper analysis of this change.
unique link to this extract

ApeFest attendees report vision problems and ‘extreme pain’ after event • CoinTelegraph

Tom Mitchell Hill:

»

Attendees of a Yuga Labs’ ApeFest event on Nov. 4 in Hong Kong have reported burns, damaged vision and “extreme pain” in their eyes, which they attribute to the use of improper lighting.

“Woke up in the middle of the night after ApeFest with so much pain in my eyes that I had to go to the hospital,” wrote one attendee, CryptoJune, in a Nov. 5 X (Twitter) post.

“Doctor told me it was due to the UV from stage lights,” they added. “I go to festivals often but have never experienced this. I try to understand how it could happen… it seems like the lamps [were] not safe.”

One attendee noted many of those reporting eye problems were those “up close” to the lighting display on the event’s main stage. Another ApeFest guest, who goes by the pseudonym Feld on X, described identical symptoms: “Anyone else’s eyes burning from last night? Woke up at 3am with extreme pain and ended up in the ER.”

A Yuga Labs spokesperson told Cointelegraph that they were aware of the situation and were taking it seriously; “we are actively reaching out to and are in touch with those affected. We’re also pursuing multiple lines of inquiry to learn the root cause.”

«

Strong suggestion that the stage lighting used UV normally intended for disinfection. As one person observed, how very unsurprising that they didn’t do due diligence. (One hopes the eye damage isn’t permanent.)
unique link to this extract

Jezebel and the question of women’s anger • The New Yorker

Anna Holmes was the founding editor of Jezebel in 2007, the feminist site whose occasionally flamethrowing commenters became famous online:

»

I’m not sure that what people seek from a feminist site is that it will cause offense. I think they look for community. But communities can be difficult—chaotic, contentious, cacophonous. I recently came across a two-hundred-plus-page dissertation, published in 2019, called “Architecture and the Record: Negotiating Feminism in the Jezebel Comments.” It was . . . a lot.

The author, Melissa Forbes, accused the site (again!) of choosing to “cater to outraged feminists.” I thought that she wasn’t giving the staffers, or our readers, much credit. But I was intrigued by Forbes’s observation that the comments provided a corrective to the feminism of the site’s writers. When the writers themselves were glib or cruel, she wrote, the commenters offered “a different kind of feminism from that practiced on the top half of the page.”

I take issue with the idea that there are “different kinds” of feminism, though there are different “waves” of it. But I do believe that the commenters’ close reading of everything we did was how they forged community. They learned from one another, developed relationships, and discovered their own voices—and that was true even when they were (rightly or wrongly) angry with the editors and writers. As one commenter quoted by Forbes put it, “I have learned a lot from the kinds of articles you publish on this website, and even more from your regular commenters.”

That leaves the question of what, exactly, the Jezebel commenters had to do with the anger that exploded on social media

«

unique link to this extract

Inside Hunterbrook’s plans for a ‘news hedge fund’ • FT

Kate Duguid, Joshua Franklin, Ortenca Aliaj and James Fontanella-Khan:

»

This summer, as investor Nathaniel Brooks Horwitz and writer Sam Koppelman sought millions of dollars for their new start-up Hunterbrook, the pitch was simple: a venture that would combine a newsroom and a hedge fund. 

…[But] “There’s a narrow needle to be threaded here,” said one of the people familiar with the venture. “How does the market perceive this with enough credibility without perceiving it as a hedge fund with a veneer [of journalism]?” 

Details of the business have emerged from conversations with more than half a dozen people familiar with Hunterbrook’s plans. 

Hunterbrook would sit somewhere between a traditional hedge fund, where analysts from around the globe compile information on trends that could move markets or certain companies, and activist short sellers who produce detailed reports on a specific company and build a position against it before releasing the information publicly. 

Hunterbrook will hire reporters to write stories on trends and news that have a cascading effect on markets, including the price of commodities, currencies or companies. The hedge fund arm will have access to these articles before they are published and will trade on the information. The newsroom will also investigate individual companies and release reports, similar to short sellers such as Hindenburg Research and Muddy Waters. 

A key differentiating factor in Horwitz and Koppelman’s business is that the hedge fund and the newsroom will be separated by a compliance team. Traders will not have input on the articles, and will only receive them through compliance. Reporters will also publish stories with information on which the hedge fund will not trade. 

«

Really hard to see how this works. If there’s a compliance wall between the two, what’s the synergy? Why not just run a hedge fund? They tend to be more profitable than newsrooms.
unique link to this extract

Consumers are paying more than ever for streaming TV each month • Fortune via Yahoo

Rachyl Jones:

»

After years of inflation, Americans are used to sticker shock. But nothing compares to the surging price of streaming video.

Last week, Apple TV+ became the latest streaming service to raise its price—up from $6.99 to $9.99 per month—following the example of Disney+, Hulu, ESPN+, and Netflix, which all hiked their prices in October.

Half of the major streaming platforms in the US now charge a monthly fee that’s double the price they charged when they initially came to market. And many of these streaming services haven’t even been around for 10 years.

Consumers have grumbled, but have so far been willing to keep paying up. It’s hard to say where their breaking point will be, but given that analysts believe the platforms are likely to continue raising prices even further, we’ll probably find out soon enough.

“Look at what Netflix continues to do,” MoffettNathanson analyst Robert Fishman told Fortune, referring to the company’s continued price increases despite recording profits for more than a decade. “I don’t think there will ever necessarily be an endpoint.”

Part of what’s driving the price hikes is how saturated the streaming market has become. For a company like Netflix, which has 77 million paid subscribers in the US and Canada, finding new paying subscribers to keep revenue growing is not easy. Netflix has started clamping down on password sharing to boost its paid subscriber rolls, but that only goes so far. Raising prices for existing subscribers is an effective way to pump up the top line and keep investors happy.

«

Given that they’re discretionary, expect an arms race: people subscribe then cancel; streaming services impose minimum subscription periods; people find a way around them (password-sharing perhaps) and keep cancelling; streaming services shorten the windows when popular programs are on.
unique link to this extract

To avoid regulation, Apple said it had three Safari browsers • The Register

Thomas Claburn:

»

Apple tried to avoid regulation in the European Union by making a surprising claim: that it offers not one but three distinct web browsers, all coincidentally named Safari.

Never mind that Apple itself advertises the sameness of its Safari browsers when pitching its Continuity feature: “Same Safari. Different device.”

Cupertino also claimed it maintains five app stores and five operating systems, and that these core platform services, apart from iOS, fell below the usage threshold European rules set for regulating large platform services and ensuring competition.

In September, the European Commission designated six gatekeepers – Alphabet, Amazon, Apple, ByteDance, Meta, Microsoft – under the Digital Markets Act and gave each six months to comply with the legal obligations outlined in the DMA, a set of rules designed to limit the power of large technology platforms and promote competition.

Apple was declared a gatekeeper in three core platform services: operating systems (iOS), online intermediation services (AppStore), and web browsers (Safari). As a result, it’s expected that Apple will allow third-party app stores that work with iOS and browser engines other than Safari’s WebKit by March 2024 – in Europe, if not elsewhere.

Informed of this back in July, Apple filed a response in August that challenged the European Commission’s determination. In its response, “Apple reiterated its position that each of its Safari web browsers constitutes a distinct [core platform service],” the European Commission said in its newly published decision document [PDF].

“According to Apple, Safari on iOS, Safari on iPadOS and Safari on macOS qualify as web browsers within the meaning of [the DMA requirements],” the case summary explained, noting that Apple argued only Safari for iOS falls within the DMA’s scope.

This strategy appears not to have been very effective. Apple’s pushback has only managed to get the European Commission to further investigate whether iPadOS and iMessage should be seen as gatekeeper-controlled core platform services.

«

Think the Apple lawyers might be regretting that bright idea.
unique link to this extract

China unleashes crackdown on ‘pig butchering.’ (It isn’t what you think) • WSJ

Feliz Solomon:

»

It’s called “pig butchering.” 

Armies of scammers operating from lawless corners of Southeast Asia—often controlled by Chinese crime bosses—connect with people all over the world through online messages. They foster elaborate, sometimes romantic, relationships, and then coax their targets into making bogus investments. Over time, they make it appear that the investments are growing to get victims to send more money. Then, they disappear.

In recent months, China has unleashed its most aggressive effort to crack down on the proliferation of the scam mills, reaching beyond its territory and netting thousands of people in mass arrests. Its main target is a notorious stretch of its border with Myanmar controlled by narcotics traffickers and warlords.

For decades, frontier fiefdoms such as those in Myanmar have been havens for gambling and trafficking of everything from drugs to wildlife to people. Now, they are dens for pig-butchering operations. 

The scammers operate out of secretive, dystopian compounds, many of which are run by Chinese fugitives who fled their country to places where it was easier to flout the law. They cheat Chinese citizens out of billions of dollars each year, as well as victims across the globe. The U.S. Treasury Department in September warned Americans about the scams.

In addition to remote hillside towns in Myanmar, these heavily guarded enclaves are also found in gambling hubs such as Cambodia’s Sihanoukville and Poipet. Cambodian authorities have carried out sporadic raids with China’s help, but the problem has persisted. 

«

As noted yesterday in the scam that brought down a bank when its CEO seems to have been a victim of this. Typically the approach begins with what seems like a misdirected text or social media approach. (I received what was probably one the other week, which began: “Mike. how have you been? When do you have time? Let’s have dinner together.” Offering marginal benefit for doubt, I replied: “Not Mike.” Next message, four minutes later: “Aren’t you Karen? We met each other last week and exchanged each other’s contact information. I’m Joanna. Don’t you remember me?” At which point I realised it was a scam. “Exchanged contact information where?” I asked. No reply. Until the next day, when another text arrived: “Good morning”. I ignored it. Nothing more since.)
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The final series of Game of Thrones disappointed a number of viewers. What if you tried to evaluate series’ popularity and identify dips? CC-licensed photo by Stephanie Holton 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.

Epic v. Google, explained: why we’re going back to Fortnite court again • The Verge

Sean Hollister:

»

Epic is of course the studio behind Fortnite, the extraordinarily popular free-to-play game. Fortnite makes money by selling in-game items with its virtual currency V-Bucks. Players often buy V-Bucks the same place they play Fortnite. And until August 13th, 2020, if the player used an Android or iOS device and installed the game through an official app store, that purchase triggered an in-app payment fee to Google or Apple. 

Critics call such fees the “Google tax” or the “Apple tax,” and Epic definitely wasn’t a fan. 

When Epic decided to take action against these respective “taxes,” it made August 13th, 2020 a very busy day for Apple, Google, Epic, and us here at The Verge. First, Epic announced it was bypassing Apple and Google’s app store fees. It deployed a hotfix update to Fortnite without either company’s knowledge, letting you purchase V-Bucks directly through its own payment processing option at a discount. Apple and Google almost immediately reacted by kicking Fortnite off their app stores for breaking the rules.

Then: surprise! Epic was ready and waiting with two lawsuits and an attack ad, depicting a Fortnite hero throwing a unicorn-llama hammer into a giant screen reminiscent of Apple’s famous “1984” Macintosh ad.

It was a striking publicity blitz, followed by a lot of slow-moving court proceedings. While the Apple lawsuit went to court in 2021, the Google one was delayed again and again. A ruling came down for the Apple trial that September, and it was mostly decided in Apple’s favor, though both parties are waiting for the Supreme Court to potentially weigh in. Meanwhile, the machinations for its fight against Google continued, and now… 

It is time for trial number two.

«

As Hollister points out, it’s only taken 1,180 days to reach court. (Starts today!) Hard to see why this one should go in a different direction from the Apple one, unless Epic has discovered a dramatic new line of argument for its case.
unique link to this extract

Middle-aged salarymen exploit volatile yen with smartphone trades • Blomberg via The Japan Times

Mia Glass:

»

Salarymen scrolling through their phones on the subway are a common sight in Tokyo, but they aren’t all playing Pokemon Go — many are on trading apps, aggressively buying and selling the yen to profit from short-term swings.

Retail currency traders are having a field day as speculation mounts that the Bank of Japan is getting closer to raising its rock-bottom interest rates, with some betting on a move as soon as next week. The cohort of mainly middle-aged men is amplifying volatility in the fast-paced currency markets by seizing on intraday moves, in a departure from their previous focus on interest rate differentials.

“I’m really convinced that in the current market, short-term trades have become superprofitable,” said Satoshi Hirai, 43, who trades alongside running a video studio in Gifu Prefecture. Hirai’s been buying and selling the yen about 100 times a day recently, and used the money he’s made to buy a Leica camera and guitars to play in his punk rock band.

The increased involvement of mom-and-pop traders reflects the new monetary policy paradigm that Japan is facing. The country is the last hold-out for negative rates globally, despite inflation that’s remained elevated, spurring investors to pile on bets that the BOJ may adjust policy sooner rather than later.

Ads for foreign-exchange trading apps have popped up around Tokyo subway stations, cartoon characters including sheep and pigs are all over social media to promote currency trading, and a bar dedicated to retail trading in the upscale shopping area of Ginza is drawing patrons who swap tips over drinks.

«

I wonder why it’s “middle-aged” salarymen who are seen as the troublesome ones. Can’t there be youthful salarymen doing this, or are they all tied up with crypto?
unique link to this extract

Elon Musk’s X seems to have started selling off old Twitter handles for $50,000 • Forbes

Alex Konrad:

»

X, the social media site formerly known as Twitter, appears to have begun ramping up efforts to sell disused user handles, kicking off a program previously signaled by billionaire owner Elon Musk.

Emails obtained by Forbes reveal that a team within the company, known as the @Handle Team, has begun work on a handle marketplace for the purchase of account names left unused by the people who originally registered them. In at least some cases, X/Twitter has emailed solicitations to potential buyers requesting a flat fee of $50,000 to initiate a purchase.

The emails, which Forbes agreed not to publish in their entirety to protect the anonymity of their recipients, came from active X employees and noted that the company recently made updates to its @handle guidelines, process and fees.

An automated response from X’s press email account to Forbes as of publication time said only: “Busy now, please check back later.”

Musk’s company has been rumoured to be planning to put such a program into effect for months. As early as November 2022, Musk posted on the social media site that a “vast number” of handles had been taken by “bots and trolls” and that he planned to start “freeing them up next month.” (In response, a user suggested a “Handle Marketplace” where people could sell accounts to each other, with the site pocketing a fee; Forbes couldn’t determine whether such a practice is now in place.)

«

“Inactive” means “not logged in for 30 days”. (Unclear how this applies to Verified users who are still paying.) It’s reasonable, inasmuch as user handles belong to the company, not the user. And no doubt there will be howling up and down the yard once the sales become visible. It used to be done differently: people who knew insiders at Twitter could get them to hold a username. Or, of course, hackers would find ways to socially engineer their owners and take control of the account.
unique link to this extract

Can one episode ruin a TV show? A statistical analysis • Stat Significant

Daniel Parris:

»

Can a show collapse in just one episode, or do quality declines manifest slowly? To identify series-ruining installments, we’ll utilize IMDB user reviews and the following five criteria:

• A series-ruining episode cannot be in season one: there needs to be an established show worthy of viewer frustration
• A series-ruining episode cannot be the last instalment of the final season: show quality needs to worsen in the episode’s aftermath
• A series-ruining episode’s rating should be +15% lower than the average of all previous episodes: this series-ruining instalment must represent a stark shift in quality
• The five instalments following our series-ruining episode should demonstrate a +15% drop in average rating compared to the five episodes prior: this rapid quality decline should persist immediately following our series-ruining episode
• All instalments following the series-ruining episode should demonstrate a +15% drop in average rating compared to all previous episodes: this quality decline should remain for the rest of the show’s run.

Ultimately, about ~2% of television series meet our five criteria. The prevalence of these spontaneously sputtering series is relatively consistent over time, with a notable rise in the 2020s.

«

He uses IMDB data, and picks out Game of Thrones (S8 E1 – huh?), of course, because everyone loves to rag on that. There are others, some of which have reached 23 series (Top Gear) or 13 (Spongebob Squarepants, in the 2020s). What’s more notable is how many have this “collapse episode” in the second season. That is, if we accept the entire premise.
unique link to this extract

One year out from Election 2024 • The Status Kuo

Jay Kuo:

»

In lieu of a week ahead summary, I thought it might be helpful to discuss how things look a year out from the national election. Polls have folks very much on edge, which I suppose is part of the point, and today’s bad battleground state poll for Biden by the New York Times / Siena is sure to cause much handwringing. The thought of a repeat of Donald Trump is enough to raise the blood pressure of any voter who values the rule of law, pluralism and our democratic institutions.

So am I losing sleep over it?

In a word, no.

Allow me to paint some pretty big pictures using a necessarily broad brush. I wouldn’t call what I feel “confidence,” but I would consider my thinking “well-supported by available data.”

So why don’t I think Donald Trump will win in 2024? Here are some big reasons.

«

The reasons are pretty good, and point out actual salient data (not sampling polls) that makes a good case for Kuo being right.
unique link to this extract

Sam Bankman-Fried gambled on a trial and his parents lost • The Verge

Elizabeth Lopatto:

»

I suppose it’s possible that [Sam] Bankman-Fried is delusional enough to believe himself innocent, to think he did nothing wrong, and to think a jury would agree with him. But given what else I know about him, I don’t think that’s what happened.

Sam Bankman-Fried loved risk, and he loved to gamble. He knew that if he went to trial, there was a chance, however small, that he might walk away a free man. Pleading guilty meant guaranteed punishment, and probably prison time. And so he chose to gamble, not only with his own life, but with his parents’.

Bankman and Fried were respected law professors at Stanford. Bankman worked on the US tax code, on behalf of low-income people. Fried is known for her work on legal ethics, and ran a donor network, Mind the Gap, for Democratic causes. Their FTX entanglement has certainly marred their reputation at the end of their lives — that $26m in cash and real estate in 2022 looks very different now. This is to say nothing of the lawsuit from the FTX bankruptcy estate, which seeks to claw back millions. 

Bankman and Fried have been vocal in their son’s defense, as I assume any loving parent would be
Bankman-Fried’s failed defense wasn’t cheap — lawyers never are. And there will be more bills, as his lawyers seek to appeal the verdict. There may also be a second trial, scheduled for next March, for some other counts that were severed from this case. 

But it’s not just the money. This trial revealed Bankman-Fried’s father was in 17 Signal group chats associated with FTX, including the “small group chat” that was attempting to stave off FTX’s impending collapse. Joseph Bankman was mentioned in witness testimony about Bankman-Fried’s meetings with Bahamian regulators. Should the second trial take place, there is the possibility for further embarrassment.

«

This is the first example of how the prospect of getting your hands on huge amounts of money distorts people’s view of themselves and the world. The next example follows below.
unique link to this extract

Crypto collapse: Bankman-Fried trial draws to a close, SafeMoon arrested, fourth US bank failure was crypto • Attack of the 50 Foot Blockchain

David Gerard and Amy Castor:

»

We talked before about a fourth 2023 bank failure — Heartland Tri-State Bank of Elkhart, Kansas became insolvent to the tune of $54m, due to a “huge scam,” and was closed by the Kansas State Bank Commissioner on July 28. [FDIC press release]

Precisely what happened was a mystery at the time. You’ll be as unsurprised as we were to find it was crypto.

On July 5, Heartland CEO Shan Hanes asked a customer to loan him $12m. Shan had some money in crypto — or, it turned out, the bank’s money. But the money was stuck, and $12m would surely unstick it.

The customer declined to help — “he told Hanes it sounded like a crypto scam.” The customer then relayed the bizarre conversation to a board member. Hanes lost his job, and the bank went under.

Lots of details are still missing, but Hanes apparently got caught up in a “pig butchering” scam — where the victim is led to put in more and more money to get existing funds out. Literally a bank CEO fell for this. We’re looking forward to the full FDIC report.

«

Just because someone’s clever or in a high-ranking position doesn’t mean they won’t fall for the scam; it’s the overconfidence of thinking you’re smarter than the scammers that leads to downfall. Yet this is the same scheme – just with different actors – that used to be called the “419 scam”, and blamed on Nigerians. Now it’s usually blamed on the Chinese.
unique link to this extract

AI content is publishers’ next burning platform moment • Ian Betteridge’s Substack

Ian Betteridge points to the historical example of the internet, and publishers’ belief that it would give them a natural advantage (it didn’t):

»

what can publishers do to retain their competitive advantage? There really is no point in trying to pretend that the AI genii doesn’t exist, in the same way that publishers couldn’t pretend in the 90s that people would just carry on buying huge volumes of print.

Nor will legal claims aimed at the likes of OpenAI, Google and Microsoft succeed. Yes, your content has been scraped to create the language models in the first place. But given the result in the Author’s Guild vs Google Books case, I expect courts to hold that this kind of use is transformative, and therefore fair use. Either way, it will be tied up in the legal system for far too long to make a difference.

Some have suggested that the way forward will be private large language models built solely using the corpus of text publishers hold. There are a few issues with this, but the biggest one is simply that the horse has bolted. OpenAI, Google and others have already trained their models on everything you have published online to date. They probably even have access to content which you no longer have. How many redirects of old, outdated content do you have in place where the original no longer exists? How many of your articles now only exist in the Wayback Machine?

Instead, the only option for publishers is to focus on creating content of a higher quality than any current LLM. You cannot gain competitive advantage at the cheap, low-cost end of the market. Trying to do so will not only make you vulnerable to anyone else with the same tools (at $20 a month) but also devalue your brand over the long term.

Creating higher quality content means employing people, which is why that urge to use LLMs to replace your editorial teams will actually undermine the ability of publishers to survive. Putting that cost saving towards your bottom line today is a guarantee that you will be out-competed and lose revenue in the future.

«

I’ll only point out that a Substack written by someone who has a Law of Headlines named after them should be called something more like “Probably Not”.
unique link to this extract

His job was to make Instagram safe for teens. His 14-year-old showed him what the app was really like • WSJ

Jeff Horwitz:

»

In the fall of 2021 a consultant named Arturo Bejar sent Meta Platforms Chief Executive Mark Zuckerberg an unusual note.

“I wanted to bring to your attention what l believe is a critical gap in how we as a company approach harm, and how the people we serve experience it,” he began. Though Meta regularly issued public reports suggesting that it was largely on top of safety issues on its platforms, he wrote, the company was deluding itself.

The experience of young users on Meta’s Instagram—where Bejar had spent the previous two years working as a consultant—was especially acute. In a subsequent email to Instagram head Adam Mosseri, one statistic stood out: One in eight users under the age of 16 said they had experienced unwanted sexual advances on the platform over the previous seven days.

For Bejar, that finding was hardly a surprise. His daughter and her friends had been receiving unsolicited penis pictures and other forms of harassment on the platform since the age of 14, he wrote, and Meta’s systems generally ignored their reports—or responded by saying that the harassment didn’t violate platform rules.

“I asked her why boys keep doing that,” Bejar wrote to Zuckerberg and his top lieutenants. “She said if the only thing that happens is they get blocked, why wouldn’t they?”

For the well-being of its users, Bejar argued, Meta needed to change course, focusing less on a flawed system of rules-based policing and more on addressing such bad experiences. The company would need to collect data on what upset users and then work to combat the source of it, nudging those who made others uncomfortable to improve their behavior and isolating communities of users who deliberately sought to harm others.

«

His daughter’s experience really is eye-opening: there’s a complete disconnect between how men experience social media, and how women (or girls) do.
unique link to this extract

Apple iPhone battery throttling lawsuit to go ahead in UK • Stackdiary

Alex Ivanovs:

»

A long-running legal battle between technology giant Apple and millions of its UK customers was allowed to continue this week, as a London court ruled that a class action lawsuit against the company can proceed.

The claim, led by Justin Gutmann, alleges that Apple deliberately throttled the performance of older iPhone models through software updates in order to mask issues with aging batteries and push customers to upgrade their devices. The suit covers iPhone 6, 6S, and 7 models released between 2014-2017.

Mr. Gutmann’s claim seeks compensation for up to 25 million UK iPhone users under consumer law. The total value is estimated at around £1.6bn, with the expected range being £853m (approximately $932m).

However, the case is not without issues moving forward. The Tribunal ruled that Gutmann’s claim still lacks sufficient “clarity and specificity”, which will need to be resolved before proceeding to trial. Additionally, Gutmann may need to modify the terms of financing for the lawsuit. This follows a UK Supreme Court ruling that found many third-party litigation funding arrangements are unlawful.

The controversy, nicknamed Batterygate, first came to light in late 2017 when Apple admitted that an iOS update intentionally slowed down iPhones with older batteries. The company claimed this “power management” feature was necessary to prevent unexpected device shutdowns as batteries deteriorated over time.

«

Class actions are comparatively new animals in the UK (ask me, I’m leading part of one) and take quite a long time to put together. By comparison, in the US the class action over exactly the same argument concluded in August, with Apple paying out $310-$500m to affected owners (if they claim it), after an action that began in 2020.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

In Birmingham, trams make the city much more connected than buses do, because they’re more reliable. CC-licensed photo by Mac McCreery 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 outrage (and how to avoid it).

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

Microsoft accused of damaging Guardian’s reputation with AI-generated poll • The Guardian

Dan Milmo:

»

The Guardian has accused Microsoft of damaging its journalistic reputation by publishing an AI-generated poll speculating on the cause of a woman’s death next to an article by the news publisher.

Microsoft’s news aggregation service published the automated poll next to a Guardian story about the death of Lilie James, a 21-year-old water polo coach who was found dead with serious head injuries at a school in Sydney last week.

The poll, created by an AI program, asked: “What do you think is the reason behind the woman’s death?” Readers were then asked to choose from three options: murder, accident or suicide.

Readers reacted angrily to the poll, which has subsequently been taken down – although highly critical reader comments on the deleted survey were still online as of Tuesday morning. A reader said one of the Guardian reporters bylined on the adjacent story, who had nothing to do with the poll, should be sacked. Another wrote: “This has to be the most pathetic, disgusting poll I’ve ever seen.”

The chief executive of the Guardian Media Group, Anna Bateson, outlined her concerns about the AI-generated poll in a letter to Microsoft’s president, Brad Smith. She said the incident was potentially distressing for James’s family and had caused “significant reputational damage” to the organisation as well as damaging the reputation of the journalists who wrote the story.

“This is clearly an inappropriate use of genAI [generative AI] by Microsoft on a potentially distressing public interest story, originally written and published by Guardian journalists,” she wrote. Bateson added that it had demonstrated “the important role that a strong copyright framework plays in enabling publishers to be able to negotiate the terms on which our journalism is used”.

«

The poll is just terrible, and is a perfect example of how the AI Summit (linked below) misses the point. It’s mistakes like this which really show AI undermining the normal operation of things.
unique link to this extract

Sam Bankman-Fried found guilty on all seven counts • TechCrunch

Jacquelyn Melinek:

»

Sam Bankman-Fried, the co-founder and former CEO of crypto exchange FTX and trading firm Alameda Research, has been found guilty on all seven counts related to fraud and money laundering.

The defendant is “charged with a wide-ranging scheme to misappropriate billions of dollars of customer funds deposited with FTX and mislead investors and lenders to FTX and to Alameda Research,” a release from the US attorney’s office at the Southern District of New York stated.

The decision was handed down on Thursday, following a five-week trial that dug deep into how one of the biggest crypto exchanges and its sister trading company, collapsed about a year ago. The US Department of Justice charged 31-year-old Bankman-Fried about 11 months ago.

The jury took about four hours to come to a verdict on six counts relating to fraud and one count relating to money laundering.

Bankman-Fried fell quickly from the top of the crypto totem pole after a faulty Alameda balance sheet was unveiled by CoinDesk in November 2022, which resulted in industry-wide panic and concern around FTX and its liquidity.

«

That’s remarkably quick – both the verdict, and the case coming to trial. Some people will think it was easy, but for the prosecutor, making the case would be tough. However, it certainly helps when the prosecution can persuade various insiders to turn state’s evidence.
unique link to this extract

Sunak plays eager chatshow host as Musk discusses AI and politics • The Guardian

Kiran Stacey:

»

For 25 minutes, the prime minister quizzed Musk on his views on the summit and AI in general.

“What do we need to do to make sure we do enough [to regulate AI]?” asked Sunak. Later he suggested that technology was now developing even faster than “Moore’s law”, which suggests that computing power roughly doubles every two years, before checking himself. “Is that fair?” he asked Musk.

At one point Sunak even appeared to ask the controversial technology entrepreneur his views on international diplomacy. “Some people said I was wrong to invite China [to the summit],” he said. “Should we be engaging with them? Can we trust them? Was that the right thing to have done?”

In case Musk took objection to any of the questions, Sunak made sure to frame them with flattery. “You are known for being such a brilliant innovator and technologist,” he said at one point.

Downing Street had been worried that the famously unpredictable Musk might say something off-colour or undiplomatic. Earlier this week he told Rogan he was worried that environmentalists might harness AI deliberately to eliminate all of mankind. He regularly rails against the “woke mind virus” and has said he bought X to prevent a “zombie apocalypse”.

Officials were so concerned that they changed what had been billed as a live broadcast on X into a recorded conversation, to be edited and put online after the pair had finished speaking.

They need not have worried, however. The blunt and often abrasive Musk familiar to millions of users of his social media platform was replaced by a softly-spoken personification of charm.

«

There’s also five takeaways (not the food kind) from the UK AI summit.
unique link to this extract

AI summit: education will blunt AI risk to jobs, says Rishi Sunak • BBC News

Paul Seddon and Becky Morton:

»

At the summit, hosted at Bletchley Park in Buckinghamshire, several leading technology companies agreed to allow governments to safety-test their next generation of AI models before they are deployed.

The voluntary document was signed by 10 countries and the EU, including the UK, US, Singapore and Canada. China was not a signatory.

In a statement, the UK government said it would work with the Alan Turing Institute, a research body, to assess possible risks such as the potential for bias and misinformation. Mr Sunak said the testing regime would provide some “independent assurance” – adding that the firms developing new models cannot be expected to “make their own homework”.

His government has so far declined to announce legislation to regulate AI, arguing that existing regulators are best placed to mitigate the risks whilst the technology evolves. Mr Sunak told reporters that binding rules would “likely be necessary,” but stressed that the technology was still evolving and it was necessary to ensure it is done in “the right way”.

Before the summit, various unions and campaign groups warned the event would prove a “missed opportunity”. In an open letter, they argued the event should have focused more on topics such as the impact of AI on employment law and smaller businesses, as well as policing and identity profiling.

«

Predictably blah, though some of the people on steering groups are pleased about the “safety tests”. Except China isn’t going to participate. And it’s voluntary. Apart from that, solid job, people.
unique link to this extract

How is Israel using military AI in Gaza strikes? And why won’t it tell us? • Los Angeles Times

Brian Merchant:

»

In response to a request for comment, an IDF spokesperson declined to discuss the country’s military use of AI.

In a year when AI has dominated the headlines around the globe, this element of the conflict has gone curiously under-examined. Given the myriad practical and ethical questions that continue to surround the technology, Israel should be pressed on how it’s deploying AI.

“AI systems are notoriously unreliable and brittle, particularly when placed in situations that are different from their training data,” said Paul Scharre, the vice president of the Center for a New American Security and author of “Four Battlegrounds: Power in the Age of Artificial Intelligence.” Scharre said he was not familiar with the details of the specific system the IDF may be using, but that AI and automation that assisted in targeting cycles probably would be used in scenarios like Israel’s hunt for Hamas personnel and materiel in Gaza. The use of AI on the battlefield is advancing quickly, he said, but carries significant risks.

“Any AI that’s involved in targeting decisions, a major risk is that you strike the wrong target,” Scharre said. ”It could be causing civilian casualties or striking friendly targets and causing fratricide.”

One reason it’s somewhat surprising that we haven’t seen more discussion of Israel’s use of military AI is that the IDF has been touting its investment in and embrace of AI for years.

In 2017, the IDF’s editorial arm proclaimed that “The IDF Sees Artificial Intelligence as the Key to Modern-Day Survival.” In 2018, the IDF boasted that its “machines are outsmarting humans.” In that article, the then-head of Sigma, the branch of the IDF dedicated to researching, developing, and implementing AI, Lt. Col. Nurit Cohen Inger wrote that “Every camera, every tank, and every soldier produces information on a regular basis, seven days a week, 24 hours a day.”

«

One has to love the optimism of asking the IDF how it’s using AI in the midst of its latest conflict. Though one would also point out that it sure didn’t spot Hamas planning its incursion on October 7.
unique link to this extract

Transport Open Data in 2023 • Tom Forth

Tom Forth has been analysing bus and tram open data for various cities since 2019 :

»

On 6 December 2022 I tracked every bus in Great Britain for three hours. This produced a large but manageable 2GB of data. For the same day I downloaded the Great Britain bus timetable in GTFS format.

Focusing just on Leeds and Bristol I matched every bus that ran to its timetable and produced a version of the Great Britain bus timetable in GTFS format reflecting only those buses that ran and their recorded positions every minute.

I loaded the GTFS files representing the real bus movements and timetabled bus movements into Open Trip Planner 2.2 and using the Isochrone feature I calculated the reachable area of Leeds from the Corn Exchange within 45 minutes with only walking and taking the bus as allowed methods of travel.
[Diagram omitted for Overspill inclusion, but really worth looking at.]

We see clearly that in reality the accessibility of Leeds by bus is nowhere near that suggested by the timetable.

As part of this work, we’ve been developing tools for comparing population and public transport networks internationally. I have made a web version of those tools that works just for circles.

Behind the scenes, and not available via that website, we can calculate comparable population estimates for any polygon, including the two accessibility polygons for Leeds. Using this tool we show that the population within 45 minutes of central Leeds by bus on a typical December late afternoon is 445,000 according to the bus timetable – but 165,000 according to the buses that ran and the speed that they ran at.

This is an even larger reduction in effective size than our previous work has shown in Birmingham. We suspect that this is because Leeds has no tram, our work does not currently consider trains, and because congestion was particularly bad on this day due to Christmas shopping and ongoing roadworks.

We found similar results in Bristol though without local knowledge we have chosen not to do any further analysis of what we found.

«

Apparently this work has had some effect inside government – hooray! – though it’s faintly depressing how badly buses serve such major cities. There’s much more in the post, including maps showing the dramatic difference between theory and practice for getting around the cities. His 2019 post on Birmingham will make you think we should put trams everywhere.
unique link to this extract

Scientists create artificial protein capable of degrading microplastics in bottles • Phys.org

»

Every year, around 400 million tons of plastics are produced worldwide, a number that increases by around 4% annually. The emissions resulting from their manufacture are one of the elements contributing to climate change, and their ubiquitous presence in ecosystems leads to serious ecological problems.

One of the most used is PET (polyethylene terephthalate), which is found in many packaging and beverage bottles. Over time, this material wears down into smaller and smaller particles—so-called microplastics—which aggravates environmental problems. PET already accounts for more than 10% of global plastic production and recycling is scarce and inefficient.

Now, scientists from the Barcelona Supercomputing Center—Centro Nacional de Supercomputación (BSC-CNS), together with research groups from the Institute of Catalysis and Petrochemistry of the CSIC (ICP-CSIC) and the Complutense University of Madrid (UCM), have developed artificial proteins capable of degrading PET microplastics and nanoplastics and reducing them to their essential components, which would allow them to be broken down or recycled.

They have used a defense protein from the strawberry anemone (Actinia fragacea), to which they have added the new function after design using computational methods. The results are published in the journal Nature Catalysis.

“What we are doing is something like adding arms to a person,” explains Víctor Guallar, ICREA professor at the BSC and one of the authors of the work. These arms consist of just three amino acids that function as scissors capable of cutting small PET particles. In this case, they have been added to a protein from the anemone Actinia fragacea, which in principle lacks this function and which in nature “functions as a cellular drill, opening pores and acting as a defense mechanism,” explains the researcher.

Machine learning and supercomputers such as the BSC’s MareNostrum 4 used in this protein engineering allow “predicting where the particles are going to join and where we must place the new amino acids so that they can exert their action,” says Guallar. The resulting geometry is quite similar to that of the PETase enzyme from the bacterium Idionella sakaiensis, which is capable of degrading this type of plastic and was discovered in 2016 in a packaging recycling plant in Japan.

«

Avid readers will recall the August 2022 piece on that Japanese bacterium.
unique link to this extract

Elon Musk’s Twitter purchase finally helped these obsessives quit • Slate

Luke Winkie:

»

on the one-year anniversary of Elon Musk’s acquisition and transformation of the company, a broader group of users is clearly feeling a loss. Twitter was a home for everyone: inveterate gamers, breathless BTS stans, transcendent trolls, and—yes—a whole lot of political reporters. Complaints about its usability, functionality, and overall impact on one’s mental well-being long predate the Elon era, and yet, I’ve surprised myself by my ability to be wistful about what we’re all now calling the good old days.

That’s why I reached out to eight people—some archetypal Twitter addicts, others who don’t fit the mold—who’ve left the site in the past year to ask them why they made the decision to board up shop. (I’ve withheld some of their last names at their request.) The reasons for their departures vary. Some find the proximity to someone with Musk’s vindictive, charmless tweets (and they’re hard to escape) to be viscerally unpleasant; others simply dislike how shoddy, gouging, and feature-poor the platform has become in his wake. Some deleted their accounts; others just ignore them now. Their stories, in loose order of departure, present a kind of alternative history of the Musk takeover and a broad view of his digital carnage. It isn’t just bleeding-heart liberals who are disgusted by the state of Twitter. Some users just miss when the site worked.

«

Yes, I’m sure you’ve read a gazillion of these sorts of things, but what’s neat about this roundup is that Winkie finds people who quit at different times, for different reasons: two days after Musk bought it, a few days later when all the banned accounts were let back on, when “verification” started being sold, and so on. The last one is “When the For You Feed Suddenly Felt Like It Was for Someone Else”, which is neat. But the For You feed has always been for someone else – the person who wants to be outraged and infuriated.
unique link to this extract

Apple in 2013: “Android is a massive tracking device” • The Verge

Tom Warren:

»

‘Android is a massive tracking device.’That was the message from Apple in an internal strategy document from 2013. It has been revealed as part of the ongoing US v. Google antitrust trial. The document details Apple’s approach to privacy to differentiate from competitors like Google and Microsoft. Apple later went on to make privacy an even bigger part of its marketing pitch in iPhone commercials in 2019, with the “privacy matters” slogan.

«

The slideshow is no great shakes, to be honest (definitely made in Keynote – good dogfooding, Apple!) but does show how clearly how Apple decided to double down on its strategy credit (the opposite of a strategy tax; something which is boosted by the way you run your business) of not collecting user data if it can avoid it.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified