Periscope lenses have been available on smartphones for some years – but still aren’t common. Will Apple change that? CC-licensed photo by Judith Jackson 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.

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

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

Elon Musk let Russia scare him • The Atlantic

Anne Applebaum:

»

The New Yorker has also described how Ukrainian soldiers abruptly lost their access to Starlink on the battlefield during a different set of land operations. Isaacson’s version of the maritime story implies that all of the drones in the operation washed ashore that evening. But recently in Ukraine, I met some of the engineers who helped design the unmanned sea vehicles, including an engineer who was involved in the first attempt to hit Russian ships in Sebastopol. They told me that not all of the drones involved were lost. Some returned back to base, undamaged.

Here is the part you might not have heard, or not registered: the same team launched a similar attack again a few weeks later. On October 29, a fleet of guided sea drones packed with explosives did reach Sebastopol harbor, using a different communications system. They did hit their targets. They put one Russian frigate, the Admiral Makarov, out of commission. The team believes that they damaged at least one submarine and at least two other boats as well.

And then? Nuclear war did not follow. Despite Musk’s fears, in other words—fears put into his head by the Russian ambassador, or perhaps by Putin himself—World War III did not erupt as a result of this successful attack on a Crimean port. Instead, the Russian naval commanders were spooked by the attack, so much so that they stuck close to Sebastopol harbor over the following weeks.

For their own security, I am choosing not to publish the names of the engineers. I was introduced to them by a tech executive I met on a previous trip to Ukraine, when I was writing about drone operations more broadly. This team has shown off its unmanned boats before, so I am not revealing secrets when I write that they are small, black, and hard to see on the water, and have a very long range—now more than 650 miles, the engineers told me.

«

Of course the Russians who spoke to Musk led him up the garden path about what they would do. Because he’s an idiot and/or easily flattered/scared.
unique link to this extract

2023 on track to be world’s hottest year on record, at 1.5ºC above pre-industrial • ABC News

Tom Saunders:

»

This year is now almost certain to become Earth’s warmest on record after a hot July and August saw global temperatures reach the Paris Agreement target of 1.5ºC above pre-industrial levels.

This is the first time the 1.5ºC threshold has been passed for more than one month, and only the second time it has ever been exceeded, behind February 2016.

Data released last week from Copernicus, a branch of the European Union Space Programme, shows August was 1.59ºC warmer than 1850-1900 levels, following a 1.6ºC increase in July.

The recent records have now lifted the year-to-date global temperature to the end of August to 1.35ºC above pre-industrial levels, just 0.01ºC behind 2016 — the current record holder, according to Copernicus data.

While the latter months of 2016 cooled off, global temperatures this year have been building, a trend likely to continue as El Niño develops further.

…The world’s waters have been so warm in 2023 that seven out of a possible eight months have entered the top 30 warmest on record, including four of the top five.

«

A marker of failure, and inertia – in the form of resistance to change. But now the planet is changing, faster and faster. Wonder if some of the American politicians will argue that since we’re now already at 1.5ºC there’s no point trying to ameliorate it.
unique link to this extract

Huawei watchers and the land of magical thinking • Digits to Dollars

Jonathan Goldberg:

»

This weekend, the big news in certain circles was the news about Huawei’s new Mate 60 flagship smartphone and the seemingly miraculous Kirin 9000s chip that powers it. This chap has riveted everyone’s attention as it seems to raise questions about the efficacy of the US government’s restrictions on Huawei and China’s access to advanced manufacturing processes.

…We know that Huawei has been investing heavily in fabs in recent years. We saw signs of that in their investments in third parties, but it is also reasonable to assume they spent a lot of time and money working with SMIC [China’s chip fab] engineers, efforts that would not show up publicly. Critically, we do not know what the yield of this chip is and whether it is remotely profitable for SMIC. SMIC has been held to DUV machinery, cut off from ASML’s EUV machines by US actions. Their 7nm process is probably the limit of what they can produce without access to EUV. And let’s not forget that labels like “7nm” are marketing terms. In terms of actual comparisons what matters most is transistor density, and on this metric this source claims SMIC is about 10% behind what TSMC labels 7nm. There is a non-zero chance that SMIC is actually struggling to build these parts and is losing a lot of money in the process. (Again, we may never see this in their public accounts, as someone, somewhere is likely subsidizing this work.)

Most people assume that SMIC is pushing the boundaries of what DUV can accomplish. Unless Huawei and SMIC have made some incredible, secret breakthrough, this means there is no future for this process. Maybe they can squeeze out a bit more density for one more round, but beyond that they would be breaking the laws of physics. So while the Kirin 9000 is a real achievement, we are highly skeptical that this really changes anything.

«

As Goldberg points out, those screaming about this chip showing the sanctions haven’t worked are essentially ignoring the reality, because doing so suits their own agenda.
unique link to this extract

Twitter appears to throttle New York Times • Semafor

Max Tani:

»

X, Elon Musk’s social media platform formerly known as Twitter, appears to be attempting to limit its users’ access to The New York Times.

Since late July, engagement on X posts linking to the New York Times has dropped dramatically. The drop in shares and other engagement on tweets with Times links is abrupt, and is not reflected in links to similar news organizations including CNN, the Washington Post, and the BBC, according to NewsWhip’s data on 300,000 influential users of X.

The drop in engagement in Times posts seems isolated to X: NewsWhip data showed that engagement with Times links shared on Facebook remained consistent relative to other outlets.

“There was a drop off in engagement for NYT compared to the other sites in late July/early August,” NewsWhip spokesperson Benedict Nicholson told Semafor.

Times employees had already taken note of the pattern, as high-profile attempts to share Times articles failed to travel on the platform. For instance, earlier this week, former President Barack Obama shared multiple New York Times articles on X about healthcare costs, which the service said reached fewer than 900,000 and 800,000 users respectively. The number was far lower than any other post shared by the former president since X began sharing that data publicly earlier this year — for comparison, a Politico link shared by the president got nearly 13 million views.

«

Hanlon’s Law may apply, but Musk makes everything worse.
unique link to this extract

Proust, ChatGPT and the case of the forgotten quote • The Guardian

Elif Batuman:

»

The other day, I was looking for a quote in Proust, so I thought I would ask ChatGPT. Here’s how it went.

EB: Is there a passage in In Search of Lost Time when Proust talks about how love affairs repeat themselves, and says that when you’re writing about a love affair in the past, you’re often drawing inspiration from a different love affair in the present?

CHATGPT: Yes, there is a passage in Marcel Proust’s “In Search of Lost Time” (also known as “Remembrance of Things Past”) where he discusses the repetition of love affairs and the influence of present experiences on writing about past love affairs. The passage can be found in the first volume of the novel, “Swann’s Way” (Du côté de chez Swann).

In this passage, Proust reflects on how our present emotions and experiences often color our memories of the past. He suggests that when we write about a past love affair, we might unknowingly be influenced by the emotions and experiences of a current or recent love affair. This blurring of past and present emotions can result in the repetition of themes and feelings in our writing.

The idea of how love affairs repeat themselves and how we are influenced by our present experiences when writing about the past is a recurring theme in Proust’s work, as he delves into the complexities of memory, perception, and the nature of human emotions.

Please note that the specific wording and context of the passage may vary depending on the translation and edition of the novel. If you’re interested in exploring this theme further, I recommend reading “Swann’s Way” to experience Proust’s beautiful prose and intricate exploration of love, memory, and the passage of time.

«

There’s a lovely meta-joke in Batuman’s efforts to pin down something “remembered” by a machine about a piece of work that is all about remembering.
unique link to this extract

What does it cost to try to stay young forever? Inside the hyped-up world of anti-ageing • The Globe and Mail

Alex Cyr:

»

Toronto’s Longevity House, a 9,000 square-foot facility that opened in 2021 and charges $100,000 for a lifetime membership, offers clients access to a suite of services such as red-light therapy, chakra therapy and electrical muscle-stimulation suits.

In the United States, startup Ambrosia started offering infusions of blood from teenagers and young adults in 2016, charging US$8,000 a litre, only to be shut down that same year after being dubbed unsafe by the FDA.

The dearth of robust scientific studies on these treatments is making some experts question their effectiveness and safety. Brett Finlay, a B.C.-based microbiologist and co-author of The Whole-Body Microbiome, says he wants to wait for results from clinical trials before believing in any trendy anti-ageing method. The problem, he said, is the length of time it takes to collect such data. One longitudinal ageing study from the Netherlands that set out to study the change in physical, emotional, cognitive and social functioning of people as they age has been continuing since 1992.

In the absence of data, Finlay fears that internet movements such as taking steroids for muscle growth or applying snake venom to wrinkles are motivating demand for quick fixes, which are clouding tried and true healthy practices.

“The truth is that we already have anti-ageing methods, but they are incredibly boring and hard: It is exercise, eating a balanced diet with lots of fibre, good sleep and cultivating a good community of people around you,” said Finlay, adding that these lifestyle practices have been shown to reduce stress and inflammation, which are correlated with some of the most prevalent causes of death such as obesity-caused illnesses, diabetes, cancer, dementia and Alzheimer’s disease.

“Trends and fads come and go, whereas science just plods along and slowly gives us answers.”

«

Apparently Peter Thiel wants to be cryogenically frozen when he dies. I say, why wait? Anyhow, anti-ageing is reckoned to be a $63bn industry worldwide. Or you could follow Finlay’s rather cheaper advice.
unique link to this extract

China sows disinformation about Hawaii fires using new techniques • The New York Times

David Sanger and Steven Lee Myers:

»

When wildfires swept across Maui last month with destructive fury, China’s increasingly resourceful information warriors pounced.

The disaster was not natural, they said in a flurry of false posts that spread across the internet, but was the result of a secret “weather weapon” being tested by the United States. To bolster the plausibility, the posts carried photographs that appeared to have been generated by artificial intelligence programs, making them among the first to use these new tools to bolster the aura of authenticity of a disinformation campaign.

For China — which largely stood on the sidelines of the 2016 and 2020 U.S. presidential elections while Russia ran hacking operations and disinformation campaigns — the effort to cast the wildfires as a deliberate act by American intelligence agencies and the military was a rapid change of tactics.

Until now, China’s influence campaigns have been focused on amplifying propaganda defending its policies on Taiwan and other subjects. The most recent effort, revealed by researchers from Microsoft and a range of other organizations, suggests that Beijing is making more direct attempts to sow discord in the United States.

«

Guess they’ll have to duke it out with the Russians trying to do the same. Also, why is China bothering with stuff that Marjorie Taylor Greene does for free using the interstellar gas in her head?
unique link to this extract

August 2020: What is a periscope lens for smartphone cameras? • How To Geek

Harry Guinness in August 2020:

»

There just isn’t enough room to add a 12mm lens to a phone that’s only 8mm thick. Unless you do it sideways.

A periscope lens works much like a periscope on a submarine. Light enters the front element and is then reflected 90 degrees by an angled mirror. It passes through any other lens elements before hitting the camera sensor and is then recorded as a photo. By changing the direction in which the light travels, longer lenses don’t have to be as deep because they can be wide.

For phone manufacturers, this is a serious advantage. It’s much more practical to find the necessary space for a longer telephoto lens horizontally than it is to shrink the sensor or make a thicker phone.

This way, manufacturers aren’t limited to 50mm-equivalent lenses with 2x optical zoom (or, at a push and with some dubious marketing, 3x). It makes 100mm- (around 5x zoom) or even 200mm-equivalent (around 10x zoom) lenses possible.

…Chinese manufacturers (Oppo and Huawei, in particular) have been playing around with them for several years. The five-camera Huawei P40 Pro+ has a 10x periscope telephoto lens that’s equivalent to a 240mm lens on a full-frame camera.

The more widely available Samsung Galaxy S20 Ultra has a 5x telephoto periscope lens that’s roughly equivalent to a 100mm.

«

So Apple’s three years late to this particular party, and is believed to be reserving this only for the top-end phone. However this article from May 2023 shows only seven phones you can buy, from five manufacturers, which incorporate periscope cameras. Maybe not that common, then.
unique link to this extract

Qualcomm to supply Apple with 5G chips until 2026 under new deal • Reuters via MSN

Stephen Nellis:

»

Qualcomm said on Monday it had signed a deal with Apple to supply 5G chips until at least 2026, at a time when the iPhone maker faces increased challenges in China and looks to reinforce its supply chains elsewhere.

The deal extends a relationship worth billions of dollars to Qualcomm for at least three years beyond what was expected and indicates Apple is not rushing out its own modem, despite moving all its computers to processing chips of its own design.

Qualcomm shares were up 4% in early afternoon trade. The company is the leading designer of modem chips that connect phones to mobile data networks. Apple shares rose 0.5%.

San Diego, California-based Qualcomm previously signed a chip supply deal with Apple in 2019, after the two companies settled a protracted legal battle.

That supply agreement ends this year, meaning that the iPhones that Apple is expected to announce on Tuesday would be the last phone debut under that deal.

«

Well this rather puts the kibosh on those stories from a couple of years ago, such as this one in November 2021 that “At today’s Investor Day event, Qualcomm CFO Akash Palkhiwala said that Qualcomm expects to supply just 20% of Apple’s modem chips in 2023.” Clearly, Qualcomm and Apple have patched things up (or that the 2019 purchase of most of Intel’s modem division for $1bn has completely fizzled).
unique link to this extract

Billpayers could miss out on £1bn a year in savings due to wind auction error • Energy & Climate Intelligence Unit

George Smeeton:

»

Offshore wind costs reached a record low strike price of around £50 per megawatt-hour (in current prices) in the Contracts for Difference (CfD) Allocation Round 4 (AR4) auction held in 2022. Despite subsequent project cost inflation due to the post-pandemic economic recovery, the gas crisis, and rising interest rates, offshore wind remains around a third cheaper than forecast wholesale prices set by gas of around £100/MWh for the middle of this decade.

However, the Administrative Strike Price (ASP) that caps CfD auction bids was not adjusted to reflect this inflation for the AR5 CfD auction held this year. By contrast, the Irish Government allowed for supply chain inflation in its recent auction, which has successfully secured 3GW of offshore wind capacity at competitive prices.

As a result, out of the UK’s 5GW of ‘shovel ready’ offshore wind projects that could be delivered in the mid-2020s, it appears that none have bid into AR5 and all could now face delays. Had these projects been able to secure contracts, even with higher strike prices they would have paid back £1bn per year in the mid-2020s, a saving that could now be delayed.

Commenting on the results, Jess Ralston, Energy Analyst at ECIU, said:

“The key point here is that even with inflation, offshore wind is still about a third cheaper than gas power stations with the price of gas set to remain higher than before the crisis.

“The more renewables, the less gas you have to buy. By failing to back offshore wind, the Government has added around £1 billion a year to energy bills in coming years.”

«

Easy to forget that it’s not a forgone conclusion that you’ll get the cheaper energy built; governments have to take the right decisions too. The UK government didn’t, here.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The user interface of streaming services is remarkably bad. But why? CC-licensed photo by Michael Sheehan 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 was another post last Friday at the Social Warming Substack: it’s about LinkedIn. Free signup.

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

Experts fear crooks are cracking keys stolen in LastPass breach • Krebs on Security

Brian Krebs:

»

In November 2022, the password manager service LastPass disclosed a breach in which hackers stole password vaults containing both encrypted and plaintext data for more than 25 million users. Since then, a steady trickle of six-figure cryptocurrency heists targeting security-conscious people throughout the tech industry has led some security experts to conclude that crooks likely have succeeded at cracking open some of the stolen LastPass vaults.

Taylor Monahan is lead product manager of MetaMask, a popular software cryptocurrency wallet used to interact with the Ethereum blockchain. Since late December 2022, Monahan and other researchers have identified a highly reliable set of clues that they say connect recent thefts targeting more than 150 people, Collectively, these individuals have been robbed of more than $35 million worth of crypto.

Monahan said virtually all of the victims she has assisted were longtime cryptocurrency investors, and security-minded individuals. Importantly, none appeared to have suffered the sorts of attacks that typically preface a high-dollar crypto heist, such as the compromise of one’s email and/or mobile phone accounts.

“The victim profile remains the most striking thing,” Monahan wrote. “They truly all are reasonably secure. They are also deeply integrated into this ecosystem, [including] employees of reputable crypto orgs, VCs [venture capitalists], people who built DeFi protocols, deploy contracts, run full nodes.”

Monahan has been documenting the crypto thefts via Twitter/X since March 2023, frequently expressing frustration in the search for a common cause among the victims. Then on Aug. 28, Monahan said she’d concluded that the common thread among nearly every victim was that they’d previously used LastPass to store their “seed phrase,” the private key needed to unlock access to their cryptocurrency investments.

«

unique link to this extract

From Netflix to HBO, the terrible design of streaming is ruining TV •

Jesus Diaz:

»

Each platform has its own flavor of bad. On Apple TV+, for example, the “continue watching” menu is hidden under the fold. Good luck watching the credits on Netflix—they automatically skip to serve up a new episode. Hulu makes viewers hunt for the very show they were just watching. And Amazon Prime has turned its streaming service into an all-you-can-eat buffet of video content that requires watchers to surf its interface as if they were looking for the best deal on toilet paper.

Across the board, the streamers have terrible landing pages and subpar curation algorithms. Playback buttons routinely fail to work properly, and using the time slide to scrub to the right point in the video still feels as fun and precise as playing Tetris with your toes.

In a silo, these UX sins are an annoying, but ultimately ignorable, trade-off for easy-to-access content. But it’s 2023, and really, there’s no excuse for an entire genre of digital tools to so blatantly ignore the basics of good design. I had to wonder: Why do the platforms that are supposed to bring us pleasure seem passionately invested in planting anti-UX mines all over their interfaces? I asked a few experts in the field for their take.

The most glaring issue with the streaming interface is how hard they make it to keep watching a show. On AppleTV+ and Disney+, in particular, finding the “continue watching” feature requires a long scroll. “As UX designers and users of these services, we find this extremely annoying,” Carsten Wierwille tells me over an email interview.

Wierwille, CEO of digital design studio Ustwo, blames the entertainment industry’s fear of the new. Instead of reinventing the TV experience, as streaming promised to do, the platforms have copied some of cable’s worst traits in a quest to hook viewers and maintain subscriptions. Wierwille ventures that the buried “continue watching” menu is likely the victim of marketing objectives. Streamers reserve the prime screen real estate to promote new or popular content, which he says fits with traditional TV design patterns of prioritizing content based on who pays the most for “preferred placement” rather than prioritizing the user needs.

«

It’s true: their design is terrible. (John Siracusa made this point, less prominently but no less well, in January 2022.)
unique link to this extract

Nobody will tell you the ugly reason Apple acquired a classical music label • The Honest Broker

Ted Gioia:

»

If Apple wanted to offer exclusive music to subscribers it wouldn’t buy a label that records so many works in the public domain. This is the first warning sign.

My favorite offerings on the BIS label are the Bach cantatas recorded by Bach Collegium Japan under Masaaki Suzuki. (I recommended them last year in my article about the cantatas.)

The BIS label has also released the full Beethoven symphony cycle—performed by the Minnesota Orchestra under Osmo Vänskä, as well as lots of Mozart, Tchaikovsky, and other core contributors to the classical repertoires. However, unlike the cantatas, these other BIS albums are seldom the most esteemed versions on the market.

But how many music fans searching for Beethoven or Mozart on streaming are picky about conductors and orchestras? If the first search results are the Minnesota Orchestra are they really going to dig deeper to find the Berlin Philharmonic?

Of course, some unique offerings can be found on BIS—especially of obscure Nordic composers. But do you really believe that Apple made this acquisition in order to corner the market on Kalevi Aho or Geirr Tveitt?

The very idea is ridiculous.

«

Gioia’s argument boils down to: Apple wants to be able to increase its profits (or reduce its losses) on Apple Music by offering music that is less encumbered by rights payments. I don’t find this persuasive: if that were the case, it would just buy a ton of music companies (which it could do, as Gioia admits, without breaking stride, apart of course from the antitrust challenge).

I think Apple is buying this classical music company precisely because people who like classical music are, indeed, picky about which performance, which conductor, which orchestra. I don’t want the Portsmouth Sinfonia’s version of Beethoven No.5; I want Karajan. (Though you may enjoy the contrast between the content on those two links.)
unique link to this extract

The untold story of Elon Musk’s support for Ukraine • The Washington Post

Walter Isaacson, Elon Musk’s biographer, detailing what happened over the derailed attack on Russian ships:

»

My phone started vibrating with messages from Musk.

“This could be a giant disaster,” he texted. I went behind the bleachers to ask him what the problem was. He was in full Muskian crisis-hero-drama mode, this time understandably. A dangerous issue had arisen, and he believed there was “a non-trivial possibility,” as he put it, that it could lead to a nuclear war — with Starlink partly responsible. The Ukrainian military was attempting a sneak attack on the Russian naval fleet based at Sevastopol in Crimea by sending six small drone submarines packed with explosives, and it was using Starlink to guide them to the target.

Although he had readily supported Ukraine, he believed it was reckless for Ukraine to launch an attack on Crimea, which Russia had annexed in 2014. He had just spoken to the Russian ambassador to the United States. (In later conversations with a few other people, he seemed to imply that he had spoken directly to President Vladimir Putin, but to me he said his communications had gone through the ambassador.) The ambassador had explicitly told him that a Ukrainian attack on Crimea would lead to a nuclear response. Musk explained to me in great detail, as I stood behind the bleachers, the Russian laws and doctrines that decreed such a response.

What the Ukrainians did not know was that Musk decided not to enable Starlink coverage of the Crimean coast. When the Ukrainian military learned that Starlink would not allow a successful attack, Musk got frantic calls and texts asking him to turn the coverage on. Fedorov, the deputy prime minister who had originally enlisted his help, secretly shared with him the details of how the drone subs were crucial to their fight for freedom.

«

So it wasn’t Musk alone who decided; Russia persuaded him. Showing precisely why you shouldn’t put these decisions in the hands of someone who doesn’t understand politics. “Russian laws and doctrines”, huh.
unique link to this extract

The value of a customer • Asymco

Horace Dediu:

»

the picture becomes clearer. The iPhone customer is 7.4 times more valuable than the Android customer. This is more impressive than the 4x rule I had 10 years ago. The reasons are mainly that my anecdotes were from developers who sold products in the US or EU whereas expansion of smartphones to 7 billion global users has drawn in more lower spending customers.

But Apple’s base has also grown to over 1 billion users (650 million store users). This highlights that Apple has effectively grown and discriminated customers effectively. It obtained not just 1 billion customers but the best 1 billion customers.

How to discriminate effectively is the holy grail of marketing. The naïve approach is to keep prices high. But that usually only results in a “luxury” branding and a small base that tends not to grow. The alternative “premium” approach is to offer functionality and multiple tiers and distribution options and financing and merchandising. There is no simple formula.

The bottom line is that Apple’s approach is attracting 650 million $10/month app spenders. When we factor in additional subscription services, we get to the juggernaut that is Apple Services. This analysis has shown how difficult it is for anyone to come close to this quality of revenue.

As we look forward to Spatial Computing, the idea of increasing that spend from $10/month for a small glass rectangle in your palm to perhaps $100/month for an immersive 360-degree 3D experience does not sound too crazy.

«

unique link to this extract

Building Meta’s Threads app (real-world engineering challenges) • The Pragmatic Engineer

Gergely Orosz got insights from the team that built Threads:

»

What was a difficult challenge which arose?

Making an engaging feed for Threads was the biggest one. On a microblogging site, the thing that is most relevant is what is happening right now. It needs to capture what everyone is talking about, and present choices for which conversations to dive into. 

There’s a balance between immediacy and helping the user find content from someone they’re most likely to engage with. At the same time, other apps in the space have proved you don’t need an extensive graph of connections in order to serve relevant and interesting content. 

We’ve talked on the Threads team about how we’d like to reduce our reliance on users manually curating follow graphs. However, to reduce reliance on manual curation, you need to understand what these posts are about, and know what is going on in the world – which are both challenging, to say the least!

«

There’s plenty more very insider-y stuff which will fascinate anyone who has had to run a skunkworks-style project inside a bigger organisation.
unique link to this extract

iPhone 15 Pro Max with periscope lens, anticipated to capture nearly 40% of new iPhone production • Trendforce

»

In regard to specifications for the iPhone 15 series, several noteworthy hardware upgrades have been made. Compliance with EU regulations has led Apple to jump on the USB Type-C bandwagon this year. The iPhone 15 and iPhone 15 Plus will come with significant camera upgrades, sporting a 48MP main sensor to align with the Pro series. Furthermore, they will also be featuring Apple’s Dynamic Island. On the other hand, the Pro series promises cutting-edge processor upgrades, increased Dram capacity, and introduces a titanium-aluminum alloy frame. The Pro Max also intends to elevate mobile photography to the next level with its exclusive periscope lens.

Advances in technology, while exciting, can also ratchet up the intricacies of mass production. Reports of component snags and assembly issues have surfaced as production of the new iPhone models revs up in the third quarter. The iPhone 15 and iPhone 15 Plus, in particular, have been grappling with lower-than-expected yield rates for their new 48MP cameras. Meanwhile, the Pro series is confronting challenges with panel and titanium alloy frame assembly. However, evidence suggests that the Pro series is likely to overcome its obstacles more swiftly than its non-Pro counterparts.

«

Trendforce has a helpful graphic that tells you everything about the models forecast to be unveiled on Tuesday. Not sure there’s anything else you need Tim Apple or the other folk to tell you apart from the price. (I think the “periscope” lens means it actually moves out from the back towards the object, rather than popping up from the top, but we await that clarification.)

unique link to this extract

Washington, D.C.’s secret carpool cabal is a daily slug fest • Car and Driver

Elana Scherr:

»

Every city has its secrets. Washington, D.C., may have more than most, but I wasn’t there to dig up bodies, corporeal or political. My interest in visiting our nation’s capital was to find out more about a covert society, an organization of carpoolers who use codes and word of mouth to work around D.C.’s notorious traffic jams and exorbitant tolls. Under cherry blossoms light as dreams and in the long shadow of the Washington Monument, I set out in search of slugs.

The origins of slugging are murky, buried somewhere in the smoggy ’70s, when the Shirley Highway (I-395) from Virginia to D.C. became the first U.S. freeway to implement a high-occupancy vehicle (HOV) lane. Those early carpool lanes were strict, initially requiring four occupants (now three), so commuting drivers would fill their seats by swinging by the bus stop and sniping riders. Eventually, the bus lines had more hopeful carpoolers than mass-transit riders, and the bus drivers began referring to the faux passengers as false coins, also known as slugs.

Undeterred by the slander, the slugs claimed the nickname and the practice grew in popularity, developing set locations and traditions. When the Virginia Department of Transportation partnered with the private toll-road operator Transurban in the early 2000s, multiperson vehicles were exempted from the expensive fees to encourage carpooling, thus keeping the fast lanes flowing free. Slugging is symbiotic—no money changes hands, but all parties benefit.

«

It’s maybe not tech (though it is fuel efficiency!) but it’s fascinating. (Thanks Paul C for the tip.)
unique link to this extract

Low-nicotine cigarette maker 22nd Century pursues strategic alternative amid financial struggles • Winston-Salem Journal

Richard Craver:

»

In May, 22nd Century provided its first annual fiscal revenue projections as a publicly-traded company at a range of $105m to $110m for 2023. That would represent a 69% to 77% increase from $62.1m in fiscal 2022.

However, to put that revenue range into perspective, the U.S. traditional cigarette marketplace has about $60bn in annual sales, according to a Goldman Sachs analysis.

Other anti-smoking advocates say an FDA emphasis on very-low nicotine cigarettes could steer tobacco consumers toward potentially less harmful products, such as electronic cigarettes, heat-not-burn cigarettes and moist snuff.

Anti-smoking advocates say that if the FDA is successful in mandating very-low nicotine cigarettes, some smokers may go to a black market to buy those made outside the U.S. with current nicotine levels.

Smokers could decide to consume more very-low-nicotine cigarettes in order to gain the same nicotine levels as they are accustomed to now.

“Their basic strategy is fundamentally flawed — tobacco use is a widely and correctly understood as a nicotine-seeking behavior,” said Clive Bates of Counterfactual, a London-based public health and sustainability consultancy.

“So, what does a product with negligible nicotine, but all the toxicity of a conventional cigarette, offer the consumer?”

«

22nd Century, which got loads of backing from the US FDA, is essentially heading for bankruptcy. There’s an amazing thread by the American Vapor Manufacturers, who make vaping products, pointing out all the positive coverage 22nd Century got, and yet now barely a peep.

I don’t hold any torch for vaping, but it’s an enormously safer (and more sociable) way of consuming nicotine than burning tobacco.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

What exactly caused the Baby Boom in the middle of the 20th century? CC-licensed photo by Erin Warren on Flickr.

There’s another post due at the Social Warming Substack at about 0845 UK time.

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

UK data watchdog to review period and fertility apps amid security concerns • The Guardian

Hibaq Farah:

»

The UK data watchdog is to review period and fertility tracking apps after users expressed concerns over data security.

The Information Commissioner’s Office (ICO) announced the move after it commissioned research showing that half of women have concerns about how their data is used by the apps.

The poll of more than 1,150 women showed that a third of women have used apps to track their periods or fertility. Women said transparency over how their data was used and how secure it was were bigger concerns – at 59% and 57% of respondents respectively – than cost and ease of use when it came to choosing an app.

More than half of the people who use the apps also noticed an increase in baby or fertility-related adverts after signing up, with 17% finding the adverts distressing.

Emily Keaney, deputy commissioner of regulatory policy at the ICO, said the review intended to establish “both the good and the bad” of how the apps work.

She said: “These statistics suggest data security is a significant concern for women when it comes to choosing an app to track their periods or plan or prevent pregnancy. That’s not surprising, given the incredibly sensitive and personal information involved. Once we have more information, we will explore next steps, but we will not hesitate to take regulatory action to protect the public if necessary.”

The regulator will look into identifying potential harms and negative impacts on users. This includes confusing privacy policies that do not allow users to easily understand what they have consented to, whether the apps request and store unnecessary volumes of data and if users are receiving upsetting targeting advertising that they did not sign up to.

«

Might also look at the app frameworks, since developers typically include third-party code that will do all sorts of intrusive things.
unique link to this extract

‘Ritual mass murder’ report in Chapel St Leonards was yoga class • BBC News

Pritti Mistry:

»

A yoga class was mistaken for a “ritual mass murder” scene after members of the public saw several people lying on the floor and reported it to police.

Five police cars descended on the North Sea Observatory in Chapel St Leonards, Lincolnshire, on Wednesday night.

Yoga teacher Millie Laws said she thought reports of her being a “mass murderer” were a “joke at first”.

Lincolnshire Police confirmed everyone was safe and well, and the call was made with “good intentions”.

The 22-year-old teacher said she was teaching seven students at the Seascape Cafe, which is inside the building, when she saw two dog walkers peering closely through the glass window during the Shavasana or relaxation stage of the class.

“They’re [students] laying down with blankets over them, their eyes are closed. It’s very dark in there. I just had candles and little tea lights lit the whole room, and I was just walking around playing my drum. I had a nice floaty top on with large bell sleeves,” she said. “A couple with some dogs just came up to the window and were having a look in, but they walked off really quickly and I didn’t think anything of it.”

Laws continued: “I didn’t know until after we left that these people phoned in saying that there was a mass murderer; they were wearing a robe and they were walking over all of the people, and it looked like some kind of ritual, and that the people on the floor were actually dead. I guess from the outside view it could look like that, because they’re all really still, very nice and relaxed.

«

What. A. Story. Dog walkers do all the hard work of finding dead bodies: have you ever noticed how many “missing body found” reports stem from people walking their dog(s)? Now including when the bodies are not yet dead, apparently.
unique link to this extract

ChatGPT fails in languages like Tamil and Bengali • Rest of World

Andrew Deck:

»

Less than a year after launch, ChatGPT is being used all over the world, generating Amazon listings in China and call center scripts in the Philippines. But while ChatGPT thrives in English, Spanish, Japanese, and other dominant languages, it struggles to produce the same quality of text in languages like Bengali, Swahili, Urdu, and Thai — languages that have millions of speakers, but appear much less often online.

When Rest of World tested ChatGPT’s ability to respond in underrepresented languages, we found problems reaching far beyond translation errors, including fabricated words, illogical answers and, in some cases, complete nonsense.

Take Tigrinya, a language which has over 7 million speakers, with the vast majority located in Eritrea and the northern part of Ethiopia. Tigrinya shares a similar script to Amharic, a more dominant Ethiopian language, but there are significant differences between the two. When asked to list examples of African countries, ChatGPT mixed up Tigrinya and Amharic, adding characters that don’t exist in Tigrinya. It created an output to this simple question that is challenging to read for native speakers of both languages.

… Much has been made of the tendency of AI chatbots to “hallucinate” — shorthand for fabrications that chatbots state as facts. This problem is common with ChatGPT responses in low-resource languages. But in multiple instances, rather than generating fake numbers or other facts, Rest of World found that ChatGPT simply makes up words.

«

This is one of those obvious-when-you-consider-it things, but that doesn’t make it any less bad. What it means is that the gap in usability of AI will grow wider between English and other languages.
unique link to this extract

Apple stock hit by China worries ahead of iPhone 15 launch • Financial Times

Patrick McGee:

»

Investor alarm that China may be cracking down on officials’ iPhone use has knocked $200bn off Apple’s market value, casting a shadow over next week’s launch of its latest smartphone.

Beijing’s reported curbs on government iPhones, alongside a resurgent Huawei, threaten to derail what should have been a moment of triumph for Apple: unseating Samsung at the top of the smartphone market.

Before the China turmoil, analysts had predicted that the launch of the iPhone 15 would put Apple within reach of becoming the world’s biggest smartphone maker by volume for the first time.

However, Apple’s shares have fallen by about 6% over the course of the past two days, as investors fretted about its fate in China, which makes up roughly a fifth of its revenue.

“A decade ago it seemed inconceivable that Apple could wrestle the top spot from Samsung but it could be that we’re on the cusp of that milestone,” said Ben Wood, analyst at CCS Insight. “They will be tantalisingly close but the Chinese market will play a pivotal role.”

Several reports surfaced this week suggesting that Beijing has ordered public officials in certain government departments not to use iPhones or other foreign devices for work.

Bank of America has estimated that China accounts for up to 50m iPhone sales annually and that such a ban could cost Apple 5-10m units a year.

«

Nothing official from the Chinese government, of course.
unique link to this extract

Understanding the Baby Boom • Works in Progress

Anvar Sarygulov & Phoebe Arslanagic-Wakefield:

»

[In the 1930s] Despite the organised resistance of groups like Alliance nationale, at least some European demographers doubted whether falling birth rates were truly reversible, or even arrestable. In 1936 Dr Carr Saunders, an English biologist, eugenicist, and later Director of LSE, wrote: 

»

once the small voluntary family habit has gained a foothold, the size of the family is likely, if not certain, in time to become so small that the reproduction rate will fall below replacement rate, and that, when this happened, the restoration of a replacement rate proves to be an exceedingly difficult and obstinate problem.

«

But even as Carr Saunders wrote those words, he was being proved wrong. Something was happening, in Europe and farther afield. Something we are still trying to understand today: the Baby Boom. 

The Baby Boom was an unexpected change in direction from the century of falling fertility that had taken place in Europe and North America. Contrary to the popular belief that it was triggered by soldiers returning home from World War Two, the Boom in fact began in the mid-1930s. It was not simply an American or British phenomenon either. The demographic wave swept over Iceland, Poland, Sweden, France, the Netherlands, Austria, the Czech Republic, Canada, Norway, Switzerland, and Finland. Thousands of miles across the sea, it even happened in Australia and New Zealand. 

And yet Carr Saunders’ doubt had been extremely well justified. 

For wherever and whenever we have had data and since the Industrial Revolution – with the crucial exception of the Baby Boom – it has been a nearly iron law of fertility that higher incomes are associated with lower birth rates. A key mechanism for this is likely to be that rising living standards effectively increase the costs of having a child, from lost wages for working women to reduced time for increasingly accessible leisure activities.

«

But that doesn’t explain the Baby Boom. This analysis does point to the multiple elements that probably do – with the Amish acting as a control group.
unique link to this extract

Buzzy AI startup for generating 3D models used cheap human labour • 404 Media

Joseph Cox and Jason Koebler:

»

An artificial intelligence company, whose founder Forbes included in a 30 Under 30 list recently, promises to use machine learning to convert clients’ 2D illustrations into 3D models. In reality the company, called Kaedim, uses human artists for “quality control.” According to two sources with knowledge of the process interviewed by 404 Media, at one point, Kaedim often used human artists to make the models. One of the sources said workers at one point produced the 3D design wholecloth themselves without the help of machine learning at all.

The news pulls back the curtain on a hyped startup and is an example of how AI companies can sometimes overstate the capabilities of their technology. Like other AI startups, Kaedim wants to use AI to do tedious labor that is currently being done by humans. In this case, 3D modeling, a time-consuming job that video game companies are already outsourcing to studios in countries like China.

…After the publication of this article, Kaedim rebranded its website to make it clearer that humans are involved in the company’s production of 3D images.

«

Smart way to get that sweet, sweet VC money.
unique link to this extract

With SBF, Gisele, and Michael Lewis at the peak of the crypto craze

Zeke Faux:

»

when we sat down, Wu and a colleague bragged to me about a Zelda-like play-to-earn game on their blockchain that had attracted 40,000 users in less than a month. They said the game was teaching people about DeFi — decentralized finance, a way of trading without a central counterparty — and letting them earn high returns. It sounded a lot like Axie Infinity. I couldn’t believe they were pitching it with a straight face after Axie’s collapse. “You can get 10% in DeFi,” Wu said. “You can be a true freelancer. There are literally people who quit their jobs. It’s not magic. If you know what you’re doing here, you’re going to change your life.”

Michael Wagner, the founder of a space-themed crypto game called Star Atlas, even cited Axie Infinity as a proof of concept. Instead of colorful blobs and Smooth Love Potions, Star Atlas players had to buy spaceship NFTs to earn atlas tokens, and he told me he’d already sold nearly $200 million worth of them. But when I asked if I could try out the game, he said it didn’t exist yet. Even though he’d already sold the spaceships, he said it would be at least five years before the game was ready. “It’s very early stage,” he said. “We believe the game could bring in billions of users.”

Another crypto executive showed me a digital image of a sneaker that he’d bought for $8 and that he said was now worth more than $1 million. He told me that recently, all owners of these imaginary sneakers had been issued an image of a box, which was it self worth $30,000. When he opened the box, he found another pic ture of sneakers and another box, each of them valuable in their own right. “It’s this never-ending Ponzi scheme,” he said happily. “That’s what I call Ponzinomics.”

«

And that always ends well, right? Taken from Faux’s new book “Number Go Up: Inside Crypto’s Wild Rise and Staggering Fall”, which sounds like a rollicking read.
unique link to this extract

New Elon Musk biography offers fresh details about the billionaire’s Ukraine dilemma • CNN Politics

Sean Lyngaas:

»

Elon Musk secretly ordered his engineers to turn off his company’s Starlink satellite communications network near the Crimean coast last year to disrupt a Ukrainian sneak attack on the Russian naval fleet, according to an excerpt adapted from Walter Isaacson’s new biography of the eccentric billionaire titled “Elon Musk.”

As Ukrainian submarine drones strapped with explosives approached the Russian fleet, they “lost connectivity and washed ashore harmlessly,” Isaacson writes.

Musk’s decision, which left Ukrainian officials begging him to turn the satellites back on, was driven by an acute fear that Russia would respond to a Ukrainian attack on Crimea with nuclear weapons, a fear driven home by Musk’s conversations with senior Russian officials, according to Isaacson, whose new book is set to be released by Simon & Schuster on September 12.

Musk’s concerns over a “mini-Pearl Harbor” as he put it, did not come to pass in Crimea. But the episode reveals the unique position Musk found himself in as the war in Ukraine unfolded. Whether intended or not, he had become a power broker US officials couldn’t ignore.

The new book from Isaacson, the author of acclaimed biographies of Steve Jobs and Albert Einstein, provides fresh insights into Musk and how his existential dread of sparking a wider war drove him to spurn Ukrainian requests for Starlink systems they could use to attack the Russians.

«

Musk is an idiot. Why would Russia hit Crimea with nukes, when that’s territory that it views as its own? And you know that any effort by the US government to educate him about the strategic realities would be a waste of time. What a situation.
unique link to this extract

Apple is reportedly spending ‘millions of dollars a day’ training AI • The Verge

Monica Chin:

»

Apple is investing millions of dollars per day into artificial intelligence, according to a new report from The Information. The company is reportedly working on multiple AI models across several teams.

Apple’s unit that works on conversational AI is called “Foundational Models,” per The Information’s reporting. It has “around 16” members, including several former Google engineers. It’s helmed by John Giannandrea, Apple’s head of AI, who was hired in 2018 to help improve Siri. (Giannandrea has reportedly “expressed skepticism to colleagues about the potential usefulness of chatbots powered by AI language models.”)

Additional teams at Apple are also working on artificial intelligence, per The Information. A Visual Intelligence unit is developing an image generation model, and another group is researching “multimodal AI, which can recognize and produce images or video as well as text.”

These models could serve a variety of purposes. A chatbot is in the works that would “interact with customers who use AppleCare”; another would make it easier to automate multistep tasks with Siri.

«

“Millions of dollars a day” sounds like someone got their calculator out on being told Apple is spending about a billion dollars on AI per year. (My guess. A billion per year is $3m; 2bn is $6m; and so on.) Given that Apple spends about $30bn per year on R+D, that isn’t actually a dramatic amount.

But look, anything that can improve Siri is going to be welcome.
unique link to this extract

Will we ever be able to detect AI usage? • TechnoLlama

Andres Guadamuz:

»

Why, you may ask, do we need to detect AI in the first place? The detection of AI use in everyday life is becoming increasingly important for a multitude of reasons. Firstly, transparency and trust are foundational to the relationship between technology and its users, as AI tools become more integrated into our daily routines, from personalised content recommendations to LLMs, it’s crucial for users to be aware of when and how these systems are making decisions on their behalf. Knowing when AI is at play can help individuals better understand the rationale behind certain outcomes, whether it’s a movie suggestion on a streaming platform or a financial advice from a robo-advisor.

Secondly, the ethical implications of AI are vast and varied. By detecting AI use, we can ensure that these systems are being employed in ways that align with societal values and norms, and with the user’s own acceptance of AI. For instance, in areas like hiring or lending, where decisions can have profound impacts on individuals’ lives, it’s essential to know if an AI is involved, potentially perpetuating biases or making uninformed decisions. Furthermore, as concerns about privacy and data security grow, being aware of AI’s presence can help individuals protect their personal information and avoid potential misuse. In essence, detecting AI use promotes accountability, ensuring that these powerful tools are used responsibly and ethically in our ever-evolving digital landscape.

Thirdly, if we’re going to continue using essays as means of assessment, we may need to find out for certain if AI is being used.

«

I tried to think of a refutation of the importance of using essays as a means of assessment, but given that the point of an essay is to get people to show they can organise their thoughts and pick the important from the wrong or trivial, it’s hard to say that doesn’t matter in the world of work.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

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.

The UK’s National Air Traffic Control has explained how a peculiar flight plan lead to chaos in a new report. CC-licensed photo by Mark Hodson Photos 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.

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

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

Government denies U-turn on encrypted messaging row • BBC News

Zoe Kleinman, Tom Gerken and Chris Vallance:

»

platforms like WhatsApp, Signal and iMessage say they cannot access or view anybody’s messages without destroying existing privacy protections for all users, and have threatened to leave the UK rather than compromise message security.

The debate has raged for several months and for some it has turned into an argument about privacy versus the protection of children. The government insists it is possible to have both.

The Online Safety Bill is due to become law in autumn and cleared its final stage in the House of Lords on Wednesday before returning to the commons.

The government has denied that its position has changed. In a statement in the House of Lords, the minister, Lord Parkinson, clarified that if the technology to access messages without breaking their security did not exist, then Ofcom would have the power to ask companies to develop the ability to identify and remove illegal child sexual abuse content on their platforms.

Indeed, the Bill already stated that the regulator Ofcom would only ask tech firms to access messages once “feasible technology” had been developed which would specifically only target child abuse content and not break encryption.

The government has tasked tech firms with inventing these tools.

“As has always been the case, as a last resort, on a case-by-case basis and only when stringent privacy safeguards have been met, [the Bill] will enable Ofcom to direct companies to either use, or make best efforts to develop or source, technology to identify and remove illegal child sexual abuse content – which we know can be developed,” said a government spokesperson.

Some security experts suggest such tech tools may never exist, and the tech firms themselves say it is not possible.

«

Lovely solution: the government is telling the tech firms to invent a totally new mathematics. I suppose that quantum computing might, somehow, someday, make such decryption possible, but essentially it’s asking them to invent the mathematical equivalent of a perpetual motion machine.
unique link to this extract

Experts warn RAAC concrete affects thousands of UK buildings – BBC News

Pallab Ghosh:

»

Tens of thousands of government and privately owned buildings should be safety checked because of crumbling RAAC concrete, experts say.

The team that alerted the government to the problems posed by the concrete said safety checks will need to be regular.

The Loughborough University team told BBC News about their research and their advice for dealing with the problem. Prof Chris Goodier said most affected buildings were probably not dangerous but should be inspected just in case. “We’ve suddenly found out that a certain proportion of our building stock is not as good as we thought it was,” he said. “It’s a small proportion but we have millions of buildings – even if its just one% of 10 million that’s 100,000,” he said.

Prof Goodier said that as well as government buildings such as hospitals, court houses and prisons, an unknown number in the private sector offices and warehouses were also potentially affected because they contained the concrete, also known as reinforced autoclaved aerated concrete (RAAC).

His team is advising the government to send surveyors to assess the affected buildings, most of which the researchers expect to be found not to have any dangerous flaws. But they will require regular inspections.

Others buildings may need areas reinforced or have certain areas closed and a small number may need to be rebuilt. But in the longer term, the team says a new approach will be needed of regular inspections and management of possibly tens of thousands of buildings, according to Prof Sergio Cavalaro from the Loughborough team.

“Buildings that were not inspected will now need to be inspected. We need to intensify these inspections. But that will be a challenge because there are so many buildings that need inspections. So we may lack the qualified people to do it in a timely fashion,” he said.

«

You simply couldn’t get a better metaphor for the neglect of the country’s infrastructure. A neat, quick, cheap construction material whose lifespan ended during the Conservative administration. And what did that administration do about it? Nothing.
unique link to this extract

Read-it-later app Matter can now transcribe your favorite podcasts • TechCrunch

Sarah Perez:

»

Matter, a GV-backed read-it-later app that competes broadly with apps like Pocket and Instapaper, though with more of a focus on reading recommendations, is today launching a new way to use its app. The company this morning is debuting “Readable Podcasts,” a feature that will let you save favorite podcasts and transcribe their audio to text. By doing so, you can use Matter’s other tools to interact with the podcast content as you would a saved article, including by doing things like highlighting, taking notes and sharing quotes.

As you listen to the podcast, the audio transcript and text are kept in sync so you can seamlessly switch back and forth between listening and reading. This allows you to pause the podcast if you want to re-read a particular section or look up something the podcast hosts were discussing.

Or, if you’re trying to get through a dense podcast more quickly, the feature would allow you to skim the transcript so you could just jump to the key areas of interest, instead of having to fast-forward or listen to the audio at a faster speed.

According to Matter co-founder Ben Springwater, the new feature is powered by the combination of a third-party transcription service along with Matter’s own proprietary system to label speakers and ensure transcript quality.

What’s more, users don’t have to give up their favorite podcast app in order to use the new functionality. That is, instead of subscribing to a podcast in Matter itself, you’ll instead share the podcast you want to transcribe from your existing podcast app using the Save extension from the iOS Share Sheet. At launch, Matter supports Apple Podcasts, Spotify, Overcast, Castro and Pocket Casts.

«

I wonder whether the transcription will remove all the little noises – the ums and ers. I can see the potential attraction (we read faster than we listen, though perhaps not with podcasts at 2x speed) but think it will show how rambling some people are.

There’s a neat circularity: a lot of podcasts now are people reading their newsletters.
unique link to this extract

Khosla Ventures backs carbon removal company with lung-like material • CNBC

Catherine Clifford:

»

The founders of the startup, named Spiritus after the Latin word for “breath,” began work in December 2021, and the company is officially coming out of stealth on Wednesday, with the announcement of an $11 million funding raise led by prominent Silicon Valley venture capital firm Khosla Ventures, with other investors including Page One Ventures.

Spiritus has built a novel approach to direct air carbon capture that relies on a material that absorbs carbon dioxide passively. Critically, Spiritus has developed a particular architecture that mimics the alveoli in the lungs in order to maximize the surface area for carbon dioxide to make contact with the material.

This lung-like material, technically called a “sorbent,” will be shaped in round balls and laid out like artificial fruits in a carbon-capture orchard, CEO Charles Cadieu and CTO Matt Lee told CNBC in a phone interview on Tuesday.

When the lung-like “fruit” have been collected from the carbon “orchard,” they will be put in a container, where low heat will be applied to remove the carbon dioxide. The desorption process will be powered by clean energy to ensure the process is a not adding emissions to the atmosphere. Once the CO2 has been removed from the lung-like fruit, the sorbent can then be returned to the carbon orchard and reused.

«

Or, you know, just plant a whole load of trees. But where’s the 100x return in trees?
unique link to this extract

The gold jewellery made from old phones • BBC Future

Anna Turns:

»

As I walk into her small demo laboratory, Hayley Messenger, a chemist specialising in sustainable precious metals, explains why nothing here is labelled: “Everything is a secret!” she says, pouring a ‘”magic green solution” into a one-litre-capacity (35oz) glass flask of fragmented circuit boards.

She and a team of chemists and chemical analysts, together with Canadian start-up Excir, have invented and patented a clean, energy-efficient way which they claim extracts 99% of gold from the printed circuit boards found inside discarded laptops and old mobile phones. Later this year, the Royal Mint is opening a new multi-million-pound factory which will be able to process 90 tonnes of circuit boards per week once fully operational, recovering hundreds of kilogrammes of gold every year.

When the luminous mixture starts to fizz, Messenger screws the lid on, then places the flask on a tumbling machine to shake the contents. In just four minutes, any gold dissolves and leaches out into the liquid.

“This all happens at room temperature and it’s very quick,” says Messenger who explains that this chemical solution gets reused up to 20 times, with the concentration of dissolved gold increasing each time.

When another mystery solution is added, the gold becomes solid metal again. This powder is filtered out and melted down in a furnace into thumbnail-sized nuggets.

«

This story felt vaguely familiar, and for good reason: the press release trailing this came out in March 2022. Progress, of sorts.
unique link to this extract

Upscale.media

PixelBin:

»

Typically, if you wish to increase the resolution of a low-quality image, you’ll just end up with a bigger low-quality image. Instead, Upscale.media’s AI technology maintains natural image details without losing the quality – quite unbelievable, if you ask us

«

Now you can get AI to upscale your photos – that is, fill in fine detail if you want to zoom in on something, or if you’ve got a pixellated photo from somewhere. Five free downloads, or quite cheap PAYG offerings otherwise.
unique link to this extract

Ryanair boss calls air traffic chaos report rubbish • BBC News

Katy Austin and Lora Jones:

»

[Ryanair boss Michael] O’Leary told the BBC that the disruption will cost the airline between £15m and £20m in refunds for hotels, food and alternative travel arrangements.

He said that “there won’t be any issues” for customers claiming costs, but demanded that Nats, which controls the UK’s air traffic services, “accepts responsibility for its incompetence”.

Marion Geoffroy, managing director at Wizz Air UK, said that it, along with its customers, had “suffered severe disruption” because of cancellations and accommodation costs.

Tim Alderslade, chief executive of Airlines UK, said: “Airlines cannot be the insurer of a last resort. We can’t have a situation whereby airlines carry the can every time we see disruption of this magnitude.”

The group represents the likes of British Airways, EasyJet, Jet2, Ryanair, Virgin Atlantic and Tui. EasyJet boss Johan Lundgren also said that “many questions are still left unanswered” after Nats published an initial report into what exactly caused the system failure.

“An incident on this scale should not have happened and must not happen again,” he added, saying that he was looking forward to a more “wide-ranging” review.

«

The NATS report itself, into a calamitous failure of the air routing systems on August 28, explains how it happened (as does the BBC story): a route was entered which had two waypoints with the same designator. Even though they were 4,000 miles apart, the routing software became confused because of the way routing is done: you start from the beginning and search forwards to the end, and then you start from the end and search backwards. But the identical waypoint designators made it seem like the end and the start were the same, and yet not the same.

No word on how old the system is (though this subsystem has been in continuous operation since 2018, apparently) or what it runs on, or is written in; but a software fix was deployed by last Monday.
unique link to this extract

IBM taps AI to translate COBOL code to Java • TechCrunch

Kyle Wiggers:

»

COBOL, or Common Business Oriented Language, is one of the oldest programming languages in use, dating back to around 1959. It’s had surprising staying power; according to a 2022 survey, there’s over 800 billion lines of COBOL in use on production systems, up from an estimated 220 billion in 2017.

But COBOL has a reputation for being a tough-to-navigate, inefficient language. Why not migrate to a newer one? For large organizations, it tends to be a complex and costly proposition, given the small number of COBOL experts in the world. When the Commonwealth Bank of Australia replaced its core COBOL platform in 2012, it took five years and cost over $700m.

Looking to present a new solution to the problem of modernizing COBOL apps, IBM today unveiled Code Assistant for IBM Z, which uses a code-generating AI model to translate COBOL code into Java. Set to become generally available in Q4 2023, Code Assistant for IBM Z will enter preview during IBM’s TechXchange conference in Las Vegas early this September.

«

Hurrah for translating old COBOL code, but — into Java? Who’s going to check that? What’s the confidence level in what is output?
unique link to this extract

LinkedIn is getting personal and deep beyond the job searches • The Washington Post

Danielle Abril:

»

After the pandemic blurred the lines between work and home, many employees reprioritized their lives, giving more emphasis to well-being and family. As a result, workers have become comfortable getting personal on LinkedIn, sharing engagement announcements, their fertility journeys, cancer diagnoses, relationship statuses, funny pet moments, even what they cooked for dinner. Though many post this content sparingly, some say it humanizes themselves to their professional network. Others say their stories tie in with lessons that could be applied to business, while some find the content annoying.

LinkedIn may be benefiting from shifts in the social media landscape. X, formerly known as Twitter, has lost many of its power users as the app becomes less functional for free accounts under billionaire owner Elon Musk. Meanwhile, after an impressive debut, Threads, the latest app from Facebook owner Meta, has seen a slowdown in activity. Instagram and TikTok continue to thrive among younger users, especially with video content.

LinkedIn says its user engagement doubled during the beginning of the pandemic and experienced a 40% rise between 2021 and 2023. The number of users who visit the site at least once a month is forecast to increase by more than 8 million to 84.1 million by 2027, with Gen Z serving as a major driver or growth, data from market research company Insider Intelligence shows. In that same period, Facebook users are expected to decline by 600,000 to 177.3 million, and Instagram’s users are forecast to grow by 20.2 million to 155.4 million.

LinkedIn says it saw a jump in personal posts during the height of the pandemic, but that has since slowed. To ensure people’s feeds stay useful, the company made changes to its algorithm. It now surfaces more posts from people’s direct connections and followers as well as those from people outside their networks that are grounded in professional knowledge and advice. Still, LinkedIn content is mirroring a larger societal shift that is happening in the workplace.

«

Still wonder about LinkedIn. What sort of social network is it really?
unique link to this extract

Mozilla report finds that new cars give out lots of your info • Gizmodo

Thomas Germain:

»

Bad news: your car is a spy. If your vehicle was made in the last few years, you’re probably driving around in a data-harvesting machine that may collect personal information as sensitive as your race, weight, and sexual activity. Volkswagen’s cars reportedly know if you’re fastening your seatbelt and how hard you hit the brakes.

That’s according to new findings from Mozilla’s *Privacy Not Included project. The nonprofit found that every major car brand fails to adhere to the most basic privacy and security standards in new internet-connected models, and all 25 of the brands Mozilla examined flunked the organization’s test. Mozilla found brands including BMW, Ford, Toyota, Tesla, and Subaru collect data about drivers including race, facial expressions, weight, health information, and where you drive. Some of the cars tested collected data you wouldn’t expect your car to know about, including details about sexual activity, race, and immigration status, according to Mozilla.

“Many people think of their car as a private space — somewhere to call your doctor, have a personal conversation with your kid on the way to school, cry your eyes out over a break-up, or drive places you might not want the world to know about,” said Jen Caltrider, program direction of the *Privacy Not Included project, in a press release. “But that perception no longer matches reality. All new cars today are privacy nightmares on wheels that collect huge amounts of personal information.”

«

Cars have become computers, but their privacy policies haven’t kept up (at least in the US; in Europe they’d be held back by the GDPR). One does wonder quite how your car collects data about your sexual activity. Perhaps better not to ask.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

There are lots of myths about Dark Mode on phones, and the Nielsen-Norman Group has opinions about them. CC-licensed photo by akitada31 on Flickr.

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

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

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

Fitbits, Apple Watches and other fitness trackers really do make you healthier • WSJ

Julie Jargon:

»

There’s been widespread scepticism in the scientific and medical community about whether fitness trackers can improve health, said researchers from the University of South Australia. So they conducted a study of 39 systematic reviews involving nearly 164,000 participants of all ages. 

The result, published last year: when people wore wearables, they on average increased their daily step counts by 1,800—and boosted their daily walking time by 40 minutes.

Fitness trackers are in fact effective in helping people lose weight, increase their aerobic activity and lower their blood pressure, the researchers concluded.

Just the practice of tracking steps can make you think more about health and adopt healthier habits, according to a clinical trial in which participants wore Apple Watches for five weeks.

And in a six-week study in Australia, when parents and their kids all wore Garmin activity trackers, they became more aware of everyone’s physical activity and inactivity. They were also more motivated to do outdoor activities together.

Despite the overall success, getting families to make a long-term habit out of fitness tracking can be a challenge. Researchers in the U.K. gave Fitbits to 24 families with kids between the ages of 5 and 9 for four weeks. The families’ Fitbit use was high in the beginning, but fell off in the final two weeks. The novelty wore off.

To ensure the success of wearables in your family, you can set goals, introduce friendly competition and participate in challenges.

«

And yet, as she points out, sales of fitness bands (eg Fitbits) have declined over the past five years, though they bumped up a bit in the pandemic.
unique link to this extract

Daniil Medvedev watches US Open on ‘illegal’ streams due to TV blackouts • Insider.com

Meredith Cash:

»

Daniil Medvedev just wants to watch some tennis.

The world No. 3 is competing at this year’s US Open — where he won his first Grand Slam in 2021. But when he’s not on the court in Flushing Meadows, Medvedev wants to tune in for some of the other matches taking place at the USTA Billie Jean King National Tennis Center.

There’s only one problem: his hotel isn’t showing the tournament.

“I guess in a lot of hotels they have Spectrum,” Medvedev said in his post-match press conference Monday. “I cannot watch it on TV anymore.”

The cable provider is currently in a standoff with Disney — the parent company of ESPN, which has exclusive broadcast rights to the tournament — over their failure to negotiate terms on a new agreement. Until the two sides agree to a new deal, channels that fall under the Disney umbrella will not be available to Spectrum customers.

But Medvedev isn’t willing to miss the US Open matches he wants to watch due to the feud. He said he’ll “have to find a way.”

“I will, I don’t know if it’s legal or illegal, … but I got internet, these pirate websites, so I watch tennis there,” the 27-year-old said with a smirk. “I have no other choice.”

«

There’s where the money that Spectrum and Disney can’t agree on is going: to the pirate websites (one way or another). Medvedev is quietly, subversively funny – always taking some pleasure in trolling crowds if they take a dislike to him, like a slightly evil (maybe “even more evil”) Jimmy Connors.
unique link to this extract

Large language models, explained with a minimum of math and jargon • Understanding AI

Timothy B Lee:

»

If you know anything about this subject, you’ve probably heard that LLMs are trained to “predict the next word,” and that they require huge amounts of text to do this. But that tends to be where the explanation stops. The details of how they predict the next word is often treated as a deep mystery.

One reason for this is the unusual way these systems were developed. Conventional software is created by human programmers who give computers explicit, step-by-step instructions. In contrast, ChatGPT is built on a neural network that was trained using billions of words of ordinary language.

As a result, no one on Earth fully understands the inner workings of LLMs. Researchers are working to gain a better understanding, but this is a slow process that will take years—perhaps decades—to complete.

Still, there’s a lot that experts do understand about how these systems work. The goal of this article is to make a lot of this knowledge accessible to a broad audience. We’ll aim to explain what’s known about the inner workings of these models without resorting to technical jargon or advanced math.

We’ll start by explaining word vectors, the surprising way language models represent and reason about language. Then we’ll dive deep into the transformer, the basic building block for systems like ChatGPT. Finally, we’ll explain how these models are trained and explore why good performance requires such phenomenally large quantities of data.

«

A good backgrounder.
unique link to this extract

Meta axes support for news in Europe • Financial Times

Daniel Thomas:

»

News publishers have hit out at Meta after the tech giant decided to axe Facebook News in Europe and end a scheme to fund local journalism in the UK. 

On Tuesday, Meta said that it would “deprecate” the dedicated tab on Facebook that showcases news stories in early December in the UK, France and Germany.

The group said that this was part of an “ongoing effort to better align our investments to our products and services people value the most”.

Meta said that publishers will continue to have access to their Facebook accounts and pages, where they can post their news article links and content, and it would “honour our obligations under all existing Facebook News agreements with news publishers in the UK, France and Germany until they expire”.

The group said that the News tab makes up less than 3% of what people around the world see in their Facebook feed, “so news discovery is a small part of the Facebook experience for the vast majority of people”.

However, newspaper executives warned the decision will deprive a valuable source of income and traffic for media groups. Meta has already axed Instant Articles, the mobile-friendly format that quickly loaded news articles on the Facebook app.

In July, Reach, the UK’s largest commercial news publisher with titles such as the Mirror and Express, blamed this move on a drop in digital revenue in the first half of the year. 

«

Another reminder that Facebook is indifferent to media except as a source of Content. And in case you needed any further demonstration…
unique link to this extract

Exclusive: Meta’s Canada news ban fails to dent Facebook usage • Reuters

Katie Paul and Steve Scherer:

»

Meta’s decision to block news links in Canada this month has had almost no impact on Canadians’ usage of Facebook, data from independent tracking firms indicated on Tuesday, as the company faces scorching criticism from the Canadian government over the move.

Daily active users of Facebook and time spent on the app in Canada have stayed roughly unchanged since parent company Meta started blocking news there at the start of August, according to data shared by Similarweb, a digital analytics company that tracks traffic on websites and apps, at Reuters’ request.

Another analytics firm, Data.ai, likewise told Reuters that its data was not showing any meaningful change to usage of the platform in Canada in August.

The estimates, while early, appear to support Meta’s contention that news holds little value for the company as it remains locked in a tense standoff in Canada over a new law requiring internet giants to pay publishers for the news articles shared on their platforms.

Meta declined to comment on the estimates.

The Online News Act, passed by the Canadian parliament in June, forces platforms like Meta and Google parent Alphabet to negotiate commercial deals with Canadian news publishers for use of their content.

Both Meta and Alphabet’s Google have said the law is unworkable for their businesses. Meta, in particular, has said links to news articles make up less than 3% of the content on its Facebook feeds and have no economic value to the company.

«

Colour me totally unsurprised. The Canadian government is now at the “find out” stage of FAAFO. Why would Facebook include news links when it doesn’t need to? It’s just delusional.
unique link to this extract

Airbnb’s NYC listings could be down 70% after September 5 • Skift

Dennis Schaal:

»

Airbnb says it will turn off calendars for stays with ineligible hosts starting September 5. 

There are several rules that determine eligibility, including: registered hosts can only book stays for fewer than 30 days and they must be present during the guest’s stay. In addition, hosts must register with NYC’s Office of Special Enforcement – just 257 applications have been approved so far, Skift reported this week. 

And how the city will enforce the law is unclear. Skift obtained an email from the city’s Public Advocate office to a host that stated: “OSE [Office of Special Enforcement] will not be proactively issuing fines unless there is an egregious violation; such as hosting a large number of people without being registered. There are a large number of applications pending; however, people who have been making significant efforts to register may be given leniency by the agency.”

Still, Airbnb has said it generated $85m in revenue from New York City in 2022, and the number of listings at risk is sizable, according to data provided to Skift by AirDNA. 

• There were 47,000 total listings as of July, of which only about 23,000 were active
• Of the 23,000, 9,500 were private rooms or shared spaces where hosts are eligible to apply for registrations from OSE. Only some of those will ultimately be approved.
• Around 13,500 were for entire homes or apartments. Of these, around 6,000 would seem exempt because they are either hotels or accept stays of 30 days or longer
• That leaves 7,500 listings mostly at risk of no longer being able to lawfully host unless they change operations.

«

The regulator comes for everyone eventually. AirBnB has lobbied hard, but without success.
unique link to this extract

Dark Mode: how users think about it and issues to avoid • Nielsen Norman Group

Tanner Kohler and Amy Zhang:

»

In a recent survey of 115 mobile users asking what mode they generally have their mobile device in, roughly 1/3 said dark mode, 1/3 said light mode, and 1/3 said a combination of both. The argument that dark mode improves the user experience (and accessibility in some cases) always seems to circle back to the same few reasons, mentioned by users, designers, and developers alike:

• Reduced eye strain
• Battery savings
• Aesthetic appeal
• Improved accessibility for those with visual impairments (e.g., cataracts)

After updating to allow for dark mode, the Google Docs application on Android touts that it’s “Easier on the eyes at night. Easier on the battery during the day.”

Let’s take a closer look at the commonly cited reasons for supporting and using dark mode.

Reducing eye strain is the most common reason users with normal vision mention for using dark mode. As one research participant put it, “My eyes have always been very sensitive to bright lights. So ideally, I use dark mode on everything I can. […] I don’t think I can go back to normal. [I have] a lot less eye strain. I don’t think I’m getting headaches as much […]”

While dark mode is not guaranteed to reduce headaches, this belief is widespread and motivates many users to permanently set their devices to dark mode.

Dark mode does slightly reduce the amount of light being taken in by the retina and might feel easier on the eyes during a single session — particularly in dark environments. However, some research has failed to find a significant difference in reported measures of eye strain and headaches when participants complete tasks in both light mode and dark mode.

…Many people simply like the way dark mode looks. Some survey respondents made comments such as, “it’s more visually appealing to me,” “I like the look of the dark screen,” and “dark mode is way cooler.”

Some designs are even built in dark mode and do not allow users to transition to light mode. In most of these cases, building in dark mode first is an aesthetic choice; it does not improve usability.

«

You can say that again. I detest dark mode. Maybe I’m just too used to paper.
unique link to this extract

Netflix is going to let DVD subscribers keep unreturned discs for free • The Verge

Jay Peters:

»

Netflix won’t charge DVD.com customers for any discs they still have after September 29th, the company announced from its DVD.com account on X on Monday. That generous offer, combined with Netflix’s recent announcement that it may send customers as many as 10 extra discs from their queues, means that some people might end up with a bunch of disc copies of movies, courtesy of Netflix.

DVD.com customers will need to visit a special link on DVD.com by August 29th to apply for the promotional offer. Netflix will then send up to 10 random discs based on the movies in the subscriber’s queue.

«

Nifty! Wonder if this will drive a wave of signups?
unique link to this extract

How Google’s augmented-reality dream and pursuit of Apple fell apart • Business Insider

Hugh Langley:

»

During its I/O conference in May 2022, Google teased a pair of Iris [augmented reality] glasses running a feature that could translate languages in real time. The demo generated positive buzz, but the group shifted away from the idea soon after, a person familiar with it said.

“Every six months there was a major pivot in the program,” they said. “They would look at it and say, ‘We want a slightly different product.'”

In early 2022, reports began surfacing that Apple’s long-rumored headset was in the final stretch, and Google leaders began to worry.

“At Google, there is a great fear that when Apple releases new products, it shifts the landscape,” a former employee said.

Around this time, Google partnered with one of its oldest hardware allies, Samsung. Samsung wanted to build a headset device similar to Apple’s Vision Pro, which would mix virtual and augmented reality, and Google would design Android-based software to power it.

The project, code-named Moohan, created a political headache inside Google. Samsung told Google it didn’t want other hardware teams working on Google AR products to be privy to the project’s technology for fear they might build a competing product based on that information, according to two people familiar with the strategy.

That created a problem for Iris. “How could you build glasses and not get Samsung riled up?” one person close to the project said. The partnership also meant Samsung would be more likely to call the shots on product features, one former employee said. “It’s easy to end up in a situation where nobody is driving,” they added.

The dynamic is already playing out. South Korea’s SBS Biz reported earlier this month that Samsung delayed the headset after the Vision Pro’s reveal over fears its device wouldn’t be enough to go toe-to-toe with Apple. The delay could push Samsung and Google’s device launch to summer 2024, the outlet reported.

«

A bit of followup on yesterday’s report which mentioned this in passing. There’s also a writeup with some extra at Ars Technica.
unique link to this extract

Meta reportedly partnering with LG for 2025 Quest Pro 2 • Upload VR

David Heaney:

»

In November last year South Korean news outlet SBS Biz reported Meta had entered talks with LG Display with the aim of securing OLED microdisplay supply for future headsets. Microdisplays have significantly higher pixel densities and thus can enable higher resolution more compact headsets.

A new report from another Korean news outlet, Maeil Business Newspaper, suggests this has now resulted in a much broader partnership with the LG Group, involving multiple of its subsidiaries, to build future Quest Pro headsets.

LG Electronics will reportedly handle production, using LG Display displays, LG Energy batteries, and other components from LG Innotek. The first product from the partnership is reportedly slated for 2025, priced at around $2000.

…This wouldn’t be the first time Meta has partnered with an experienced consumer electronics company for a VR headset. Oculus Go was manufactured by Xiaomi, and Oculus Rift S was manufactured and co-designed by Lenovo. In both cases, the headset bore the partner company’s logo on the side as well as the Oculus branding on the front. How LG’s branding will appear on the reported 2025 headset is not yet known.

This also isn’t the only software-first Big Tech company partnering with a Korean hardware company on an XR headset. Samsung announced in February it is building a headset, with Google handling the software.

«

The VR-only headset space remains a confused mess, with no clear message about what they’re for. Is it gaming? Is it business meetings? If it’s both, which is the market to focus on? Apple, as usual, is going for “none of the above”, and yet seems most likely to come through the middle. As usual.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Wild boar in Europe are more radioactive than they should be. But the reason isn’t Chernobyl. CC-licensed photo by Torsten Behrens 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 about 0845 UK time. Free signup.

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

AI systems have learned how to deceive humans. What does that mean for our future? • The Conversation

Simon Goldstein and Peter Park:

»

Perhaps the most disturbing example of a deceptive AI is found in Meta’s CICERO, an AI model designed to play the alliance-building world conquest game Diplomacy.

Meta claims it built CICERO to be “largely honest and helpful”, and CICERO would “never intentionally backstab” and attack allies.

To investigate these rosy claims, we looked carefully at Meta’s own game data from the CICERO experiment. On close inspection, Meta’s AI turned out to be a master of deception. In one example, CICERO engaged in premeditated deception. Playing as France, the AI reached out to Germany (a human player) with a plan to trick England (another human player) into leaving itself open to invasion.

After conspiring with Germany to invade the North Sea, CICERO told England it would defend England if anyone invaded the North Sea. Once England was convinced that France/CICERO was protecting the North Sea, CICERO reported to Germany it was ready to attack. This is just one of several examples of CICERO engaging in deceptive behaviour. The AI regularly betrayed other players, and in one case even pretended to be a human with a girlfriend.

…advanced AI systems can autonomously use deception to escape human control, such as by cheating safety tests imposed on them by developers and regulators.

«

The authors expand on this in an arXiv paper on “AI Deception: a survey of examples, risks and potential solutions”.
unique link to this extract

New iPhone, new charger: Apple bends to EU rules • BBC News

Chris Vallance and Zoe Kleinman:

»

Apple’s latest iPhone will almost certainly feature a USB-C charge point when it is unveiled on 12 September.

The firm’s phones currently use its proprietary Lightning adaptor, unlike rivals, including Samsung.

A European Union law requires phone manufacturers to adopt a common charging connection by December 2024 to save consumers money and cut waste.

Most new Apple products such as the latest iPads already use USB-C, but the firm had argued against the EU rule.

When it was introduced in September 2021, an Apple representative told BBC News: “Strict regulation mandating just one type of connector stifles innovation rather than encouraging it, which in turn will harm consumers in Europe and around the world.”

«

As I’ve said previously, I’m sure Apple will portray the move to USB-C as something it did first (on Macs) and that now the iPhone’s cameras can capture so much data it’s time to move on from Lightning, which first appeared in 2012, well before USB-C, to something that can handle much faster data rates.

But this story? Ah, read the comments and be transported back to those smartphone boom days. Some people still carry all the same irrational prejudices.
unique link to this extract

‘A psychological weapon’: inside a Ukrainian factory making decoy kit • The Guardian

Emma Graham-Harrison:

»

In a dusty workshop, a unique group of Ukrainian weapons experts race to produce artillery guns that will never be fired, radar trucks that cannot detect anything, and missiles without explosives.

The pieces are decoys that aim to draw Russian fire, wasting enemy ammunition, missiles and drones while protecting real equipment and the soldiers manning it.

The team’s skill, honed over more than a year, is shaping plastic, scrap wood, foam and metal into copies of advanced weapon systems, precise enough to convince Russian operators of drone cameras and battle-seasoned troops on the ground that they are real military targets.

They measure success by how quickly their products are destroyed. “When the military come to us and says ‘we are out of these’, it means we were totally successful in our job,” says one.

A cupboard near their workshop is stuffed with expensive souvenirs of that success, including the engine and crumpled fragments of an Iranian-made Shahed suicide drone and the crashed wing of a Russian-made Lancet loitering drone, both lured to attack the fake equipment.

Hitting a decoy is a costly mistake for Russia, and also means one less attack on a real Ukrainian position. “These can save the lives of our guys, our friends who are serving,” the worker adds. “We have an agreement with the military to share pictures and remains of attacks (on decoys), as proof we did a good job.”

…Three senior managers at the firm came up with the idea of making decoy weapons at the start of the war, when Ukraine’s troops seemed dangerously outgunned. The influx of western weaponry that has helped hold Russia at bay had only just begun flowing across the border.

“We thought if the Russians saw a lot of weapons, they might be scared to move forward, or to shell an area. It’s a psychological weapon,” says one of them. “The company fully supported it.”

«

unique link to this extract

Burning Man organisers plot ‘exodus’ of people trapped in desert mud • Financial Times

Antoine Gara and Amanda Chu:

»

Organisers of the Burning Man festival have announced an “exodus” of thousands of attendees stranded in the Nevada desert after torrential rains transformed an event that has become a magnet for technology executives, venture capitalists and social media influencers into a “nightmare” of muck and broken toilets.

“Exodus likely to begin around noon [Pacific] today, Monday 9/4,” the organisers of the event in Black Rock City, Nevada said, days after they urged attendees to shelter in place and conserve food and water as deepening mud shut roads away from the venue.

The burning of a human effigy and temple that culminates the annual event is now planned for Monday night.

News that roads away from the event would reopen as the mud dries ends days of misery for thousands of festival-goers attending what was once the US’s signature counterculture event, but now attracts celebrities alongside older hippies and families.

«

Apparently the rain has also brought out three-eyed shrimp that live in the desert floor. Bet that’s fun on the drugs.
unique link to this extract

LogicMonitor customers hit by hackers, because of default passwords • TechCrunch

Lorenzo Franceschi-Bicchierai:

»

Some customers of the network security company LogicMonitor have been hacked due to the use of default passwords, TechCrunch has learned.

A LogicMonitor spokesperson confirmed to TechCrunch that there’s “a security incident” affecting some of the company’s customers.

“We are currently addressing a security incident that has affected a small number of our customers. We are in direct communication and working closely with those customers to take appropriate measures to mitigate impact,” LogicMonitor’s spokesperson Jesica Church said in a statement.

The incident is due to the fact that, until recently, LogicMonitor was assigning customers default — and weak — passwords such as “Welcome@” plus a short number, according to a source at a company that was impacted by the incident, and who asked to remain anonymous as they were not authorized to speak to the press.

“When you set up an account with [LogicMonitor], they define a default password and all user accounts for your organization/account are made with that password,” the source told TechCrunch. “They also didn’t require the changes, nor were they temporary passwords, until this week. Now the setup password lasts 30 days and must be changed on first login.”

«

So, so disastrous. If you get something from a company and the default password isn’t randomised in some way, either figure out a way to return it, or get ready to be hacked – as here.
unique link to this extract

‘That doesn’t reflect the price for the planet’: France wants to put a stop to mega low airfares • Euronews

Gael Camba:

»

France’s Transport Minister has said he will propose that the EU brings in a minimum price on flights to reduce carbon emissions.

Clément Beaune says he wants to “fight against social and environmental dumping” and that “ten euro plane tickets aren’t possible anymore”, referring to low-cost airline fares. He said a mega low ticket price “doesn’t reflect the price for the planet.”

But are more expensive flights the solution to reducing aviation’s carbon footprint?

“Anything that makes airlines pay a fair share of the environmental cost that they create is a good thing”, says Jon Worth, travel expert and founder of Trains for Europe campaign. “But we should be dealing with frequent flyers and this does not deal with them. It might reduce nice city weekends for some people but it’s not going to stop or reduce this regular flying elite.”

In France, 2% of people take half of all flights, according to research published by the climate campaign group Possible; 15% of Brits take 70% of flights and 8% of the Dutch take 42%.

…[However] on average, European train tickets are twice as expensive as flights, according to a Greenpeace report from July.

Only 12 train lines were found to be fast, reliable and cheaper than flights, over the 112 most important routes analysed by Greenpeace.

«

Air travel has been ridiculously subsidised – through avoiding fuel taxes, and by not having to pay for its climate effects – since forever.
unique link to this extract

In dispute with Disney over ESPN, Spectrum owner says cable model is broken • The New York Times

Benjamin Mullin:

»

One of the biggest cable companies in the United States has a message for media companies, its major partners in a decades-old business: The traditional cable-TV model is broken, and it needs to be fixed or abandoned.

Cable TV has become too expensive for consumers and providers, Charter Communications said in an 11-page presentation to investors on Friday, adding that cord-cutters and rising fees are contributing to a “vicious video cycle.”

The presentation comes amid negotiations between Charter and the Walt Disney Company, owner of popular cable channels including ESPN and FX, which will not be available to Charter’s nearly 15 million pay-TV subscribers until both sides agree on how much Charter will pay Disney to carry its channels. Subscribers to Charter’s Spectrum TV service will be without access to the U.S. Open tennis tournament and college football games during a holiday weekend.

These so-called carriage fights are commonplace in the media industry, with channels going dark for days or weeks on cable systems while the two sides — cable providers and content creators — haggle over how much the channels are worth and how to bundle them. But Charter’s suggestion that parts of its own business model are in disrepair adds a new wrinkle to the crisis facing the cable-TV business.

The fight comes at a time of declining subscriptions: more than five million Americans end their cable-TV subscriptions annually, according to research from SVB MoffettNathanson.

…Adding to the challenges, tech companies like Apple and Amazon are willing to pay top dollar to acquire live sports rights, further driving up programming costs. Cable companies, for their part, have weaned themselves off depending wholly on traditional TV revenue, by offering services like wireless internet.

«

Cable doesn’t work, streaming doesn’t work.. what the hell does work then?
unique link to this extract

Revealed: Home Office secretly lobbied for facial recognition ‘spy’ company • The Guardian

Mark Townsend:

»

Senior officials at the Home Office secretly lobbied the UK’s independent privacy regulator to act “favourably” towards a private firm keen to roll out controversial facial recognition technology across the country, according to internal government emails seen by the Observer.

Correspondence reveals that the Home Office wrote to the Information Commissioner’s Office (ICO) warning that policing minister, Chris Philp, would “write to your commissioner” if the regulator’s investigation into Facewatch – whose facial recognition cameras have provoked huge opposition after being installed in shops – was not positive towards the firm.

An official from the Home Office’s data and identity directorate warned the ICO: “If you are about to do something imminently in Facewatch’s favour then I should be able to head that off [Philp’s intervention], otherwise we will just have to let it take its course.”

The apparent threat came two days after a closed-door meeting on 8 March between Philp, senior Home Office officials and Facewatch.

Facewatch uses cameras to check faces against a watch list and, despite widespread concern over the technology, it has already been introduced in hundreds of high street shops and supermarkets.

The use of facial recognition has provoked fierce criticism over its impact on privacy and human rights, with the European Union seeking to ban the technology in public spaces through proposed legislation.

«

Facewatch has been going for a long time, but its tight focus on just being used in commercial outlets (initially to identify thieves, subsequently to ban people on a blacklist) has largely kept it out of the public eye.
unique link to this extract

Scientists finally know why Europe’s wild boar are surprisingly radioactive • The Washington Post

Kasha Patel:

»

On April 26, 1986, the infamous explosion at a Chernobyl nuclear power plant unleashed large amounts of radiation into the atmosphere, an event that contaminated wildlife across country lines. The radiation levels seen in animals as a result has decreased in recent years — with the exception of one animal: the wild boar.

For years, scientists questioned why levels of a radioactive isotope known as caesium-137 have remained surprisingly high in wild boars rooting around Germany and Austria, while decreasing in other deer and roe deer. In a new study released last week, a team of researchers finally solved this “wild boar paradox.” They uncovered that the main radioactive source is not the Chernobyl accident but nuclear weapons testing from the 1960s.

“My mind was blown when I realized how relevant this source of radioactive contamination in general still is,” said Georg Steinhauser, a radiochemist at TU Wien and author of the new study. Steinhauser said people might not think that 60 years after a nuclear weapons explosion, wild boar populations would still be contaminated with radiation levels well above the regulatory food limit.

Researchers have previously proposed other explanations to no avail. Some thought the contamination levels in wild boar were actually decreasing, but the data didn’t show up in the limited number of studied samples. Steinhauser previously suggested that perhaps the caesium dissolves better in fat tissue in wild boars and stays present longer, but subsequent research did not support that assertion either.

…Even if the Chernobyl accident had never happened, “some of the wild boars would actually still exceed the regulatory limits for food safety limits only because of the weapons tests today,” said Steinhauser. “I think this is pretty mind-blowing because they were 60 years ago.”

«

But but but! You’d have to eat “insane” amounts to show any “meaningful” radiation. We’re really good at detecting radioactivity, less good at evaluating its (low) risks.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Machine learning-based systems can now pilot drones through an obstacle course better than the best humans. CC-licensed photo by Ed Schipul on Flickr.

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

Did you miss last Friday’s Social Warming Substack? It’s about political advertising on social media. Free signup.

A selection of 11 links for you. Self-piloting. I’m @charlesarthur on Twitter. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

Foxconn puts its empty buildings in Wisconsin up for sale • The Verge

Nilay Patel:

»

remember when Foxconn bought a bunch of buildings around Wisconsin just before an election and said it was building “innovation centers” around the state in a transparent attempt to build support for the giant tax credits it was given to build an LCD factory that never arrived? Yeah, it’s selling two of those buildings. The news was first reported by Wisconsin Public Radio, which got a quote from Foxconn saying that “selling its Green Bay property, known as the Watermark building, will add to the vibrancy of the city’s downtown.” Very good.

You might recall two of these buildings, in Eau Claire and Green Bay, because The Verge’s Josh Dzieza went and looked in the windows months after they were purchased and noticed they were empty. This groundbreaking reporting prompted Foxconn’s Alan Yeung to say that the buildings were not empty at an event celebrating the purchase of yet another building in Madison. That building has never been occupied by Foxconn, and two of its floors are now for lease, as reported by the Wisconsin State Journal.

“I can assure you that they are not empty,” Yeung said at the time. “We do have a plan and we actually will make sure the building is adequate and well-equipped before we move people in. So you will see a lot more coming in the next months, the next year or so,” he added.

“I can assure you it will not be empty and they’re not empty right now,” he said.

Anyway, we checked in again a year later, and the buildings were still empty.

Foxconn never built an LCD factory or created 13,000 jobs in southeastern Wisconsin, even after getting a sweetheart tax deal from the state and local government and a groundbreaking ceremony featuring then-President Donald Trump wielding a golden shovel.

«

Shocking, eh. Such a longrunning boondoggle, but The Verge has been on its tail all the time.
unique link to this extract

Officials investigate death at Burning Man as thousands stranded by floods • The Guardian

Michael Sainato:

»

Over 70,000 attendees of the annual Burning Man festival in the Black Rock desert of Nevada are stranded as the festival comes to a close on Monday due to heavy rains that have cut off access to the site.

Attendees have been ordered to shelter in place and to conserve food, water, and fuel, although no shortages have been reported. A death that occurred at the festival is currently under investigation, but no details have been released, including the identity of the deceased or the suspected cause of death.

All traffic apart from emergency vehicles in and out of the festival site has been halted. More rain is forecast at the festival site on Sunday afternoon. Local officials said some attendees have been walking out of the site, but conditions remain too wet and muddy for vehicles to get out and could trap many attendees at the site for days.

“We do not currently have an estimated time for the roads to be dry enough for RVs or vehicles to navigate safely. Monday late in the day would be possible if weather conditions are in our favor. We will let you know. It could be sooner, and it could be later,” said an update on the Burning Man website on Saturday evening.

“We are also deploying buses to Gerlach to take people to Reno who might walk off the playa. See our recommendations on when walking is viable or not. This is not likely a 24-hour operation at this time.”

The update said the site is working to set up mobile cell service and internet trailers at the festival site and trying to configure the organization’s on-site wifi for public access.

«

Rain isn’t unknown at Burning Man, but people seem to be shocked that it would turn the desert into mud and have the audacity to interrupt their travel plans.
unique link to this extract

Person died after Cruise cars blocked ambulance, SFFD says • SF Gate

Ariana Bindman:

»

On Aug. 14, two stalled Cruise vehicles delayed an ambulance from leaving the scene of a crash in which a driver had hit a pedestrian with their car, according to reports from the San Francisco Fire Department. The pedestrian later died of their injuries, which first responders linked to the delay in getting them to the hospital. 

“The fact that Cruise autonomous vehicles continue to block ingress and egress to critical 911 calls is unacceptable,” one emergency responder wrote in a report. Cruise spokesperson Tiffany Testo countered that one of the cars cleared the scene and that traffic to the right of it remained unblocked. “The ambulance behind the AV had a clear path to pass the AV as other vehicles, including another ambulance, proceeded to do,” she wrote in a statement to SFGATE. 

According to several reports written by first responders, first obtained by Forbes, emergency personnel arrived at Seventh Street and Harrison in SoMa and began treating a “critically injured” pedestrian who had been struck by a car. The patient was quickly loaded into an ambulance, but the ambulance driver was unable to immediately leave the scene, according to two reports written by members of the ambulance team.

Two autonomous Cruise vehicles and an empty San Francisco police vehicle were blocking the only exits from the scene, according to one of the reports, forcing the ambulance to wait while first responders attempted to manually move the Cruise vehicles or locate an officer who could move the police car.

«

So it wasn’t entirely the autonomous cars at fault. But they were quite at fault. Maybe they need an emergency service override.
unique link to this extract

X, formerly known as Twitter, may collect your biometric data and job history • CNN Business

Brian Fung and Clare Duffy:

»

X, the social media platform formerly known as Twitter, said this week it may collect biometric and employment information from its users — expanding the range of personal information that account-holders may be exposing to the site.

The disclosures came in an update to the company’s privacy policy, which added two sections related to the new data collection practice.

“Based on your consent, we may collect and use your biometric information for safety, security, and identification purposes,” the policy read.

In addition, under a new section labeled “job applications,” X said it may collect users’ employment and educational history.

The company also said it could collect “employment preferences, skills and abilities, job search activity and engagement, and so on” in order to suggest potential job openings to users, to share that information with prospective third-party employers or to further target users with advertising.

For X Premium users, the company will give an option to provide a government ID and a selfie image for verification purposes. The company may extract biometric data from both the government ID and the selfie image for matching purposes, the company told CNN in a statement.

«

This sounds like a terrible idea from start to finish. So no doubt Musk will try to make it compulsory. Wanting to become LinkedIn is a nice wish, but LinkedIn is going to defend that turf with everything it’s got.
unique link to this extract

Population collapse almost wiped out human ancestors, say scientists • The Guardian

Hannah Devlin:

»

Early human ancestors came close to eradication in a severe evolutionary bottleneck between 800,000 and 900,000 years ago, according to scientists.

A genomics analysis of more than 3,000 living people suggested that our ancestors’ total population plummeted to about 1,280 breeding individuals for about 117,000 years. Scientists believe that an extreme climate event could have led to the bottleneck that came close to wiping out our ancestral line.

“The numbers that emerge from our study correspond to those of species that are currently at risk of extinction,” said Prof Giorgio Manzi, an anthropologist at Sapienza University of Rome and a senior author of the research.

However, Manzi and his colleagues believe that the existential pressures of the bottleneck could have triggered the emergence of a new species, Homo heidelbergensis, which some believe is the shared ancestor of modern humans and our cousins, the Neanderthals and Denisovans. Homo sapiens are thought to have emerged about 300,000 years ago.

“It was lucky [that we survived], but … we know from evolutionary biology that the emergence of a new species can happen in small, isolated populations,” said Manzi.

Prof Chris Stringer, the head of human origins at the Natural History Museum in London, who was not involved in the research, said: “It’s an extraordinary length of time. It’s remarkable that we did get through at all. For a population of that size, you just need one bad climate event, an epidemic, a volcanic eruption and you’re gone.”

The decline appears to coincide with significant changes in global climate that turned glaciations into long-term events, a decrease in sea surface temperatures, and a possible long period of drought in Africa and Eurasia. The team behind the work said the time window also coincides with a relatively empty period on the fossil record.

«

It’s an amazing story, and the researchers think it partly explains why there are substantial gaps in the human fossil record: because there were so few humans alive to become fossilised.
unique link to this extract

Champion-level drone racing using deep reinforcement learning • Nature

Elia Kaufmann et al from the University of Zurich’s Robotics and Perception group, and Intel:

»

First-person view (FPV) drone racing is a televised sport in which professional competitors pilot high-speed aircraft through a 3D circuit. Each pilot sees the environment from the perspective of their drone by means of video streamed from an onboard camera.

Reaching the level of professional pilots with an autonomous drone is challenging because the robot needs to fly at its physical limits while estimating its speed and location in the circuit exclusively from onboard sensors.

Here we introduce Swift, an autonomous system that can race physical vehicles at the level of the human world champions. The system combines deep reinforcement learning (RL) in simulation with data collected in the physical world.

Swift competed against three human champions, including the world champions of two international leagues, in real-world head-to-head races. Swift won several races against each of the human champions and demonstrated the fastest recorded race time. This work represents a milestone for mobile robotics and machine intelligence2, which may inspire the deployment of hybrid learning-based solutions in other physical systems.

«

This is why SF films that show the human pilot taking the spaceship through the crowded asteroid field is nonsense. You’d leave that stuff to the computer. This work shows that we’re already on the cusp of (or past) the “Kasparov moment”, when the machines do it better. (There’s video with this tweet.)

And: coming soon to a warzone near you!
unique link to this extract

Power laws have more power than you think • Every

Doug Shapiro:

»

The Long Tail, published in 2004, argued that because the internet dramatically lowered the cost to store and transport information goods, it would result in practically unlimited shelf space. Faced with far more choice, consumers would shift most of their consumption to the “tail,” heralding the end of mass culture and waning importance of hits. If anything, Anderson underestimated the size of the tail because he didn’t anticipate social media. The tail is not Icelandic synth pop, as it turns out, but an endless amount of user-generated content.

That the internet would yield more choice and, therefore, more fragmentation was intuitive then and is indisputable now. But it only tells half the story. Though it seems contradictory, the internet both fragments and concentrates attention. 

Understanding those dynamics matters. The contention that there are still hits may seem uncontroversial and certainly feels right intuitively. We know that when Beyonce or Taylor Swift releases an album, or the next season of Stranger Things or Game of Thrones drops, the collective attention of popular culture, much like the eye of Sauron, will be trained on it—at least until the next thing comes along. But understanding why there are still hits provides insight into whether this will persist as the supply of content keeps growing faster than demand.

The internet concentrates attention because it connects everyone in a big network. And networks are subject to powerful feedback loops. Since consumers increasingly both discover and consume content through information networks, their decisions are increasingly influenced by other people’s decisions. These feedback loops amplify the popularity of a small number of choices—hits.

«

The inevitability of the power law is so difficult to describe, but you only have to look at what has happened with the new social networks – Bluesky, Threads – to see that the big get bigger, and the minnows remain small.
unique link to this extract

OpenAI reportedly on track to generate more than $1 bln revenue over 12 months • Reuters

»

OpenAI is on track to generate more than $1bn in revenue over the next 12 months from the sale of artificial intelligence software and computing capacity that powers it, The Information reported on Tuesday.

Earlier, the ChatGPT maker projected $200m in revenue for this year.

The Microsoft-backed company is generating more than $80m in revenue per month, compared to just $28m in the entire last year, the report added.

OpenAI did not immediately respond to a Reuters request for comment.

Besides ChatGPT, it makes money by selling API access to its AI models for developers and enterprises directly and through a partnership with Microsoft, which invested over $10 billion into the company in January.

«

That would be quite the uptick, and it would be revealing to see just where that money’s coming from. My suspicion is it’s programmers using it to write frameworks.
unique link to this extract

China reaches peak gasoline in milestone for electric vehicles • Bloomberg Hyperdrive newsletter

Colin McKerracher:

»

Earlier this month, Chinese oil giant Sinopec made a surprise announcement that mostly flew under the radar. It’s now expecting gasoline demand in China to peak this year, two years earlier than its previous outlooks.

The main culprit? The surging number of electric vehicles on the road.

As I’ve written previously, calling peaks is often a no-win endeavor for industry analysts. The call will either be correct but seem obvious after the fact, or wrong and lead to years of mockery. But this isn’t an analyst calling a peak; it’s China’s largest fuel distributor. Sinopec knows the fuel business, and more importantly, it has an interest in the business remaining robust. Saying it’s all downhill from here for gasoline is quite a statement.

China has been the largest driver of global growth for refined oil products like gasoline and diesel over the last two decades. But EV adoption rates in China are now soaring, with August figures likely to show plug-in vehicles hitting 38% of new passenger-vehicle sales. That’s up from just 6% in 2020 and is starting to materially dent fuel demand.

Fuel demand in two and three-wheeled vehicles is already in structural decline, with BNEF estimating that 70% of total kilometers traveled by these vehicles already switched over to electric. Fuel demand for cars will be the next to turn, since well over 5% of the passenger-vehicle fleet is now either battery-electric or plug-in hybrid. The internal combustion vehicle fleet is also becoming more efficient due to rising fuel-economy targets.

«

I’m fascinated by the use of the word “culprit” in that second paragraph. Not using fossil fuels is a bad thing now? Just saying “reason” would have been just as accurate without offering a bizarre judgement on what’s happening.
unique link to this extract

Ukraine cannot win against Russia now, but victory by 2025 is possible • Financial Times

Richard Barrons is a former British Army general, and previous commander of Joint Forces Command:

»

Ukraine’s current counteroffensive will not throw Russia out — not that anyone expected it to. Nor is it likely to cut the occupation in half before the winter, which might have been one of the more optimistic aims. It has, however, shown how the Russian army can be beaten. Not in 2023, but in 2024 or 2025. Thus the refrain among western allies of supporting Kyiv “for as long as it takes”.

The modest progress achieved this summer shows that, while overcoming a well-prepared conventional battlefield defence may be one of the hardest operations in war, it can be done. The Ukrainian military has only breached the first line of trenches to take Robotyne in the south, having battled for weeks through minefields to get there. Progress is about eight miles with another 55 miles to go (through three lines of defences) before reaching the sea. The aim is to cut the land bridge to Crimea. To the north and south of Bakhmut, advances amount to about five miles with 10 miles to the Russian main defensive line and 60 miles to the border.

The presumed assassination of Wagner chief Yevgeny Prigozhin and the top leadership of his mercenary group has had no effect on the fighting, save perhaps for stiffening the troops’ loyalty to Vladimir Putin. Russian forces are stretched, worn out and short of reserves but unless they simply give up, this will still be a long haul.

«

Basically just checking in on this. Progress in wars is slow, and then fast; gradual, and then sudden.
unique link to this extract

Algospeak is driving us mad • New Statesman

Marie Le Conte:

»

There are moments when being online makes you feel insane. I had one of those recently, scrolling down my Instagram time line. There was a picture of a stripper wearing a jaunty little pink thong, vertiginous plastic heels and kneeling by a pole. That part was normal; I have a rich and varied social life. There’s often a lot of skin on my social media.

The madness-inducing cognitive dissonance came from the caption, in which it was explained that the dancer in question loved embracing her “seggsuality”. It made me want to throw my phone at the wall. Here was about 90% of a woman’s arse, out for all to see, yet the account felt unable to spell out a word that describes a fundamental part of adult human life.

Strip clubs are places where you can look but you cannot touch, and the internet is fast becoming a space where you can ogle but cannot spell. If “seggs” made your teeth itch, then I can only apologise for introducing you to “corn” (porn), “le dollar bean” (lesbian), “mascara” (penis) and “unalive” (suicide).

The last word here may feel like the odd one out, but it isn’t. They’re all “algospeak”; lingo developed to evade the algorithmic content filters now used by social media platforms. We’re all products of sex and we’ll all die one day but apparently we’re not allowed to talk about either.

Well, it isn’t clear what we’re allowed to talk about. Many of those phrases originated on TikTok but TikTok says it does not ban or penalise sensitive content. Do we trust it? Apparently the teens do not. Other moderation policies are equally as opaque. It isn’t always clear what they will and will not allow, so content creators self-censor just in case, as do the ones who follow them.

«

It’s an outgrowth of the Scunthorpe phenomenon (look it up if you need to) combined with these networks coming from the absurdly prudish American culture. (Requires a free login to read in full.)
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Introducing a speed limit of 80mph (130kph) would bring substantial benefits, research says – but there’s no political will for it. CC-licensed photo by John M 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 political advertising.

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

Apple’s decision to kill its CSAM photo-scanning tool sparks fresh controversy • WIRED

Lily Hay Newman:

»

This week, a new child safety group known as Heat Initiative told Apple that it is organizing a campaign to demand that the company “detect, report, and remove” child sexual abuse material (CSAM) from iCloud and offer more tools for users to report CSAM to the company. 

Today, in a rare move, Apple responded to Heat Initiative, outlining its reasons for abandoning the development of its iCloud CSAM scanning feature and instead focusing on a set of on-device tools and resources for users known collectively as Communication Safety features. The company’s response to Heat Initiative, which Apple shared with WIRED this morning, offers a rare look not just at its rationale for pivoting to Communication Safety, but at its broader views on creating mechanisms to circumvent user privacy protections, such as encryption, to monitor data. This stance is relevant to the encryption debate more broadly, especially as countries like the United Kingdom weigh passing laws that would require tech companies to be able to access user data to comply with law enforcement requests.

“Child sexual abuse material is abhorrent and we are committed to breaking the chain of coercion and influence that makes children susceptible to it,” Erik Neuenschwander, Apple’s director of user privacy and child safety, wrote in the company’s response to Heat Initiative. He added, though, that after collaborating with an array of privacy and security researchers, digital rights groups, and child safety advocates, the company concluded that it could not proceed with development of a CSAM-scanning mechanism, even one built specifically to preserve privacy.

“Scanning every user’s privately stored iCloud data would create new threat vectors for data thieves to find and exploit,” Neuenschwander wrote. “It would also inject the potential for a slippery slope of unintended consequences. Scanning for one type of content, for instance, opens the door for bulk surveillance and could create a desire to search other encrypted messaging systems across content types.”

«

Apple says it’s going to introduce scanning for CSAM: result, outrage.

Apple says it’s not going to do scanning for CSAM: result, outrage. Though I’m puzzled by Apple’s assertion that scanning for CSAM would “create new threat vectors for data thieves”. What’s the thinking there, exactly? It isn’t explained.
unique link to this extract

The real story of Elon Musk’s Twitter takeover • WSJ

Walter Isaacson:

»

[Musk, who had agreed to join Twitter’s board after buying a ton of its shares] then flew to Larry Ellison’s Hawaiian island, Lanai. He had planned the trip as a quiet rendezvous with one of the women he was occasionally dating, the Australian actress Natasha Bassett. But instead of using it as a relaxed mini-vacation, he spent his four days there figuring out what to do about Twitter.

He stayed awake most of his first night stewing about the problems Twitter faced. When he looked at a list of users who had the most followers, they were no longer very active. So at 3:32 a.m. Hawaii time, he posted a tweet: “Most of these ‘top’ accounts tweet rarely and post very little content. Is Twitter dying?”

About 90 minutes later, Twitter CEO Agrawal sent Musk a text message: “You are free to tweet ‘Is twitter dying?’ or anything else about Twitter, but it’s my responsibility to tell you that it’s not helping me make Twitter better in the current context.” It was a restrained text, carefully worded to avoid implying that Musk no longer had the right to disparage the company.

When Musk got the text, it was just after 5 a.m. in Hawaii, but he was still going strong. He shot back a scathing reply: “What did you get done this week?” It was the ultimate Musk put-down.

Then he texted back a fateful three-shot volley: “I’m not joining the board. This is a waste of time. Will make an offer to take Twitter private.”

Agrawal was shocked. “Can we talk?” he asked plaintively.

Within three minutes, Taylor, the Twitter board chair, texted Musk with a similar plea to talk. “Do you have five minutes so I can understand the context?” he asked Musk.

“Fixing Twitter by chatting with Parag won’t work,” Musk answered. “Drastic action is needed.”

«

And, well, you know what happened after that. Isaacson of course is Musk’s official biographer, though it’s not fawning. (The link may let you jump the WSJ paywall.)
unique link to this extract

With five old phones and some Pew data, the BBC’s Marianna Spring monitors social media from the inside • Nieman Journalism Lab

Sophie Culpepper:

»

Marianna Spring realizes it’s a little bizarre to carry around five old cell phones.

Spring, the BBC’s misinformation and social media correspondent, keeps them in a tote bag, and uses stickers to distinguish them. It’s challenging, among other things, to keep them all charged, she tells me in an interview. But for nearly a year, that clunky, antiquated tech has served a cutting-edge reporting purpose for the 27-year-old London-based reporter: helping her get inside the black box of social media algorithms, and get a firsthand sense of what they feed to American voters with different political and personal profiles.

Since September 2022, just before the U.S. midterm elections, Spring has maintained social media accounts that correspond to five different “voter profiles” she developed using Pew Research Center data:

• Larry, a “faith and flag” conservative, is a 71-year-old white retired insurance broker living in Oneonta, Alabama
• Britney, a 50-year-old white school secretary living in Texas, is a Populist Right voter
• Gabriela, a 44-year-old Hispanic nanny living in Florida, is a “stressed sideliner” who is not that interested in politics
• Michael, a 61-year-old Black protestant and a teacher in Milwaukee, is a “Democratic mainstay”
• Emma, a 25-year-old graphic designer who lives in New York City with her girlfriend, is a Progressive Left voter.

Each character’s accounts are confined to a single phone to avoid contaminating Spring’s findings. The reporter maintains accounts with computer-generated profile photos on a range of platforms — Facebook, Instagram, Twitter (now X), TikTok, and YouTube — and, essentially, lurks. She Likes posts, but the accounts are all private and are not “messaging people or commenting on stuff,” she said. “They’re very much passive social media users to an extent — so all I have to do is feed the algorithm: watch content, like content, follow content. But they’re not deceiving people in any way.”

«

Very smart approach by Spring; as you’d expect, her fake IDs get very different views of the world.

unique link to this extract

Car smuggling from Chile to Bolivia is booming • Rest of World

Daniela Dib:

»

TikTok is full of videos of chuteros, or car thieves, venturing into the impressive Bolivian salt flats as they avoid police raid sites. They usually do stunts to show off the cars they’ve stolen [from neighbouring Chile; 20% of cars in Bolivia are smuggled], mostly high-end SUVs or expensive sports cars, attracting praise from wannabe chuteros who ask for information on how to become one. Those interested in buying stolen cars inquire about prices of specific models in the comments. It makes for a striking feed: hundreds of TikTok videos set to the beat of “Chutero Yo Soy,” a song about the chutero lifestyle.

On the Chilean side of the border, car theft victims are whipping out their phones to try and tackle the problem themselves. Hugo Bustio, a former policeman, founded a volunteer group called Vehicle Search Group (GBV) after realizing the local police were overwhelmed by reports of stolen cars. Its members are mostly car theft victims. They patrol the streets, taking photos and sharing them on GBV WhatsApp groups as they try to find stolen cars before they’re taken to Bolivia.

GBV has even partnered with a Chilean AI startup called SafeByWolf to identify cars faster. Felipe Lobos, SafeByWolf’s founder and CEO, told Rest of World his app is able to identify licence plates on moving cars. Its algorithm is allegedly smart enough to identify if licence plates have been swapped, and it then feeds the information into the police database to find a match in the stolen car reports. The company was specifically created to deal with this issue, as the Chilean car insurance industry — Lobos’ main customer — has been increasingly disheartened by the smuggling situation in recent years.

«

At least a numberplate isn’t personally identifying information.
unique link to this extract

Broadband ISP Brsk faces full fibre anti-pole campaign in Burnley • ISPreview UK

Mark Jackson:

»

Broadband ISP and network builder Brsk, which has so far deployed their own gigabit speed Fibre-to-the-Premises (FTTP) network to 250,000 UK premises (RFS) – mostly in the Midlands (rollout plan), is facing a petition from 40 residents in Burnley who want to see their newly installed telecoms poles removed.

Like most full fibre builders, Brsk has been deploying plenty of poles (telegraph / telecoms poles), which are usually made of wood and stand around 8-9 metres high. This approach tends to be much more cost-effective and less disruptive to local residents than digging trenches for underground cables. The lower cost impact can often mean the difference between building into an area or skipping it entirely.

Brsk is supported by at least £259m of funding from Advencap and the Ares Management Corp. The operator aspires to cover 1 million homes with FTTP by 2026 – focusing on the Greater Manchester, Lancashire, West Yorkshire and the West Midlands of England.

However, poles also have a growing tendency to divide public opinion, particularly when built into an area that previously only enjoyed the benefits of underground infrastructure. Complaints often focus on their negative visual appearance, concerns about the risk of damage from major storms (example), and the lack of prior consultation.

In the past it was harder to deploy new poles, but the rules have long since been softened to aid the rollout of gigabit broadband. Today poles are built using Permitted Development (PD) rights, which means they don’t have to go through the usual planning process and can pop up quickly, often without residents getting much of a say. Operators usually only need to give the most minimal of prior notification (e.g. sticking a notice to a lamp post).

…A second online petition for the area has also been set up, which rather oddly moans that “these 10m poles, erected outside our homes and cluttering our streets with wires, are hindering our progress towards faster broadband connectivity … We must consider alternative methods that do not compromise the aesthetics of our neighbourhood while still providing us with access to advanced technologies“.

«

Mad. People are mad. This is Britain’s productivity and growth problem in a nutshell.
unique link to this extract

Worldwide smartphone shipments forecast to reach lowest volume in a decade • IDC

»

Worldwide smartphone shipments are forecast to decline 4.7% year over year in 2023 to 1.15 billion units, the lowest volume in a decade, according to the International Data Corporation (IDC) Worldwide Quarterly Mobile Phone Tracker. This is a downward revision from IDC’s previous forecast of -3.2%, driven by a weaker economic outlook and ongoing inflation, which has dampened consumer demand and lengthened refresh cycles. Despite the lower forecast for 2023, IDC expects the market to recover in 2024 with 4.5% year-over-year growth followed by growth in the low single digits through the remainder of the forecast, resulting in a five-year compound annual growth rate of 1.7%.

“Although inventory levels have normalized, the majority of OEMs remain extremely cautious in their business planning for the short term, yet again kicking the recovery can down the road.” said Nabila Popal research director with IDC’s Mobility and Consumer Device Trackers. “While the market will undoubtedly return to growth, longer refresh cycles are tapering the rate of growth over the long run, preventing the total available market from reaching pre-Covid levels.”

«

At the same time, also forecasts that iPhone sales will nudge up to their highest-ever share, at just under 20%. Nearly one in five new smartphones sold in 2023 will be an iPhone? What a stunning statistic – particularly compared to PCs, where Apple has never, post-Windows, been above 10%.

But again, most smartphones now do the job perfectly. Camera? Great. Speed? Absolutely fine. All they really need is the occasional new battery. There are more than 6bn smartphones in use worldwide; this suggests a replacement cycle on average longer than every five years.
unique link to this extract

UK government seeks expanded use of AI-based facial recognition by police • Financial Times

Anna Gross and Madhumita Murgia:

»

The move comes after privacy campaigners and independent academics criticised the technology for being inaccurate and biased, particularly against darker-skinned people. MPs have previously called for a moratorium on its use on the general population until clear laws are established by parliament.

The government is calling for submissions from companies for technologies that “can resolve identity using facial features and landmarks”, including for live facial recognition which involves screening the general public for specific individuals on police watch lists.

In particular, the Home Office is highlighting its interest in novel artificial intelligence technologies that could process facial data efficiently to identify individuals, and software that could be integrated with existing technologies deployed by the department and with CCTV cameras.

Facial recognition software has been used by South Wales Police and London’s Metropolitan Police over the past five years across multiple trials in public spaces including shopping centres, during events such as the Notting Hill Carnival and, more recently, during the coronation.

The Financial Times has previously revealed that the private owners of King’s Cross in London were using facial recognition on the general public, scanning for known troublemakers and sharing the data with the Metropolitan Police. They have since stopped using the technology.

«

I’m presently reading (for a review in The Guardian) the new book by Kashmir Hill, the NYT reporter, about how she discovered the existence of Clearview AI, which scraped the web for billions of photos and set an algorithm loose on doing facial recognition. The point in the first paragraph above is the key one: if you aren’t white, these algorithms can do a poor job – and if the police are white, the result can be even worse.

But the scary point is that last extracted paragraph: people/organisations doing it for themselves. This toothpaste isn’t returning to the tube.
unique link to this extract

The hidden link between ad blockers and better news consumption • AdGuard

Ekaterina Kachalova:

»

About 37% of Internet users now use ad blockers for some or all of their online browsing. This implies that users are not entirely happy with the trade-off between ads and content.

One question that is often ignored is how ads affect the user experience in terms of engagement, specifically whether ads discourage users from returning to sites and following the news. Many publishers simply assume that users view ads as a necessary evil and will come back no matter what. But this assumption has never been tested, so a group of researchers took pains to do just that.

The 2022 study called “How does the Adoption of Ad Blockers Affect News Consumption?” compares the news consumption habits of ad blocking and non-ad blocking users, with some curious results. In particular, the researchers found that users who started blocking ads also started reading more articles and about more topics on a news site than those who browsed it without ad blockers.

They also found that, perhaps, surprisingly, ad blockers may be a boon for news media. They discovered that ad-blocking users were more likely to return to the site, become loyal readers and hence potential subscribers. This way they could generate more revenue for publishers in the long run if the latter decide to move to a subscription-based model.

«

Obviously, AdGuard has an interest in boosting this article, but the research is by legitimate academics, and they don’t seem to have received funding from any makers of adblocking software.
unique link to this extract

Autobahn speed limit would cut carbon and bring €1bn in benefits, study says • Carbon Brief

Molly Lempriere:

»

Bringing in a speed limit of 130kph (80mph) across all of Germany’s motorways could result in nearly €1bn in “welfare” savings, a new study says, with “avoided warming” leading the benefits. 

While more than 96% of German roads are subject to a permanent speed limit, just 30% of the 13,000km network of motorways – or autobahn – has permanent or temporary speed limits, according to the German Association of the Automotive Industry (VDA). 

Germany is one of the only major countries in the world not to have a nationwide speed limit, a topic which has garnered increasing attention as the country struggles to meet its transport decarbonisation targets. 

But, according to a new paper published in the Ecological Economics journal, bringing in a 130kph speed limit on autobahns could bring about nearly €300m in welfare savings from avoided carbon dioxide (CO2) emissions, along with a host of other benefits. 

A speed limit enjoys majority public support and there is no “rational argument” against bringing it in, yet the measure is unlikely due to opposition from certain political parties, researchers tell Carbon Brief.

Calls for the introduction of an autobahn speed limit, on the basis of environmental benefits, have been growing since the introduction of the country’s first national climate law in 2019. 

«

To misquote Douglas Adams, people love speed limits – they make a wonderful whooshing sound as you go past them. It’s odd that the political will isn’t there, though, if it has public support. What would the speeders do – protest by having a go-slow? (As a reminder, in the 1976 oil crisis, when the world price of oil quadrupled in three months, Jimmy Carter instituted a 56mph (90km/h) speed limit on American motorways, producing enormous fuel efficiency benefits.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

The arrival of weight loss drug Ozempic poses an existential challenge for companies reliant on selling diets and weight loss plans. CC-licensed photo by Chemist4U 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.

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

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

I tracked an NYC subway rider’s movements with an MTA ‘feature’ • 404 Media

Joseph Cox:

»

In the mid-afternoon one Saturday earlier this month, the target got on the New York subway. I knew what station they entered the subway at and at what specific time. They then entered another station a few hours later. If I had kept monitoring this person, I would have figured out the subway station they often start a journey at, which is near where they live. I would also know what specific time this person may go to the subway each day.

During all this monitoring, I wasn’t anywhere near the rider. I didn’t even need to see them with my own eyes. Instead, I was sitting inside an apartment, following their movements through a feature on a Metropolitan Transportation Authority (MTA) website, which runs the New York City subway system.

With their consent, I had entered the rider’s credit card information—data that is often easy to buy from criminal marketplaces, or which might be trivial for an abusive partner to obtain—and punched that into the MTA site for OMNY, the subway’s contactless payments system. After a few seconds, the site churned out the rider’s travel history for the past seven days, no other verification required.

The news presents a significant privacy risk from a feature that is supposedly designed for individuals to check their own travel history, but which in reality is wide open to abuse.

…“Obviously this is a great fit for abusers who live with their victims or have physical access, however brief, to their wallets,” Eva Galperin, the director of cybersecurity at activist organization the Electronic Frontier Foundation (EFF) and who has extensively researched how abusive partners use technology, told 404 Media. “​​Credit card info is not a goddamn unique identifier.”

«

Galperin makes a reasonable point: the extreme situation is very undesirable. Adding a username/password barrier would make sense (though in the case of people living with their abusers, the latter will just demand the username and password).
unique link to this extract

The $76bn diet industry asks: what to do about Ozempic? • WSJ

Andrea Petersen, Rolfe Winkler and Sara Ashley O’Brien:

»

Annick Lenoir-Peek, a lawyer from Durham, N.C., has struggled with her weight since adolescence. She has tried Atkins and keto and spent thousands of dollars over decades on weight-loss efforts and programs such as Noom, Nutrisystem and WeightWatchers.

Since starting [weight loss drug] Ozempic in late November, she has lost around 30lb [14kg]. Her cholesterol and glucose levels have improved, and she can eat far fewer calories without feeling hungry, she says. She has felt few side effects and has more energy than when she tried calorie-restricted diets. Currently on a trip through Eastern Europe, she says she is doing more tours than she would have at a higher weight. 

People such as Ms. Lenoir-Peek—among the diet business’s most reliable customers—are sparking an existential crisis for the industry, which rang up $76bn in sales in 2022 from weight loss and medical programs, diet soda and low-calorie frozen food, gym memberships and other categories, according to research firm Marketdata LLC. 

Drugs such as Ozempic, Wegovy and Mounjaro have upended the business of losing weight in America. They are shaping up to be blockbusters for Novo Nordisk, which makes Ozempic and Wegovy, and Eli Lilly, which makes Mounjaro. The drugs are also ripping up long-held beliefs that diet, exercise and willpower are the way to weight loss. 

“I think they [the new drugs] are going to transform the industry of weight loss in a pretty big way,” says Alex Fuhrman, senior research analyst at Craig-Hallum Capital Group LLC. “The traditional approach to weight loss has been the only game in town for a very long time except for more drastic surgical interventions. The behavioural approach to weight loss is going to be under pressure now.”

«

This was always going to be the collateral damage – or existential risk – from Ozempic and other GLP-1 drugs: why do you need to torture yourself with diets when you could just take this completely safe medication?
unique link to this extract

Researchers discover common origin behind major childhood allergies • MedicalXpress

»

Several major childhood allergies may all stem from the community of bacteria living in our gut, according to a new study led by researchers at the University of British Columbia and BC Children’s Hospital.

The research, published in Nature Communications, identifies gut microbiome features and early life influences that are associated with children developing any of four common allergies—eczema, asthma, food allergy and/or hay fever. The findings could lead to methods of predicting whether a child will develop allergies, and ways to prevent them from developing at all.

“We’re seeing more and more children and families seeking help at the emergency department due to allergies,” said Dr. Stuart Turvey, professor in the department of pediatrics at UBC and an investigator at BC Children’s Hospital Research Institute, and co-senior author on the study. “Hundreds of millions of children worldwide suffer from allergies, including one in three children in Canada, and it’s important to understand why this is happening and how it can be prevented.”

The study is one of the first to examine four distinct school-aged pediatric allergies at once. While these allergic diseases each have unique symptoms, the Turvey lab was curious whether they might have a common origin linked to the infant gut microbiota composition.

“These are technically different diagnoses, each with their own list of symptoms, so most researchers tend to study them individually,” says Dr. Charisse Petersen, co-senior author on the paper and postdoctoral fellow in the Turvey lab. “But when you look at what is going wrong at a cellular level, they actually have a lot in common.”

«

Very interesting: a shift in the microbiome could explain the apparent rise in allergic incidence. Bacteria as the underlying triggers of disease and the microbiome are the new frontiers in medicine; so much coming out now.
unique link to this extract

The Mystery of the Bloomfield Bridge • Tyler Vigen

Tyler got intrigued by something one day:

»

Why is this bridge here?

This pedestrian bridge crosses I-494 just west of the Minneapolis Airport. It connects Bloomington to Richfield. I drive under it often and I wondered: why is it there? It’s not in an area that is particularly walkable, and it doesn’t connect any establishments that obviously need to be connected. So why was it built?

I often have curious thoughts like this, but I dismiss most of them because if I answered all of them I would get nothing else done. But one day I was walking out of a Taco Bell and found myself at the base of the bridge. That only raised MORE questions! Why did the bridge just lead to some grass? Why isn’t there a sidewalk? What is the point? It makes no sense!

«

It’s a fabulous tale, with a bit of the shaggy dog about it. (TL;DR it does actually bridge the road.)
unique link to this extract

Google to sell maps data to companies building solar products • CNBC

Jennifer Elias:

»

The company plans to sell access to new APIs (application programming interfaces) with solar and energy information and air quality, according to materials viewed by CNBC.

Among the new offerings will be a Solar API, which could be used by solar installers like SunRun and Tesla Energy and solar design companies like Aurora Solar, according to a list of example customers viewed by CNBC. Google also sees customer opportunities with real estate companies like Zillow, Redfin, hospitality companies like Marriott Bonvoy, and utilities like PG&E.

Some of the data from the Solar API will come from a consumer-focused pilot called Project Sunroof, a solar savings calculator that originally launched in 2015. The program allows users to enter their address and to receive estimated solar costs such as electric bill savings and the size of the solar installation they’ll need. It also offers 3D modeling of the roofs of buildings and nearby trees based on Google Maps data. 

Google plans to sell API access to individual building data, as well as aggregated data for all buildings in a particular city or county, one document states. The company says it has data for over 350m buildings, according to documents, up significantly from the 60m buildings it cited for Project Sunroof in 2017.

One internal document estimates the company’s solar APIs will generate revenue between $90m and $100m in the first year after launch. There’s also a potential to connect with Google Cloud products down the line, documents state.

«

As Elias points out, the revenue will be tasty for Google, which wants to make more from maps (which remain expensive to maintain).
unique link to this extract

The SynthID watermark from Google DeepMind can detect AI-generated images • The Verge

David Pierce:

»

[Google DeepMind CEO, Demis] Hassabis and his team have been working on a tool for the last few years, which Google is releasing publicly today. It’s called SynthID, and it’s designed to essentially watermark an AI-generated image in a way that is imperceptible to the human eye but easily caught by a dedicated AI detection tool. 

The watermark is embedded in the pixels of the image, but Hassabis says it doesn’t alter the image itself in any noticeable way. “It doesn’t change the image, the quality of the image, or the experience of it,” he says. “But it’s robust to various transformations — cropping, resizing, all of the things that you might do to try and get around normal, traditional, simple watermarks.” As SynthID’s underlying models improve, Hassabis says, the watermark will be even less perceptible to humans but even more easily detected by DeepMind’s tools.

That’s as technical as Hassabis and Google DeepMind want to be for now. Even the launch blog post is sparse on details because SynthID is still a new system. “The more you reveal about the way it works, the easier it’ll be for hackers and nefarious entities to get around it,” Hassabis says. SynthID is rolling out first in a Google-centric way: Google Cloud customers who use the company’s Vertex AI platform and the Imagen image generator will be able to embed and detect the watermark. As the system gets more real-world testing, Hassabis hopes it’ll get better. Then Google will be able to use it in more places, share more about how it works, and get even more data on how it works.

«

So this is essentially steganography – hiding a message in an image. That usually gets broken (ie, the message is lost) through other conversions. And the problem is also that just because an image doesn’t contain this sikrit c0d3 doesn’t mean it isn’t AI-generated – just that SynthID (probably) wasn’t used on it.
unique link to this extract

We analyzed millions of ChatGPT user sessions: visits are down 29% since May; programming assistance is 30% of use • SparkToro

Rand Fishkin:

»

Programming is the largest use case, with 29.14% of all prompt series falling into this use-case. It’s also the clearest/least ambiguous. I hand-checked 100+ of each prompt series (an arduous, but fascinating task) to confirm the classifier’s accuracy, and programming help (with writing specific bits of code, formatting code, catching errors in code, and more) was present in every one ChatGPT marked as such.

As others have often pointed out, the tool excels at programming-related tasks. Little wonder it’s such a popular use case.

Next up is education — but not just primary or secondary education. Personal knowledge or interest pursuits and professional knowledge for work purposes are both included here as well. Same with content creation — some is clearly personal (D&D dungeon masters needing riddles or quests for their adventures was a recurring favorite in the dataset) while others are professional (“write me a 500 word blog post about detroit plumbing problems” – presumably a content marketer tired of writing their own material).

Sales and marketing use-cases overlap with content creation, but I chose to keep these separate to help see only those sessions that could only be classified as helping sales+marketing professionals with their tasks (analysis of analytics, questions about which channels to promote their products in, ad optimization tasks, and even messaging/promotion help were all in the dataset).

«

“SEO” appeared in just over 2% of prompts, suggesting that it’s already being used to game web search. When programming stops being the top use, I think we’ll be able to say that ChatGPT has properly arrived.
unique link to this extract

Fukushima: China accused of hypocrisy over its own release of wastewater from nuclear plants • The Guardian

Amy Hawkins and Justin McCurry:

»

As China bans all seafood from Japan after the discharge of 1m tonnes of radioactive water from the ruined Fukushima Daiichi nuclear power plant into the Pacific Ocean, Beijing has been accused of hypocrisy and of using the incident to whip up anti-Japanese sentiment.

Scientists have pointed out that China’s own nuclear power plants release wastewater with higher levels of tritium than that found in Fukushima’s discharge, and that the levels are all within boundaries not considered to be harmful to human health.

On Thursday, Tokyo Electric Power (Tepco), the company that manages the plant, began pumping water containing radioactive tritium into the sea, starting a wastewater discharge process that is expected to take at least 30 years. The plan has been approved by the UN’s atomic watchdog, the International Atomic Energy Agency (IAEA) and the Japanese government.

On Friday, Tepco said rapid tests of samples taken from the first batch of released wastewater showed radioactivity levels well within safe limits. “We confirmed that the analysed value is equal to the calculated concentration and that the analysed value is below 1,500 becquerels per litre,” a Tepco spokesperson, Keisuke Matsuo, told reporters. The national safety standard is 60,000 becquerels a litre.

“We will continue to conduct analysis every day over the next one month and even after that, maintain our analysis effort,” Matsuo added. “By providing swift, easy to understand explanations we hope to dispel various concerns.”

China has condemned the discharge, with the customs agency saying it risks the “radioactive contamination of food safety”. China’s foreign ministry said it was an “extremely selfish and irresponsible act”.

«

Good that someone is pointing out that China is simply playing a PR game here. People are Losing Their Minds over the Fukushima water discharge, which I’d say is safer to drink than many – most? – English rivers. The radioactivity levels are absolutely infinitesimal, but people hear that word and starting doing Münch painting impressions.
unique link to this extract

Most cancer screenings don’t extend life, study finds – but don’t cancel that appointment • CNN

Jen Christensen:

»

Early diagnosis has been shown to improve cancer outcomes, which is why the American Cancer Society and the World Health Organization say routine screening is an important public health strategy. Overall cancer mortality worldwide has decreased significantly, falling 33% since 1991, in part due to early detection as well as advances in treatment and declines in smoking.

The latest study, published Monday in the journal JAMA Internal Medicine, found that of the six most common cancer screenings, only colorectal cancer screening with sigmoidoscopy — in which doctors check the lower part of the colon or large intestine for cancer — seemed to make a difference in extending someone’s life. It may extend life by a little more than three months, the research says.

The researchers looked at clinical trials that involved at least nine years of follow-up reporting and found no significant difference in lifetime gain with the other most common cancer screening tests: mammography for breast cancer, colonoscopy, fecal occult blood testing or endoscopy (FOBT), prostate-specific antigen tests, and computed tomography for current or former smokers.

“We do not advocate that all screening should be abandoned,” the researchers wrote. “Screening tests with a positive-benefit-harm balance measured in incidence and mortality of the target cancer compared with harms and burden may well be worthwhile.”

The authors of the new research suggest that rather than emphasize that cancer screenings save lives, doctors should be clearer about their absolute benefits, harms and burdens.

«

The argument against is that inevitably there are false positives, which lead to anxiety and more screening. Compared to cancer, though, anxiety and more screening seems a small price. (Plus what about the upside of relief if the false positive is discovered?)
unique link to this extract

Managed to death: how Canada turned its forests into a carbon bomb • Bulletin of the Atomic Scientists

Jessica McKenzie:

»

To illustrate the scale and pace of our metastasizing forest carbon crisis, I turned to data in Canada’s official national greenhouse gas inventory, plus recent wildfire data from the European Union’s Earth Observation Program. The resulting chart shows the cumulative amount of CO2 that’s been added to the atmosphere from Canada’s managed forest since 1990.

The falling green line at the start of the chart shows that in the early 1990s, the forest was a valuable carbon sink, helping to slow global warming. Back then, new forest growth absorbed more CO2 from the air than was emitted by logging, wildfire and decay.

That all changed after 2001, the tipping point year for Canada’s managed forest.

As the rising red line on the chart shows, since that year, the forest has emitted more CO2 than it has absorbed. A lot more. Logging, wildfires, insects and the many forms of decay are now turning trees into CO2 faster than the forest can grow back.

That pumped billions of tonnes of climate fuel into the atmosphere—even before accounting for this year’s epic wildfires (shown by the dashed line). With those included, the cumulative total since the tipping point year is now around 3,700 million tonnes of CO2 (MtCO2).

«

unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Some ebooks about mushroom foraging on Amazon seem to have been written by a chatbot – and so could kill their readers. CC-licensed photo by Stephen Bowler 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 about 0845 UK time. Free signup.

A selection of 9 links for you. Back in the office? I’m @charlesarthur on Twitter. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.

‘Life or death:’ AI-generated mushroom foraging books are all over Amazon • 404 Media

Samantha Cole:

»

A genre of AI-generated books on Amazon is scaring foragers and mycologists: cookbooks and identification guides for mushrooms aimed at beginners.

Amazon has an AI-generated books problem that’s been documented by journalists for months. Many of these books are obviously gibberish designed to make money. But experts say that AI-generated foraging books, specifically, could actually kill people if they eat the wrong mushroom because a guidebook written by an AI prompt said it was safe.

The New York Mycological Society (NYMS) warned on social media that the proliferation of AI-generated foraging books could “mean life or death.”

“There are hundreds of poisonous fungi in North America and several that are deadly,” Sigrid Jakob, president of the New York Mycological Society, told me in an email. “They can look similar to popular edible species. A poor description in a book can mislead someone to eat a poisonous mushroom.”

A quick scan of Amazon’s mushroom and foraging books revealed a bunch of books likely written by ChatGPT, but are sold without any indication that they’re AI-generated and are marketed as having been written by a human when they’re very likely not.

“Edwin J. Smith” is the author listed on two books—The Ultimate Mushroom Books Field Guide Of The Southwest: An essential field guide to foraging edible and non-edible mushrooms outdoors and indoors and Psilocybin Mushroom Book: Field Guide To Identification, Growing, and Microdosing Psilocybin Mushroom for Safe Use and Health Remedies—but doesn’t have any other books, or an online presence otherwise. The only Edwin J. Smith I could find was a Professor Emeritus of medicine at Indiana University from a staff list that’s more than a decade old.

…After 404 Media reached out for comment and sent the company links to these suspected AI books, Amazon deleted The Ultimate Mushroom Books Field Guide Of The Southwest, Psilocybin Mushroom Book, and WILD MUSHROOM COOKBOOK FOR BEGINNER.

«

Once again, journalists get the job of saving people from the bad decisions made by platforms – in this case, having no gatekeeping or quality function because it’s cheaper (= more profitable) not to.
unique link to this extract

The AI revolution is coming, but not as fast as some people think • The New York Times

Steve Lohr:

»

Lori Beer, the global chief information officer of JPMorgan Chase, talks about the latest artificial intelligence with the enthusiasm of a convert. She refers to AI chatbots like ChatGPT, with its ability to produce everything from poetry to computer programs, as “transformative” and a “paradigm shift.”

But it’s not coming soon to the nation’s largest bank. JPMorgan has blocked access to ChatGPT from its computers and told its 300,000 workers not to put any bank information into the chatbot or other generative AI tools.

For now, Ms. Beer said, there are too many risks of leaking confidential data, questions about how the data is used and about the accuracy of the AI-generated answers. The bank has created a walled-off, private network to allow a few hundred data scientists and engineers to experiment with the technology. They are exploring uses like automating and improving tech support and software development.

Across corporate America, the perspective is much the same. Generative AI, the software engine behind ChatGPT, is seen as an exciting new wave of technology. But companies in every industry are mainly trying out the technology and thinking through the economics. Widespread use of it at many companies could be years away.

Generative AI, according to forecasts, could sharply boost productivity and add trillions of dollars to the global economy. Yet the lesson of history, from steam power to the internet, is that there is a lengthy lag between the arrival of major new technology and its broad adoption — which is what transforms industries and helps fuel the economy.

Take the internet. In the 1990s, there were confident predictions that the internet and the web would disrupt the retailing, advertising and media industries. Those predictions proved to be true, but that was more than a decade later, well after the dot-com bubble had burst.

«

Unlike the web3/crypto argument that “it’s only useless because it’s like the early internet!”, I think this argument holds water.
unique link to this extract

The cheap radio hack that disrupted Poland’s railway system • WIRED

Andy Greenberg:

»

On Friday and Saturday, August 25 and 26, more than 20 of Poland’s trains carrying both freight and passengers were brought to a halt across the country through what Polish media and the BBC have described as a “cyberattack.” Polish intelligence services are investigating the sabotage incidents, which appear to have been carried out in support of Russia. The saboteurs reportedly interspersed the commands they used to stop the trains with the Russian national anthem and parts of a speech by Russian president Vladimir Putin.

Poland’s railway system has served as a key resource in the facilitating of Western weapons and other aid into Ukraine as NATO attempts to bolster the country’s defense against Russia’s invasion. “We know that for some months there have been attempts to destabilize the Polish state,” Stanislaw Zaryn, a senior security official, told the Polish Press Agency. “For the moment, we are ruling nothing out.”

But as disruptive as the railway sabotage has been, on closer inspection, the “cyberattack” doesn’t seem to have involved any cyber at all, according to Lukasz Olejnik, a Polish-speaking independent cybersecurity researcher and consultant, and the author of the forthcoming book Philosophy of Cybersecurity. In fact, the saboteurs appear to have sent simple “radio-stop” commands via radio frequency to the trains they targeted. Because the trains use a radio system that lacks encryption or authentication for those commands, Olejnik says, anyone with as little as $30 of off-the-shelf radio equipment can broadcast the command to a Polish train—sending a series of three acoustic tones at a 150.100 megahertz frequency—and trigger their emergency stop function.

“It is three tonal messages sent consecutively. Once the radio equipment receives it, the locomotive goes to a halt,” Olejnik says, pointing to a document outlining trains’ different technical standards in the European Union that describes the radio-stop command used in the Polish system.

«

Seems like a bit of an oversight when you have a potential aggressor just over your border.
unique link to this extract

South Africa’s Competition Commission takes aim at Google • Rest of World

Damilare Dosunmu:

»

According to the commission’s report, Google’s model disadvantages new, small, and underfunded South African businesses. The report indicates that the US tech giant has not only become direct competition to its clients by offering some of the services they render (including shopping and travel), but it is also playing unfairly by prioritizing its offerings over theirs. 

The commission released a set of remediations for Google to follow, including providing a South African badge and search filter to enable consumers to find and identify local platforms quickly. It also asked Google to introduce a new feature that displays smaller South African platforms relevant to consumer search, especially in travel and shopping. Google also has to provide support programs worth 330m rand ($17.6m) over five years, with 180m rand ($9.6m) of that investment going into advertising credits. The remaining amount goes to technical training, credits for other Google products, startup training, and networking, among other things.

Industry experts and analysts were divided on whether Google’s power over the digital economy needed more regulation or if the watchdog’s move was anti-market. They told Rest of World the commission’s report might begin a decade-long regulatory back-and-forth with the tech giant.

«

unique link to this extract

Meta rejects recommendation to suspend former Cambodian prime minister • The Hill

Rebecca Klar:

»

Meta, the parent company of Facebook, rejected a recommendation from its Oversight Board to suspend the account of former Cambodian Prime Minister Hun Sen, according to a decision announced Tuesday. 

Meta said it would not be suspending Hun Sen’s Facebook or Instagram page after determining that doing so would “not be consistent with our policies, including our protocol on restricting accounts of public figures during civil unrest.” 

Hun Sen, who handed power to his son Hun Manet after July’s national election, had preemptively removed his Facebook page after the Oversight Board recommendation in June, and banished Facebook representatives from operating in the country.

However, he returned to the platform three weeks later, after briefly using Telegram as his main tool of public communication. His Cambodian People’s Party won a large majority in the National Assembly after the main opposition party was barred from competing.

«

Thus demonstrating that the Oversight Board is a pointless exercise in pencil-sharpening. It’s not even a figleaf now. May as well dissolve it and give the money to charity.
unique link to this extract

As Canada burns and California floods, Facebook and Twitter are MIA • The Washington Post

Will Oremus:

»

As wildfires ravage western Canada, Canadians can’t read the news about them on Facebook or Instagram. This month, Facebook parent company Meta blocked links to news organizations on its major social networks in Canada to protest a law that would require it to pay publishers for distributing their content.

As a freak tropical storm flooded swaths of Southern California over the weekend, residents and government agencies who turned to X, formerly known as Twitter, for real-time updates struggled to discern fact from fiction. That has gotten far more difficult, officials say, since Elon Musk jumbled the site’s verification policies, removing the blue check marks from verified journalists and media outlets — instead granting them to anyone who pays a monthly fee.

Facebook and Twitter spent years making themselves essential conduits for news. Now that government agencies, the media and hundreds of millions of people have come to rely on them for critical information in times of crisis, the social media giants have decided they’re not so invested in the news after all.

Tech titans Mark Zuckerberg and Musk may not agree on much. But both have pulled back, in different ways, from what their companies once saw as a responsibility, to both their users and society, to connect people with reliable sources of information. A drumbeat of natural disasters, probably intensified by climate change, is highlighting the consequences of that retrenchment.

“Just a few years ago, Twitter was a really valuable way for us to communicate with the public,” said Brian Ferguson, deputy director of crisis communications for the California Governor’s Office of Emergency Services. “It’s much more challenging now because of some of the changes that have happened.”

«

These two things aren’t quite like each other, though. Twitter is simply abrogating the responsibility it used to feel as the place where the world found the news – which Musk is still claiming was his desire for it. Meta, on the other hand, has never truly pretended to be the place where you find the news about the world – and the Canadian government wrote a terrible law which offered Meta a simple get-out by not including links. Bad laws that you can evade are going to be evaded.
unique link to this extract

Why note-taking apps don’t make us smarter • The Verge

Casey Newton:

»

Earlier this year, like many productivity tools, Notion added a handful of AI features. I use two of them in my links database. One extracts the names of any companies mentioned in an article, creating a kind of automatic tagging system. The other provides a two- or three-sentence summary of the article I’m saving.

Neither of these, in practice, is particularly useful. Tags might theoretically be useful for revisiting old material, but databases are not designed to be browsed. And while we publish summaries of news articles in each edition of Platformer, we wouldn’t use AI-written summaries: among other reasons, they often miss important details and context.

At the same time, the database contains nearly three years of links to every subject I cover here, along with the complete text of thousands of articles. It is here, and not in a note-taking app, that knowledge of my beat has been accreting over the past few years. If only I could access that knowledge in some way that went beyond my memory.

It’s here that AI should be able to help. Within some reasonable period of time, I expect that I will be able to talk to my Notion database as if it’s ChatGPT. If I could, I imagine I would talk to it all the time.

Much of journalism simply involves remembering relevant events from the past. An AI-powered link database has a perfect memory; all it’s missing is a usable chat interface. If it had one, it might be a perfect research assistant.

«

Like Newton, I’ve tried a lot of note-taking apps, and concluded they’re never going to replace the connections you make in your head. For assembling lots of information, I like Scrivener (used it to write all three of my books), but it didn’t make my thinking any clearer.
unique link to this extract

Elon Musk to remove headlines from news articles shared on X • Fortune

Kylie Robison:

»

X, the social media platform previously known as Twitter, is planning a major change in how news articles appear on the service, stripping out the headline and other text so that tweets with links display only an article’s lead image, according to material viewed by Fortune.

Roughly four hours after the publication of this article, Elon Musk confirmed these plans, posting that “this is coming from me directly,” and it “will greatly improve the esthetics.”

The change means that anyone sharing a link on X—from individual users to publishers—would need to manually add their own text alongside the links they share on the service; otherwise the tweet will display only an image with no context other than an overlay of the URL. While clicking on the image will still lead to the full article on the publisher’s website, the change could have major implications for publishers who rely on social media to drive traffic to their sites as well as for advertisers.

According to a source with knowledge of the matter, the change is indeed being pushed directly by X owner Elon Musk. The primary objective appears to be to reduce the height of tweets, thus allowing more posts to fit within the portion of the timeline that appears on screen. Musk also believes the change will help curb clickbait, the source said.

“It’s something Elon wants. They were running it by advertisers, who didn’t like it, but it’s happening,” the source said, adding that Musk thinks articles occupy excessive space on the timeline.

«

You can imagine how this panned out: Musk scrolling and declaring that “all these articles [which he doesn’t read – he doesn’t strike me as a person consumed by curiosity] take too much space. How many more ads could we show if we cut these? That many, huh? OK let’s do that.” It’s not intended for users. It’s him and his personal plaything; there’s no sense of having any responsibility to all the other users of Twitter, who may derive value from the headlines and text.
unique link to this extract

Apple’s September iPhone event: how to watch and what to expect • The Verge

Emma Roth:

»

One of the biggest changes coming to this year’s iPhone lineup is the addition of USB-C. For the first time, the iPhone 15 is expected to come with the widely used port instead of Apple’s proprietary Lightning connector. This doesn’t come as a surprise, as Apple confirmed last year that it would make the change to USB-C to comply with the European Union’s incoming regulations.

Although reports indicate that all phones in the iPhone 15 lineup will get the USB-C port, Apple analyst Ming-Chi Kuo says only the Pro and Pro Max will benefit from higher data transfer rates. Both premium models will come with “at least” USB 3.2 or Thunderbolt 3, while the base iPhone 15 and 15 Plus will support USB 2.0, according to Kuo. Either way, 9to5Mac reports that all iPhone 15 models should have faster 35W charging rates thanks to the switch.

Additionally, supply chain analyst Ross Young reported last year that all models of the iPhone 15 will come with the Dynamic Island. That’s a change from what Apple currently offers, as it only includes the pill-shaped cutout on the iPhone 14 Pro and Pro Max. This time around, the most significant changes coming to the iPhone 15 Pro and Pro Max are reported to be a titanium frame, thinner bezels, and a potentially more expensive price.

Both premium models may also come with an action button similar to the one on the Apple Watch Ultra, as reported by MacRumors and 9to5Mac. This button is supposed to replace the mute toggle and could allow you to assign shortcuts to different apps and utilities. And while early rumors suggested that Apple will replace the volume rocker and the power button on the iPhone 15 Pro and Pro Max with solid-state toggles, recent reports from Kuo and leaker Unknownz21 suggest they’ll stay the same for now.

«

The event is on Tuesday September 12. Also coming: faintly updated Apple Watches. And there you go – enough to drive hundreds of billions of dollars of revenue. But the days when iPhone reveals were a) dramatic surprises and b) actually real news events are long past.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified