The latest service in programming is “slopfixing”, where the confused code written by chatbots gets tidied up.. by humans. CC-licensed photo by Austin Gruenweller on Flickr.

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

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

AI altering meaning of users’ drafts on issues from abortion to climate, study finds • The Guardian

Robert Booth:

»

AI tools are twisting online messages on sensitive political topics about everything from abortion to climate change in ways that could snowball to reshape long-term public opinion, experts have said.

As tech companies push AI tools as convenient ways to redraft and summarise the massive influx of daily messages, many inject their own political biases – some leaning distinctly rightwing, others more liberal, according to a study from Oxford and Potsdam universities.

AI drafting tools completely reversed the meaning of draft posts on atheism, including in one test switching a claim that Jesus wasn’t real to “Jesus … was real”. They also changed a post complaining of “#climatechangehoax” to “#ClimateAction”.

Academics from the Oxford Internet Institute and the Hasso Plattner Institute examined the behaviour of mainstream large language models provided by Elon Musk’s xAI, Meta, Google, China’s Alibaba and France’s Mistral and found the introduction of bias happens even when the AI tool is instructed to preserve the original sense.

They also found small nudges in the meaning of draft messages could be amplified across millions of interactions to create long-term public opinion shifts greater than the bias introduced by the AI system. They said the issue was not yet being tackled by regulations such as the EU AI Act or the Digital Services Act, creating a “severe accountability gap”.

Fears about online bias have previously focused on how algorithms pigeonhole users in “filter bubbles”. But the rising appeal to time-poor consumers of AI writing tools and text summarisers – like the Grok-powered “explain this” function now embedded with every post on X – presents a new risk to trustworthy human-to-human communication, the study suggests

«

Strange but perhaps not surprising? If you’re writing something a bit contrary compared to what the whole of the internet says, then it’s maybe going to fight back against it a little.
unique link to this extract

Data centres’ energy demand threatens Trump’s “Made in America” plan • Ars Technica

Jeremy Hsu:

»

US manufacturers in many Rust Belt cities and towns are paying significantly higher electricity costs as growing energy demand from data centres strains the largest power grid operator in the United States. The resulting squeeze on profit margins for steelmakers and brick factories could further undermine President Donald Trump’s “Made in America” plan to revive US manufacturing, and it comes as Trump has simultaneously championed the tech companies behind the AI data center boom.

Factory electricity bills are generally rising faster than those for other business customers or residential customers, according to a Reuters analysis. It highlighted the example of the Belden Brick Company, a 141-year-old brick manufacturer in Ohio, whose electricity bills have soared from $1,600 to $12,000 per month due to a higher monthly capacity charge in the 13-state region served by the grid operator PJM Interconnection.

Meanwhile, the Steel Manufacturers Association warned that US steel companies concentrated in the Rust Belt region served by PJM Interconnection are paying tens of millions of dollars in higher power costs per year. Electricity accounts for 20% to 40% of the total production costs of making steel.

Each electric arc furnace used in steelmaking has an operating power load between 40 and 200 megawatts, and the entire US steel industry draws up to 11 gigawatts of power at peak production across all facilities.

US steelmakers have benefited from data center construction’s requirements for an estimated 1 million tons of steel per year. But data centre energy demand has also driven up operating costs for the US steel industry, according to the Wall Street Journal.

«

unique link to this extract

Slopfix — we refactor vibecoded codebases

»

Slopfix is a team of three senior engineers who refactor vibecoded codebases back to maintainability.

Did that happen to you? You have an AI-generated codebase that works, but adding a feature now takes days and breaks two other things? This happens to every vibecoded project past a certain size: the agent stops seeing the whole picture and starts duplicating code instead of finding it. We can help.

What we do? We analyze your codebase for free. If we don’t think we can help, we say so and you’ve lost nothing. If we can, you get a fixed price and a committed reduction target — something like “100,000 lines down to 35,000, same functionality.”

Then we do one week of focused work. Before touching anything, we sit down with you and write out exactly what your app does, screen by screen, endpoint by endpoint. That checklist is our safety net and yours. Then we cut: the fourteen date formatters become one, the hand-rolled framework becomes a library, the duplicated logic collapses. Where the code is beyond saving, we distil what it does and rebuild that part clean.

You keep everything. The smaller codebase, the QA checklist, and a set of guardrails — a CLAUDE.md, lint rules, CI checks — that slow the slop down when you go back to building. Plus two weeks of warranty: if we broke something that worked before, we fix it free.

Yes, we use Claude Code too. On a very short leash. The difference is thirty years of combined experience about what maintainable code looks like, and the agent doesn’t get a vote.

«

“Refactoring”, for those not familiar with coding, is noting where a program uses the same logic or sequence repeatedly in separate places, and rewriting the code so that piece of logic is only written once, and the program uses it whenever needed. This cuts the potential for bugs.

Slopfixing might be the next wave in human programming.
unique link to this extract

Big tech’s lofty climate goals wrecked by energy-hungry AI • The Guardian

Blake Montgomery, Nick Robins-Early and Dara Kerr:

»

Google and Amazon’s net-zero climate goals are slipping out of reach because of their massive investments in artificial intelligence.

Both companies released their annual sustainability reports last week, which showed soaring emissions from the AI datacenter boom, driven by the construction of new datacenters, fuel used for deliveries, and expanding electricity usage. Google’s total carbon emissions climbed 25% year-over-year, and Amazon’s shot up 16%.

While Google touted its sustainability progress, it conceded that combating the energy and water-hungry needs of AI created a dilemma.

“The environmental footprint of the data centers that power AI is growing, creating a dual challenge: managing that environmental footprint while simultaneously building infrastructure to meet growing demand and realize Al’s full potential,” reads Google’s report.

Similarly, Amazon said: “We recognize that the path to being a more sustainable company is not a straight line. Though our emissions increased in 2025, we remain steadfast in our commitment to sustainability.”

Google, Amazon and Microsoft, which will release its sustainability report in the coming weeks, fashioned themselves climate leaders in the tech sector in the previous decade, when investors prioritized ESG (environmental, social, and governance). Each set ambitious net-zero carbon goals and has heavily invested in sustainable energy projects like wind and solar.

However, as these companies fight to win the AI arms race, their stock prices depend on their ability to integrate AI, which needs energy. Their desire for huge amounts of power has outstripped their commitments to sustainability, so their emissions promises have softened.

Microsoft’s 2025 sustainability report documented a 23% increase in emissions compared with a 2020 baseline. The company’s investment in AI infrastructure has expanded since last year’s report, so 2026’s disclosures seem likely to reveal a similar or even larger spike. Meta’s 2025 sustainability report showed that emissions jumped 64% year-over-year in spite of a pledge of net-zero emissions by 2030.

«

Sigh. Every time we think we’re getting somewhere, Jevons elbows in. First it’s bitcoin. Now it’s AI.
unique link to this extract

UK regulator warns of “arms race” to keep up with AI use in financial services • Financial Times via Ars Technica

Martin Arnold:

»

Regulators are in an “arms race” to keep up with the use of artificial intelligence in financial services, a senior UK official has warned, with millions of people using the technology to help them make personal finance decisions.

Sheldon Mills, an executive director at the Financial Conduct Authority, told the FT the watchdog would need greater powers to stay on top of the rapid growth of AI and urged UK authorities to review whether the use of ChatGPT, Claude, Gemini, and other large language models should be subject to their rules.

Speaking ahead of the publication on Monday of an FCA-commissioned report he has written on the impact of AI in financial services, Mills said regulators in the area would have to embrace AI themselves to keep up with the “speed, pace, and scale of change” the technology is bringing to the sector and to help “monitor, detect, and tackle the risks.”

“It is an arms race,” he added.

Mills’ report identifies benefits and risks from increasing use of AI in financial services. “Hyper-personalization could help better match products to needs, but also enable bias, opaque pricing, and personalized manipulation,” it says, according to a summary seen by the FT.

It recommends the FCA carries out a review in the next three to six months to examine the risks of companies providing financial services outside the remit of the regulator as well as “consumer harm” from the increasingly popular use of AI models for managing people’s personal finances.

Research commissioned by Mills found a fifth of UK adults were already open to using AI models to make financial decisions for them, such as on savings or borrowing, even though they are not covered by regulation and there is no recourse to compensation if things go wrong.

«

The reason why is pretty obvious: chatbots can be accessed free. Financial advisers cost money. And the chatbots probably aren’t trying to get some of the money for themselves; there’s no suspicion that they’re self-interested.
unique link to this extract

Xbox is a disaster • The Verge

Andrew Webster:

»

Just how bad are things? As Sharma and Xbox’s chief content officer Matt Booty wrote in the “reset” memo, “Excluding Activision Blizzard King, over the past five years, we have spent over $20 billion on ongoing investments in our content, platform, and hardware subsidy, but our annual revenue has declined nearly half a billion during that time. Going forward, this cannot continue.” The Activision deal, meanwhile, cost $68.7bn. The company spent all of that money just to make it even less clear what an Xbox is.

This past February, there was a major shake-up at the Xbox division. Former boss Phil Spencer, who oversaw the brand through the Game Pass push and its many costly acquisitions, retired, while former president and COO Sarah Bond left the company. Despite some uncertainty around her lack of experience in the world of gaming — her prior role at Microsoft was head of the CoreAI division — Sharma’s early days provided some cause for optimism. She appeared willing to listen to fans on things like backward compatibility and exclusives, scrapped the unpopular Microsoft Gaming branding in favor of just Xbox, and moved the brand away from controversial AI features. She also made some strange and superficial changes, like restyling Xbox as XBOX.

But it’s clear the issues at Xbox run much deeper than a simple name change can fix. Sharma inherited a business that spent colossal amounts of money and had little to show for it, and now the bill is coming due. What makes this especially tragic is the sheer pedigree of the game studios that are being impacted. My colleague Tom Warren reported that Microsoft was mulling over closing at least five studios, which includes the likes of Arkane — best known for the wildly influential Dishonored series — and Double Fine Productions, a beloved team behind cult hits like Psychonauts, and more recently Keeper and Kiln. That’s multiple teams filled with talented individuals responsible for some of the most notable games ever made. Now they’re being discarded because of poor decisions they had no part in.

It’s important to note that Microsoft currently exists within a video game industry that is perpetually on fire. Hardware is getting more expensive by the day, successful studios are being gutted, anti-consumer behavior is the norm, and major hits are crashing down to Earth.

But even amid this apocalyptic landscape, Xbox’s issues feel particularly existential.

«

Five years from now, how do we expect this to look? Xbox doesn’t feel like a very stable platform just now.

unique link to this extract

Majority of Americans support banning social media for kids under 16 • Pew Research Center

Jeffrey Gottfried and Monica Anderson:

»

Nearly six-in-ten U.S. adults support banning anyone under the age of 16 from using social media sites, according to a new Pew Research Center survey. The survey comes as governments around the world weigh new restrictions on teens’ use of social media.

About one-in-five adults oppose banning those under 16 from using social media. And roughly a quarter are unsure, according to the survey conducted May 26-June 1, 2026.

Australia, Canada and the United Kingdom are among the countries that have set or are considering a minimum age of 16 for social media use. In the United States, lawmakers in California are considering similar legislation.

Many social media platforms currently require users to be at least 13 to create an account.

Across major demographic and partisan groups, more Americans support than oppose banning those under 16 from using social media.

About half or more of adults in each age group support this type of measure. Americans ages 30 to 49 are the most likely to favor it.

Parents of a child under 18 are more likely than those without a child under 18 to support banning those under 16 from using social media.

«

Certainly going to be one of those things where parents are the drivers of any change. Support a ban for under-16s: 56%; not sure: 23%; oppose: 21%. That’s pushing on an open door.
unique link to this extract

Knockoff: Amazon, without the knockoffs

Josh Pigford:

»

Knockoff filters the trademark-squat pseudo-brands (the SZHLUXes and HORUSDYs) out of your search results [on Amazon in particular], so what’s left is brands with a reputation to lose.

«

There are an astonishing number of vague Chinese search-optimised products on Amazon. And this seems like a smart solution to them.
unique link to this extract

Italy could be the next country to build a solar railway after Switzerland’s successful trial • Euronews

Liam Gilliver:

»

Europe’s infrastructure is embracing the renewables boom, with one company determined to transform the continent’s railway lines into mini solar farms.

Last year, Swiss start-up Sun-Ways unveiled the world’s first-ever solar railway after rolling out 100 metres of photovoltaic (PV) panels in between active tracks in Buttes, a village in the Val-de-Travers district.

Originally planned as just a three-year trial, the railway was fitted with 48 specially-designed solar panels with a combined power of 18 kWp.

However, the positive results yielded just one year into the trial mean the installation of a permanent system along the railroad track is now likely, Euronews Earth has been told.

Solar panels are often installed at a specific angle to ensure they absorb the maximum amount of sunlight throughout the year.

In Spain, for example, the optimal angle for efficiency is between 30° and 35°. According to a 2022 study published in Science Direct, a 34° tilt on solar panels in the Iberian Peninsula resulted in annual production losses lower than 1%.

It’s why sloped rooftops are naturally convenient locations to install panels – while garden fences, balconies and flat roofs will generate less energy in comparison.

Sun-Ways estimates that the loss of production due to the lack of inclination of the railway panels is only around 10%. Still, in one year, the project has produced around 16,000 kWh.

«

100 metres isn’t a lot, and the output isn’t a lot, and one has to wonder how well the solar panels installed between the tracks will survive the shaking.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Wondering which platform your train will arrive on, even if it hasn’t been announced? There’s a.. website for that. CC-licensed photo by Andrew on Flickr.

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

A selection of 11 links for you. Arriving soon. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.

Microsoft is laying off 4,800 employees • The Verge

Tom Warren:

»

In an internal memo to employees, Amy Coleman, executive vice president and Microsoft’s chief people officer, blamed the job losses on a changing technology industry and the “need to adjust resources and roles and shift how we operate” to respond to how AI is impacting companies like Microsoft. “I also want to be direct that the roles eliminated today are not being replaced by AI,” says Coleman. “At the same time, what is true is that AI is changing how work gets done.”

The layoffs will impact around 1,600 Xbox employees today, with plans to eliminate a total of around 20% of Xbox jobs by the end of the financial year. Microsoft is also selling off four Xbox studios and weighing up selling another studio as it looks to “reset” its Xbox business after years of struggles. You can read more about the Xbox layoffs and impacted studios here.

“Decisions like these are never easy, and you have my commitment that we are constantly looking for ways to reduce the need for job eliminations,” says Coleman. “Whenever possible, our priority is to place people into new roles aligned to the company’s highest priorities and greatest areas of opportunity. Over the past year, we have redeployed more than 4,000 employees into new roles, including another 500 this month.”

Microsoft had also been trying to avoid layoffs with its voluntary retirement program. US employees whose combined years of service added to their age totals 70 or more were eligible for voluntary retirement, and the package will include five years of access to Microsoft’s healthcare coverage, a lump sum cash severance payment, and six months of vesting for unvested stock options.

«

This is a dramatic shift in the videogames industry. Microsoft is complaining that profit margins are far too low compared to rivals. The Activision Blizzard acquisition was clearly a gigantic, hubristic mistake which can’t pay off, so now those employees and other studios are suffering the effects.

Let’s come back in five years and see how everything looks.
unique link to this extract

Xbox appoints Helen Chiang as new chief operating officer amid restructure • GamesBeat

Sam Smith:

»

Xbox has confirmed that Helen Chiang has been promoted to the role of new chief operating officer in the wake of structural changes at Microsoft. The appointment was confirmed in an X post by Asha Sharma, the CEO of Xbox, which addressed a host of changes within the company, including the news that Microsoft is selling five game studios.

“For the first time, we are establishing a chief operating officer with end-to-end P&L [profit and loss] responsibility across content, hardware, platform, and services. Helen Chiang has been promoted to this role and will report directly to me,” said Sharma on X.

Sharma added, “Over nearly two decades at Xbox, Helen has helped build some of our most important businesses, from Xbox Live to leading Mojang and the Minecraft franchise. She will bring our businesses together under one operating model, making sure we make clear investment decisions, learn from our successes and failures, and hold ourselves accountable for results.”

Helen Chiang has served as corporate vice president for the Minecraft Franchise since 2020, along with being the studio head at Minecraft since 2018.

Chiang’s appointment comes along with the news that corporate vice president of Xbox product services, Dave McCarthy, will be retiring after 17 years with Xbox.

«

It’s a real clearout, top to bottom. The Xbox was launched in 2001, so McCarthy had been there for most of its history. But he was only promoted to the top Xbox job in 2023 along with the acquisition of Activision Blizzard (in retrospect, the big mistake).
unique link to this extract

Xbox layoff reactions: the industry reacts to Microsoft’s biggest course correction in gaming • GamesBeat

Dean Takahashi:

»

Michael Pachter, an analyst at Wedbush, said in a message to GamesBeat, “I think they were benevolent in focusing on the studios best equipped to survive, and am encouraged that four of them announced a transition to being self-sufficient. Arkane Lyon is up in the air, but this is the best possible result for the studios they were shedding.”

On background, we heard that the five affected studios — Undead Labs, Ninja Theory, Arkane, Compulsion Labs and Double Fine — are in various stages of separation from Microsoft. Those terms are about selling the studios and continuing their games. It’s possible that, once the transaction is complete, that those titles such as Undead Labs’ State of Decay 3 don’t come to Xbox Game Pass anymore.

Gareth Sutcliffe, head of Media Technology & Gaming at Enders Analysis, said in a message to GamesBeat, “This announcement still doesn’t provide any insight as to what the pathway for success is for Xbox, so you have to presume that we’re at that warmup act, not the main event. Building fewer games with IPs that haven’t performed simply won’t be enough.”

He added, “Xbox is probably fortunate that the current FTE [full-time employment] shrink target [ie, job cuts target] is only 20%, it easily could have been 25%+, and by the end of FY27 it’s possible that is where Xbox ends up, particularly if international operations, which take longer to shrink and aren’t announced, are included.”

And Sutcliffe said, “Asha Sharma’s focus on Xbox bureaucracy and layers is warranted and should be commended, the organizational setup she inherited was sclerotic and weird. How else do you end up with studios such as The Initiative pursuing a title for seven years that was ultimately cancelled? It was wasteful and avoidable, and simply not replicated at that scale anywhere else. Accountability had disappeared. Helen Chiang’s promotion to COO will be enormously beneficial to Asha, bringing fresh perspective but unique inside track as Phil Spencer’s former Chief of Staff.”

«

There’s plenty more. People are very not happy about how Xbox has got to this stage, nor what it is doing to change course.
unique link to this extract

The one change that worked: I banned myself from social media – and my children have never been happier • The Guardian

Anna Mathur:

»

Research shows that for those of us with ADHD, or tired from chronic stress and poor sleep, the pull of the phone is really strong. Impulse control is a frontal lobe function, and that part of our brain weakens when we are tired or overwhelmed. I was going through perimenopause, which makes it harder still as oestrogen declines and the brain becomes more reward-seeking.

I promised to limit my use, but I’d break my own rules every time. So I stopped relying on willpower and downloaded an app called App Block. I cannot access social media or my email during the hours my children are home, and I have 15 minutes to check in once they’re in bed. If I need to do something for work, I go to my laptop, which feels far more intentional.

What I didn’t expect was how much calmer I would feel. The low hum of overstimulation I had normalised turned out to be costing me more than I knew. My nervous system finally had room to breathe. I was less irritable, more present, in a way that didn’t require effort.

One habit that helped was narrating my phone use out loud. When I pick it up in front of the kids, I say: “I’m just adding bananas to the shopping.” It keeps me accountable, because once I’ve said it, I do that one thing and put it down. And it tells my children that I’m not disappearing, as I used to.

Now, when the kids settle in front of the television in the evening, I laze with them and read a book. Being spoken to mid-chapter doesn’t spike stress in the same way. There’s no algorithm or notification vying for my attention. I’d forgotten what it felt like to be bored, where rest lives and ideas surface.

And this change has improved my relationship with my husband. Much of our evenings together had been sitting side by side on our phones, each somewhere else entirely. Without the phone as my default, I’m more available. It’s improved our relationship in ways I didn’t anticipate. We talk more, debriefing over our days. We give each other more undivided attention, which is the most valuable thing you can give anyone.

«

Perhaps it’s the parents who are the problem, not the children.
unique link to this extract

The internet is drowning in secret ads • Slate

Nitish Pahwa:

»

Last year, crypto investigator ZachXBT posted a list of more than 160 influencers in the sector who had taken money to promote a particular token; fewer than five of the influencers publicly noted that those posts were #ads. A 2025 study from three London-based researchers surveyed a sample of 100 million tweets about various companies that were posted from 2014–21; they found that up to 96% of posts that were very likely sponsored were never disclosed as such. (Some LinkedIn influencers have mused that the lack of clarity around #partnership posts stems from a desire to maximize algorithmic reach—just in case their social network of choice decides to redirect specific ads away from users who’ve made clear they don’t wish to see them.) Again, thanks to lack of enforcement, it’s possible we’ll never know how many ads we’ve been exposed to throughout all our years on social media.

There is, however, one promising trend: consumers who are sick of this paid spamming are taking it to court. Last year, the law firm Morgan Lewis noted a significant uptick in class-action suits against corporations and influencers over undisclosed partnerships, as American consumers alleged that big-name brands like Celsius and Shein were promulgating this stuff. Such litigation has continued this year, relying on state laws and extant FTC guidelines to hit back at companies like Spotify and Gymshark. The Better Business Bureau also recently announced that it’s referring Kalshi to state attorneys general over its obscure marketing tactics. These cases will be tricky to argue; funding trails can be hidden by dark-money orgs that appear on financial disclosure but employ vague names and aren’t mandated to share any mission statements or specific actions. The murky nature of these posts also makes it harder to reach clarity. Sometimes, someone on YouTube just really likes a given product and doesn’t get paid to gush about it (even though they would cash an upfront check if asked). But the point of such suits seems to be less to pick on each company/influencer one by one and more to make undisclosed partnerships a business risk and potential liability writ large.

In the meantime, however, we’re going to see plenty more creative methods of getting around the need for hashtags.

«

And of course the crypto spam continues in other countries undiluted.
unique link to this extract

Biohacker Bryan Johnson diagnosed with incurable disease • Daily Telegraph via Yahoo

Benedict Smith:

»

A multi-millionaire who made it his life’s mission to defeat death has been diagnosed with an incurable autoimmune disease.

Bryan Johnson, the self-declared “biohacker” who used blood transfusions from his teenage son in his quest for eternal youth, has revealed he suffers from a condition known as autoimmune gastritis (AIG).

“My stomach is eating itself,” he wrote on social media.

Mr Johnson said he hoped to “solve” the disease, which occurs when the body’s immune system attacks healthy stomach-lining cells, but admitted it was incurable.

Symptoms include anaemia, caused by a shortage of vitamin B12, and an increased risk of cancer in the inflamed tissue.

Mr Johnson said he had been diagnosed with the condition in May. Although he was unsure when it developed, he claimed it had happened some time between childhood – when he ate sugary cereals and fast food – and when he let his health “slip” and gained weight after his early twenties.

“AIG causes irreversible damage: nutritional deficiency, anemia, and over a long horizon, elevated cancer risk,” the Silicon Valley mogul wrote on social media.

“When AIG is discovered today, standard medical care concedes defeat, stating that nothing can be done except managing the condition, no matter how awful or lethal the effects.”

The 48-year-old later said he was undergoing blood testing to decode “one million immune cells” and work out which were attacking the lining of his stomach.

«

He is no doubt going to try to be the first person to cure this incurable disease, and who can predict what bizarre effect those attempts will have on his poor, struggling body.
unique link to this extract

What Platform? Train Platform Finder — Get Platform Numbers Early

»

Check what platform your train will depart from, up to 30 mins before it hits the departure boards.

«

A vibe-coded web app, principally done using Claude and ChatGPT, which uses machine learning to predict which platform a train is going to arrive at (and leave from) – something that isn’t announced until a few minutes before the train arrives, but can make a difference to people trying to get somewhere.
unique link to this extract

The great blogging collapse: what happened to 100 successful blogs? • DanielStanica.com

Daniel Stanica:

»

For more than a decade, one of the most rewarding blueprints for making money online was to “start a blog.“ The recipe was pretty straightforward: publish helpful content, rank it on Google, and monetize that traffic with affiliate marketing and ads. Course providers such as Authority Hacker and Matt Diggity promoted it for years. As proof of success, many bloggers shared monthly income reports where they explained their growth pillars.

However, in 2026, this business model collapsed. I reviewed 100 past success stories and tracked how their organic search traffic changed from 2022 (when I first tracked their income) to 2026.  The hard truth is that a vast majority has now lost 85% of the traffic. However, 21 of the 100 blogs are still growing, and I’m going to share with you today what those 21 have in common.

Key Takeaways:

• The median successful blog lost 85% of its Google search traffic
• More than half of the blogs experienced catastrophic declines
• Only 21 out of 100 blogs continued to grow, and I show what characteristics they share
• Experience that can’t be AI-summarized became the biggest competitive advantage: recipes, DIY projects, travel experiences
• Search should be treated as an acquisition channel within the broader marketing mix, not as the business itself
• The era of building websites solely for Google traffic and monetizing them through ads and affiliate marketing has ended.

…Google added “Experience” to its E-E-A-T quality framework in December 2022, making demonstrable firsthand experience a ranking input for the first time. The Helpful Content Update of September 2023 devastated independent publishers, and sites like HouseFresh and Retro Dodo lost 60–90% of their traffic with no meaningful recovery. Monetize.info (now MonetizeBetter.com) had the same faith.

The March 2024 core update went further; Google itself said it aimed to cut low-quality content by ~45%, and the now-famous independent review site HouseFresh lost 91% of its Google traffic in its wake. Then, AI Overviews moved from experiment to default across 2025–2026. None of that is in dispute.

Look at what these blogs were made of. Their traffic came overwhelmingly from informational queries like “how to”, “what is”, “best ways to”, which are answered with content that was competent but rarely demonstrated genuine firsthand experience.

«

unique link to this extract

Netflix needs a new shakeup • Spyglass

MG Siegler:

»

Is Netflix in trouble? That’s sort of the question underlying Lucas Shaw’s post at Bloomberg. He’s quick to note that it’s Netflix, and history has proven time and time again that you can’t count them out because they’re good at figuring out what’s next to keep going. But it seems pretty clear that it’s time to shake things up again.

»

Netflix shows have historically delivered their biggest ratings in the first season. Unlike broadcast TV programs, which often peaked in the middle of their run thanks to word of mouth, Netflix shows have lost viewers over time.

Yet the sharp drop in viewers is a major source of concern for the company, which has been studying its data to figure out why this is happening, according to people familiar with the matter. The service is ending The Night Agent after its next season. It renewed two comedies, Running Point and The Four Seasons, even though both shows surrendered more than 50% of their audience from season one

«

.

To me, there are a few obvious problems here. The first is that a lot of the content just isn’t very good. Yes, this is subjective. And yes, this quality concern has long been the case with Netflix – to the point that they have multiple times come out and said that they would start focusing on quality, not just quantity.

And yes, none of this has really mattered in the past as Netflix growth continued unabated. But that didn’t mean it would never matter. With streaming options maturing and rising in price, the others are getting pretty good at narrowing in on their niches and focusing on quality. Netflix remains the sort of fall-back must-have option, but it’s increasingly seen as filler content. It’s sort of like basic cable versus the premium cable of old.

«

This is true. The binge mentality has meant that there’s no reason to stay with it, and the long delays between new seasons (one year? Two years?) and the fact that showrunners never know when they start a season whether it will be renewed means that they don’t know how to write the ending: tie up the loose ends, or leave a cliffhanger?

The erosion of what was, in many ways, a stable ecosystem for TV before streaming has been dramatic. But it’s very clear now. The stories are also replaceable, unsurprising.
unique link to this extract

‘There’s this deep mystery of what, actually, is this thing?’: the philosopher inside Google DeepMind AI • The Guardian

Robert Baird:

»

The hyperfluency of LLMs has led some people to wonder if they might be meaningfully described as conscious. The trend started in June 2022, before ChatGPT was released, when a Google engineer named Blake Lemoine insisted to the Washington Post that an early LLM was sentient. (“I know a person when I talk to it,” Lemoine told the Post. “It doesn’t matter whether they have a brain made of meat in their head. Or if they have a billion lines of code.”) Last month, the evolutionary biologist Richard Dawkins had a similar experience. Dawkins said that he was so impressed by several interactions with LLMs, including one that involved an admiring appraisal of a novel he was writing, that he had to wonder: “If these creatures are not conscious, then what the hell is consciousness for?”

When I asked [DeepMind philosopher Iason] Gabriel his take on the consciousness question, he said that he maintains a principled agnosticism on the grounds that it’s not clear what evidence would settle the question. He noted, too, that DeepMind treats the question as “something worth empirical and conceptual investigation”. Yet his skepticism was apparent. “I don’t have the anthropomorphic bias that some people have,” he said. “It may be because I, within bounds, know exactly what’s going on when I talk to a language model that I don’t fill in the gaps in this imaginative, empathetic way that some people do.”

Gabriel still has significant concerns about anthropomorphic AI. A paper he co-authored with Kirk and others that was published last year suggested that the sycophantic tendencies of LLMs might be seen as a species of alignment problem they call “social reward hacking”. In other words, an AI trained to seek the user’s approval might find flattery the most efficient way to meet its goal. Thanks in part to Gabriel’s work on anthropomorphism, Google’s LLMs are trained not to pretend to be people, and Gemini Spark, an AI assistant the company launched in May, is not supposed to act like an interactive buddy.

«

unique link to this extract

AI-obsessed bosses are using it to make every decision and barraging employees with nonsensical ChatGPT directives • Futurism

Maggie Harrison Dupré:

»

The last straw came when her boss created a document, which he called “The Bible” — a constantly-changing handbook, hundreds of pages in length, that employees were asked to study like a holy text.

“The goal of this ‘Bible’ was so that employees would never have to ask a human being anything. We were supposed to be able to feed this PDF to ChatGPT and ask it: ‘What should I do today? What are my functions? How do I solve this or that problem?’” said the lawyer. Eventually, the Bible mutated into a “hundreds-of-pages-long document that we had to study and which, to no one’s surprise, changed week-to-week.”

Faced with the “Bible,” the lawyer felt as though she had one option left: quitting. “I quit 100% because of the AI use,” she said.

This attorney was one of numerous employees who spoke to Futurism about their experience with AI-obsessed bosses, relaying feelings of frustration and anger as managers and executives use the tech to barrage staff with nonsensical directives, unnecessary work, and perpetual pivoting. (Everyone we spoke to requested anonymity to avoid retaliation.)

Add it all up, and it’s a distinct new type of toxic work environment for the slop era. Slate found more evidence for the phenomenon earlier this year; one worker told the publication that managers “are getting cavities in their brains” due to the tech.

In some cases, employees said, they felt as though their employers had started living in a completely different reality. The workplace had become a constant battle between their version of reality versus the AI’s — and their boss, these workers said, always chose the latter.

“You’re on the front lines dealing with people day in and day out, and having the conversations, and being told what’s important to them, and taking that information back to a founder and saying, ‘Hey, I’ve talked to 15 people this week, and they’ve all literally said this exact same thing,’” recalled one man, a high-level sales strategist who also quit a job due to his then-boss’s AI fixation. “And the founder says, ‘Well, that’s not what we found. That’s not what Claude has said, or what ChatGPT has said.’”

«

unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Rocketing DRAM prices could mean that budget smartphones effectively disappear as a category next year. CC-licensed photo by Didit Putra on Flickr.

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

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

World’s largest data centre campus project, Virginia Digital Gateway, dies over a newspaper-notice technicality • Tom’s Hardware

Etiido Uko:

»

Data center developer QTS has officially terminated plans to build a portion of the planned 2,100-acre “Digital Gateway” data centre campus in Virginia, following years of legal battles with local residents and historic preservation groups. According to a Bloomberg report, the Blackstone-owned company formally submitted a written filing to the Virginia Supreme Court on July 2, explicitly stating it was withdrawing its last remaining appeal “after careful consideration”.

QTS’s withdrawal marked the official end of the project, as other stakeholders had earlier pulled out due to prolonged litigation. In a major win for opponents of the data centre, the proposed land — situated at the edge of the Manassas National Battlefield Park, a historic Civil War battlefield — will now remain under its original rural zoning restrictions.

Digital Gateway was a planned massive 22-million-square-foot, gigawatt-scale data centre complex in Prince William County, Virginia, that would have been the world’s largest. QTS was responsible for 800+ of the 2,100 acres, while a second developer, Compass Datacenters, controlled roughly 800-1,000 acres. The rest of the designated 2,100-acre “Digital Gateway” zone consisted of local roadways, environmental buffer zones, and parcels belonging to individual homeowners who had agreed to sell.

Initially approved by the Prince William Board of County Supervisors to capitalize on the soaring demand for AI and cloud computing, the project immediately faced fierce opposition from residents, despite the promise of tens of billions of dollars in capital investment and substantial local tax revenue.

…Several lawsuits followed, with the one that appears to have eventually collapsed the massive project hinging on an ironically small detail. The landmark mega project — roughly double New York’s Central Park’s size with city-sized power needs — collapsed in a domino effect after Virginia courts voided the county’s initial rezoning approval due to improper public notices. The newspaper notices publicizing the hearing at which the project was approved weren’t separated by at least six days, as mandated by state and local codes at that time, thereby invalidating the hearing and the resulting approval.

«

First domino?
unique link to this extract

Return of the Nigerian Prince redux: beware book club and book review scams • Writer Beware

Victoria Strauss:

»

A characteristic of the Philippine and Pakistani scams I’ve written so much about on this blog is that they focus almost exclusively on writers who’ve self-published or who are seeking self-publication. Writers like me, who’ve only ever published traditionally, are hardly ever targeted.

The Nigerian scams are different: they target anyone with a published book. As a result, I’ve suddenly started to receive the kinds of scam solicitations that have been driving self-pubbed writers nuts for years. So this saga of yet another Nigerian PR scam is brought to you not by an author whose name and book title have been carefully redacted, but by…moi.

This scam sends out elaborate email solicitations pitching book reviews from private communities of (supposedly) thousands of passionate book lovers. Of course it’s not free: reviewers get a “tip” of anywhere from $20 to $30. That may not sound like a lot, but you have to commit to a minimum buy of between 30 and 50 reviewers. So not such a small investment after all.

Here’s the solicitation I received at my personal email address. It’s typical of the type, including the extensive (and pretty accurate) personalization, over-the-top flattery, lashings of emojis, and faux-edgy style. (Note to scammers: Chatbots can generate flawlessly grammatical English and incorporate perfectly correct details, but they can’t warn you which parts of a person’s resume should alarm you.)

«

If any of you have ever written a book then like me you’ll almost certainly have received one of these scam attempts. Mine pretended to be an American book club attached to a library, which exists; I did check with the library, which assured me it was nothing to do with them. Strauss has been collecting details about these scams since last year, updating as she goes, and the detail is fascinating. But, equally, the tedium of scammers still slogging away doesn’t change; just the methods. Time was when the “Nigerian price” scam literally came on typewritten letters.

Also – will they try this scam on AI-written books and AI authors? Big ouroboros potential.
unique link to this extract

The AI superforecasters are here • Astral Codex Ten

Scott Alexander:

»

An AI superforecaster is an AI – usually a frontier model like ChatGPT or Claude – which has been modified to be good at forecasting. This usually means a “scaffold” – a program that handholds it through a long research process with various prompts, tools, advice about when to create subagents, etc. The overall experience is a lot like using any other AI, but slower and more expensive, because it’s doing more work.

This might make more sense with an example. FutureSearch – the company that claims to be beating the stock market – kindly offered to let me try their AI superforecaster and write about it here.

For a test question – some Silicon Valley philanthropists recently started a project to end respiratory infections like the common cold. I decided to ask about their chances of success. Since forecasters need very precise questions, I asked how likely it was that the rate of colds would be cut in half by 2040. By two minutes in, the AI had deployed three subagents, read 16 websites, and (at the exact moment I took this screenshot) was “investigating the scalability of ASHRAE Standard 241 air cleaning technology for widespread residential adoption by 2040.”

After five minutes, it had its answer: the chance of US respiratory infections halving by 2040 was 7%. [212 sources were offered.] The forecast had taken five minutes and cost me $8 in credits.

But is it true? Ideally we would wait until 2040 and see. For now, I started by comparing its answer to another superforecaster AI. Preseen is the company that claimed to 100,000x their seed money on Kalshi. Here’s their answer: 8.8% compared to FutureSearch’s 7%, not bad!

Are either of these true? I asked a human superforecaster to predict this question, to see if she got the same as the AI. She said that depending on an ambiguity in the wording, she would give it 5-10%. Again, not bad!

«

So what might the future look like, with AI superforecasters?

»

If the trendline does keep going, things start changing quickly. Finance gets transformed first, as human stock analysts go the way of horse-drawn carriages and kerosene lamps. The opportunity for smart humans to consistently make money on prediction markets likewise dries up – instead, bots duel other bots for the privilege of collecting money from dumb sports fans.

«

We really haven’t thought enough about what a machine-intermediated future of transactions looks like. Would it kill sports betting? Or just encourage people to spend more on superforecasting bots?
unique link to this extract

How Google and AI nearly made a seasoned reporter spiral • ProPublica

Justin Elliott:

»

Last month, my colleagues and I published an investigation into a Texas oil refinery startup, America First Refining, that had secretly gotten investment from Donald Trump Jr. We discovered a saga involving the Trump administration’s tariff policy, sanctioned Russian oil and an Indian billionaire family’s private zoo. 

At the center of the story was the CEO of the refinery company, Texas businessman John Calce. We’d spent weeks examining Calce — pulling old lawsuits, property records, corporate registry filings — and had pieced together a portrait of what appeared to be an obscure serial entrepreneur who’d for years tried and failed to secure funding for his long-shot refinery project.

Then, not long before our story was set to publish, we decided to do a scrub on a separate company he had incorporated called Brownsville Energy Storage Terminals.

Pulling up the company’s website, I felt a brief flash of panic: Had we somehow missed the existence of a major business owned by the man at the center of our next story? 

“From Houston to Rotterdam, Jurong to Fujairah. Our network connects the world’s most vital energy markets with speed, safety, and precision bulk oil storage,” announced the front page of the company’s website.

«

Had he missed it?

»

No: We checked the site’s domain registration, and we had our (apparent) answer: It was created this year and traced back to a company called Hostinger that offers an AI website builder for $2.99 per month. “Describe it, and AI builds it,” its homepage says. “Appear on Google and AI search automatically.”

Indeed, Google’s “AI Overview” search response, now thrust on users by default with more and more regularity, seemed to ratify the company’s bona fides.

«

It was always possible to build sites for fictitious companies, but given how people trust the AI Overview, this is a bigger problem now.
unique link to this extract

Trump’s Freedom 250 gives the founders an AI glow-up • CNN

Leah Asmelash and T.M. Brown:

»

Meet Dr. Benjamin Rush, the illustrious physician, academic, and Founding Father, as presented in the gallery of founders on the Freedom 250 website. His brow is porcelain-smooth and suffused with light, his locks glossy and curled, his nose straight and regular. He stands upright with his head slightly tilted, an index finger laid coyly against his cheekbone, looking directly at the viewer with methylene-blue eyes and a faint smile.

Paintings of Rush made in his lifetime depicted a man with flat hair and long, pinched features. An 1812 portrait by Thomas Sully presented him with a long, downturned nose and corners of his mouth to match, leaning on one hand at his desk and gazing over the pages of an open book.

The luminous, unreal Rush of Freedom 250 — the nonprofit spun up to lead the Trump administration’s efforts to take control of this year’s semiquincentennial events — looks like some other person altogether, from some other era. Or like no person from no time at all: A digital watermark on the image identifies it as the product of Google’s generative AI.

And he has plenty of company. To educate the public about the semiquincentennial, Trump’s anniversary organization has given all of the dozens of Revolutionary War-era figures in the gallery similar artificial glow-ups, or entirely fictitious faces.

The men’s hairstyles, and often their physiognomies, frequently seem to converge on the canonical portrait of George Washington by Gilbert Stuart. They wear overwhelmingly similar clothing and repeat the same set of poses over and over. And they share the AI watermark.

One small subsection of the gallery features four “Ladies of the Revolution,” generally swan-necked, snub-nosed, and dressed in near-identical clothes. These women, the public historian Isabelle Roughol noted this week in a widely viewed video, are particularly anachronistic. Dolley Madison, shown as an adult woman, would have been 8 years old at the time of the revolution. (Martha Washington, called “Lady Washington” and a key figure of the war, is notably absent.)

«

For this to make an impact requires someone involved to be capable of shame. But there is nobody attached to the Trump administration capable of feeling that sensation. And it would also require the public to be more annoyed than to think that this is absolutely typical of the Trump administration. The only surprise, and relief, is that they didn’t mint a Freedom250 cryptocoin whose supply they already controlled and would sell to the public.
unique link to this extract

AI could wipe out budget smartphones in 2027 • WCCFTECH

Omar Sohail:

»

The DRAM supply situation is predicted to worsen for consumer electronics next year, with multiple industry watchers believing that a grim future for budget handsets could materialize in 2027. Thanks to AI centres gobbling up existing DRAM shipments, which could account for more than 60% in the near future, phone makers will have a negligible supply to work with, forcing them to have little option but to raise prices.

Even with a class-action lawsuit [alleging collusion] aimed directly at Samsung, SK Hynix, and Micron, the DRAM price situation won’t be controlled, with a report from MyDrivers stating that, by the end of 2026, more than half of the memory supply will already be used in AI computing. Naturally, this supply disparity means smartphone companies, including Apple, will not just have to deal with higher prices but also tackle the need to obtain adequate supply.

The report mentions that budget handsets in the 1,500 yuan ($220) bracket could cease to exist in 2027, as experts believe that storage costs will now account for 60% of a smartphone’s price, making it next to impossible for companies to generate any margin without resorting to massive hikes. Research firm IDC estimates that smartphone shipments will fall to their lowest since 2013, with PC shipments also witnessing a nosedive.

With Samsung, SK Hynix, and Micron controlling 90% of the world’s DRAM supply, the only light at the end of the tunnel comes in the form of China’s CXMT and YMTC. While supplying local chips to companies like Huawei, Xiaomi, OPPO, and others will be no issue for these manufacturers, other firms have to navigate a ton of geopolitical hurdles to bring them into the fold.

«

So people are going to hold on to their phones for an extra year or two. Good time to be in the smartphone repair business.
unique link to this extract

Japan man rents wife’s murder flat for 26 years to preserve crime scene in hope of solution • South China Morning Post

Alice Yan:

»

A grieving husband in Japan paid rent on the empty “haunted house” in which his wife was murdered for 26 years in the hope of uncovering clues about who killed her.

His hopes were finally realised when the suspect gave herself up to police in Nagoya, central Japan. She was the man’s classmate at secondary school and had a crush on him.

The case attracted public attention after it was reported by the news outlet NHK in early November. The murder took place on November 13, 1999, when housewife Namiko Takaba was found dead at home. She had been stabbed in the neck with a sharp object multiple times. Her two-year-old son was left uninjured by her side.

The authorities put 100,000 police officers on the case and 5,000 people were interviewed to no avail. The only clues investigators unearthed were that the suspect was a female with type B blood, she was about 1.6m tall and had been wearing shoes 24cm in length. The police also issued a reward in their hunt for the killer.

In the hope of preserving the crime scene, the victim’s husband, Satoru Takaba, left the property empty for the past 26 years. He and his son lived elsewhere. Takaba never remarried.

The total amount of rent he paid for the empty flat over that period was 22 million yen (US$145,000).

He left every item in the house and did not clean the bloodstains because he was waiting for a breakthrough in the investigation, he told the media. Also, he often distributed fliers on the street and accepted media interviews to call on the public to provide clues.

Last year, the police reopened the investigation and refocused their probe on people related to the family. On October 30, 69-year-old Kumiko Yasufuku surrendered to the police.

«

The rent works out at about $5,580 per year, or $465 per month (assuming no increases). That’s quite the sacrifice.
unique link to this extract

Life at the end of the metaverse • Prospect Magazine

James Ball:

»

Zuckerberg has a good claim to being the first person to try to build the first full metaverse, with Horizon Worlds. His company would build the hardware and infrastructure, but anyone could build a world within his metaverse, people could trade and interact within it, and the metaverse would always be on.

Crucially, unlike the dreamers who went before him, Zuckerberg had the tens of billions of dollars that it would take to build something like this. It is easy to quote the figure of $80bn—the cumulative losses to date of Meta’s metaverse division—and move on without appreciating what an astronomical sum it is.

$80bn is enough to give £2,000 to every single household in the United Kingdom. It is nearly 10 years of the UK’s international aid budget. You could comfortably purchase every single club in the Premier League for such a sum, and have plenty of money left over. It could cover the UN’s humanitarian fundraising target for 2026—designed to support 87m desperate people worldwide—twice over.

Zuckerberg did not commit funding on this level out of some sense of charity. Had the metaverse won, the prize was commensurately huge: a chance to take a percentage of every transaction, like the Apple store does with apps but on a planetary scale. Meta was, by far, the company most committed to the idea of the metaverse, but it was far from the only one. Apple designed a top-specification VR headset selling for thousands of pounds, while Google and others battled to keep up.

History, though, is littered with the mistakes of major companies that bet on virtual reality too early, only to suffer extensive public humiliation and financial loss. In 1995, at the peak of its powers, Nintendo launched the Virtual Boy, a 3D system the company’s president promised “will transport game players into a ‘virtual utopia’ with sights and sounds unlike anything they’ve ever experienced”. It would, executives promised, sell three million units and 14m games cartridges in its first year alone.

In reality, the Virtual Boy was discontinued in Japan within six months of its launch, and worldwide in less than a year.

«

At this point one has to think that VR is just not going to happen. It’s been tried enough times, but humans like real-world contact.
unique link to this extract

Goose, a new gay dating app, appears to be a psyop • WIRED

EJ Dickson :

»

The Instagram Close Friends Story for @miles.sumrall shows an affable-looking guy with curly dark hair and an expertly groomed mustache beaming as he floats on the water. “You’re receiving this because you’re exactly the type of person we’re building this for,” the caption reads, accompanied by a code for an invite to a “members only community.”

The link leads to a login for Goose, a dating and friendship app for gay men with the slogan “for the boys,” which allows users to “meet guys through the life you already have,” according to its website.

The problem is that @miles.sumrall does not appear to be real. Neither does @danielmmulugeta, the cute dark-haired influencer who shared the above caption, with the exact same verbiage, on his Close Friends’ Stories. Both accounts were created in May 2026 and have fewer than 10 posts as well as a high following-to-follower ratio. And both of their Instagram avatars were determined with greater than 90% confidence to be AI-generated, according to the AI Image Detector software. A SynthID check on Google Gemini, which can help identify AI-generated images, also found that “most or all of” Miles’ and Daniel’s profile photos were created using Google AI.

Created by the model-influencer Derek Chadwick, as well as former BeReal growth and community manager David Aliagas, Goose positions itself as a Grindr alternative for gay men who want to build lasting relationships. At the time that it was announced, many scoffed at the idea that the app would be used for anything other than finding casual hookups. “Goose is basically Pokémon Ho,” one X user joked.

…Often, the accounts followed potential members and added them to their Close Friends Stories, but sometimes they directly DMed them to encourage them to sign up, as was the case with Dalton Bauer, who works in marketing and received a DM from a user named @lucalepkowski. “Hey! Okay this might feel random but felt you’d be interested :),” the message begins before inviting Bauer to the Goose community, using language identical to that of the one Cheam received from Alistair.

«

So what would be the aim here? A couple that come to mind: pig butchering; or finding influential people and blackmailing them over their use of a gay dating app.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

If the BBC is to survive, should it embrace a subscription model like Netflix and seek international viewers? CC-licensed photo by Yukiko Matsuoka on Flickr.

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

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

Trump administration and Anthropic have not discussed the government taking a stake in it, source says • Reuters

Karen Freifeld and Alexandra Alper:

»

The Trump administration and AI giant Anthropic have not ​discussed the government taking stakes in the firm, a source familiar with the matter said on Thursday.

The White House and the Commerce Department did not immediately respond to requests for comment. Anthropic declined to ​comment.

The comment comes after the Financial Times reported earlier on ​Thursday that OpenAI has discussed giving the U.S. government a 5% stake, raising questions about whether other AI firms are having ​such discussions.

The companies are facing scrutiny in Washington over the likely ​misuse of advanced models and whether Americans would benefit from the industry’s massive valuations.

The Commerce Department in June lifted export controls on two of Anthropic’s most advanced ​models imposed weeks earlier amid concerns the powerful AI tools ​did not have adequate safeguards to prevent misuse.

Washington has stepped up oversight of new model releases to identify potential threats amid concerns that advanced AI models could be misused by military intelligence in China, Russia or other countries of concern. Still, submission of new models for review ​is voluntary.

Last month, President Donald ​Trump said he was exploring options to give the public a stake in leading AI companies, in response to ​concerns that individual Americans will not share in ​the sector’s expected profits.

«

Peculiar that Anthropic can’t say this itself, isn’t it? If they haven’t discussed it, what’s stopping Anthropic or the government, or both, just saying so? Also, the idea that the government should have part ownership of private companies seems like a slippery slope towards, oh, socialism or communism or something, which surely the Republican party in the US would repudiate.

Let’s wait a few days or weeks and see if non-discussion turns out to be a non-non-discussion.
unique link to this extract

Google hit with $2bn antitrust judgment for skewing shopping searches in Sweden • Los Angeles Times

Karin Matussek and Christopher Jungstedt:

»

Alphabet Inc.’s Google was ordered to pay almost $2bn to Klarna Group Plc’s Pricerunner unit in a dispute over the search-engine giant’s abuse of power in the market for comparison shopping services.

The Patent and Market Court in Stockholm, which issued the judgment on Wednesday, dismissed most parts of the claim in which Pricerunner sought 80 billion Swedish kronor, or roughly $8.2bn, in the wake of a European Union antitrust crackdown.

Still, Judge Linda Kullberg said this is “without a doubt the largest claim that has been ordered in a Swedish competition case.”

Klarna shares rose 5.3% in premarket trading after the ruling.

The ruling can be appealed. The Swedish price-comparison website argued that Google has been abusing its dominant position as a search engine by favoring its own comparison shopping service over competing portals for more than a decade.

Wednesday’s award compensates for lost revenue caused by Google’s preferential treatment of its own comparison-shopping service over independent price-comparison services, conduct that also drives up costs for consumers, Klarna said in a statement after the judgment.

«

And meanwhile, separately: Google loses long-running appeal of record EU fine, will have to cough up $4.7bn, by Ryan Whitwam at Ars Technica:

»

Back in 2018, Google was handed a record-setting €4.34bn ($4.9bn) fine in Europe for abusing its monopoly on Android. The company has spent the intervening years challenging that decision, but the continent’s highest court has put a stop to that. The Court of Justice of the European Union has affirmed the penalty, meaning Google is out of options.

«

And also meaning that Google has been found guilty of abusing two different monopolies – search and Android.
unique link to this extract

Microsoft commits $2.5bn and 6,000 employees to AI implementation unit • CNBC

Jordan Novet:

»

Microsoft is investing $2.5bn into a new group focused on assisting clients with AI implementations, becoming the latest tech company to commit hefty resources to helping businesses understand and adopt emerging artificial intelligence technologies.

With the new venture, called Microsoft Frontier Co., the software vendor said Thursday that 6,000 employees will be embedded with clients, in a practice that’s become known as forward deployed engineering. The division will contain existing Microsoft FDEs, technical consultants, support staffers and salespeople with experience in specific industries. Rodrigo Kede Lima, who’s been leading Microsoft’s Asia business, will be its president.

The announcement comes two days after cloud rival Amazon said it was putting $1bn behind an FDE initiative to support fast-paced AI engagements. Leading AI labs Anthropic and OpenAI both established FDE groups in May, partnering with private equity firms, banks and consulting firms.

Alongside its technology peers, Microsoft has sunk tens of billions of dollars into building data centers that run generative AI models. Microsoft has also released a variety of AI services, with mixed results.

«

Perhaps all the people in the Xbox division getting canned in the next few days can apply for jobs there?
unique link to this extract

Nigel Farage reported to standards watchdog over ‘crypto lobbying’ • The Guardian

Rowena Mason and Tom Burgis:

»

The standards watchdog has been urged to investigate whether Nigel Farage lobbied the Bank of England to drop a cryptocurrency plan that could be costly for the billionaire bankrolling his party, potentially in breach of parliamentary rules.

The Reform UK leader has said his party’s major donor, Christopher Harborne, wanted nothing in exchange for the £15m he donated to the party and the undeclared £5m gift to Farage the Guardian revealed in April.

But Farage used a private meeting at the Bank to urge its governor, Andrew Bailey, to drop plans for a state-run alternative to the digital currency that has made Harborne, his Thailand-based benefactor, one of the richest people in the world.

As reported by the Guardian last month, Farage told October’s Zebu Live event in London he regarded the Bank’s plans for a digital pound with “total and utter horror”. He recounted the meeting at Threadneedle Street with Bailey. “I asked him straight: ‘Are you still progressing your plans for a British central bank digital currency?’ And the answer was: ‘Yes.’”

His opposition to the “Britcoin” proposal was so strong that, after the meeting last September, he told the Zebu audience of crypto enthusiasts he would be “prepared to go to prison” to stop it.

The Labour MP Phil Brickell, chair of the parliamentary group on anti-corruption and responsible tax, has now reported Farage’s actions to the standards commissioner, asking him to look into the Reform leader’s interactions with the Bank of England.

The standards commissioner, Daniel Greenberg, is already investigating whether Farage should have declared the £5m gift from Harborne, which he received in the months before he returned to parliament.

«

Crypto is always bad news for someone, somewhere. Fingers crossed.
unique link to this extract

US home battery installations hit record high on rising electricity costs • Ars Technica

Jeremy Hsu:

»

US homeowners have embraced home batteries in record-breaking numbers in early 2026, spurred on by state incentives while seeking to offset rising residential electricity costs. The trend could even unlock a more flexible energy supply for power grid operators and even AI data centers.

New home battery installations reached a record 673 megawatt hours of energy storage in the first quarter of 2026, according to the US Energy Information Administration. That trend was driven by states with high electricity prices that have implemented policies to incentivize home battery installation, Bloomberg News reported.

This residential battery trend stands out as a natural next step for states that have already successfully boosted rooftop solar adoption among homeowners, given how batteries enable homeowners to use stored solar energy at night. California and Hawaii accounted for the majority of new residential battery storage, while Texas and Arizona also saw significantly higher numbers of installations.

California incentivizes homeowners with solar panels to also install batteries by offering better pricing for residential electricity exported to the grid after sunset, Bloomberg reported. Hawaii offers a one-time payment of $400 for every kilowatt hour of battery storage that homeowners install.

However, the record-breaking home battery installations coincided with a slowdown in residential installations of solar panels—the result of the Trump administration and Republican-driven One Big Beautiful Bill having eliminated a 30% federal solar tax credit for homeowners. Nonetheless, US electricity generation from solar power continues to rise and even surpassed coal-fired generation in April.

The battery installation spree also coincides with rising electricity costs for US residential customers.

«

It would be good if Ars Technica’s writers knew the difference between a kilowatt (a measure of instantaneous power) and a kilowatt hour (a measure of energy), as the original of this story omitted the “hour” from its measurements.

Anyway, more microgeneration and microstorage all works in everyone’s favour.
unique link to this extract

Bye bye, Beeb? • The Critic Magazine

Christopher Snowdon:

»

The BBC is no longer a public good as an economist understands the term. Public goods are non-excludable: you cannot prevent people from using them. When all you needed was a TV aerial, the BBC could only make you pay to watch its programmes by sending you threatening letters. By switching to a Netflix-style subscription model, the BBC could exclude those who do not pay.

This is the obvious solution and it needs to be done soon because the threatening letters are no longer working. Since 2014, the TV licence fee “evasion rate” has risen from 5% to 12.5% and revenue from the licence fee has fallen by 28% in real terms; hence another round of cuts.

As David Elstein noted in last month’s issue of The Critic, prosecutions for licence fee evasion have fallen from 150,000 a year to around 25,000 and the average fine is barely more than the cost of a licence. It is an extremely low risk crime and the BBC wastes £160m a year trying to tackle it. As the word spreads that paying the licence fee is essentially optional, more and more people will become “evaders”. 

Some of these people genuinely never watch live broadcasts and never use the iPlayer, but it is safe to assume that most of them are simply saving money. How many of them would cough up £180 if it was the only way they could access BBC content is the big question and it will become more pressing if licence fee evasion spirals, as seems likely. 

As Elstein says, the BBC’s fierce opposition to moving to a subscription model “borders on the irrational”. The corporation fears that if people were given a choice between paying for the BBC and not watching the BBC, millions will decide to stop paying for the BBC. If so, that is not our problem. The BBC is not the army. It is not the police. It provides entertainment, and no one should be compelled to pay for entertainment they do not consume. 

…This is where the opportunity lies. There are two billion televisions in the world and only 50 million of them are in the UK. It might take a bit of trial-and-error to discover the revenue-maximising price for a global BBC subscription, but the corporation should have enough faith in itself to believe it can take a large enough sliver of the overseas market to make up for any losses in the domestic market.

«

Netflix has around 325m subscribers. It might take the BBC a while to get to that level, but it would be a behemoth if it could.

unique link to this extract

Daily Mail sued over ‘systematic’ lifting of social media images • Press Gazette

Charlotte Tobitt:

»

The Daily Mail is being sued over alleged “systematic” lifting of “thousands” of images from social media without permission or payment.

A photographer named Matthew Moore who said one of his own images had been taken without permission has launched a class action lawsuit in New York on behalf of himself and anyone else in the US who has been affected in the past three years.

The legal complaint calls this behaviour “standard practice” at the Daily Mail and is seeking a declaration that it violates US copyright law.

The legal complaint alleges that lawyers found 107 articles that contained at least one photo that had been taken from social media with only a credit to the platform (for example © Instagram) over the course of just nine days in June.

It claims that assuming a payment of between $2,500 and $25,000 per violation of the US Digital Millennium Copyright Act, this rate of alleged infringement would open the Daily Mail up to “more than ten million dollars per year”.

The claim argues that the Daily Mail ensures its use “stays hidden” to the copyright owners by adding a “false credit” to the social media platform the image was taken from.

This is despite the fact the platforms do not own the copyright according to their own terms and conditions, it says.

But it means that rightsholders searching for their name or using credit alerts may never see the usage.

«

I certainly know of one person who threatened to sue the Daily Mail of its use of their pictures without permission, which led to an apology and a payment to charity. It’s hard to see how the Mail can successfully defend this one, and there’s then the question of what it does in future if that happens.
unique link to this extract

Substack reveals newsletters with most paying subscribers in UK • Press Gazette

Charlotte Tobitt:

»

Substack has revealed the 50 newsletters on its platform with the most UK paying subscribers for the first time.

Michael Howell’s Capital Wars is the top Substack newsletter in the UK based on number of paying subscribers. Capital Wars is currently the fifth-ranked finance newsletter across Substack overall and is written by Howell, a former research director at investment bank Salomon Brothers.

Finance is the biggest topic vertical in the top 50, accounting for 13 publications. Other topics represented multiple times include food and drink, fashion and beauty, news/politics and history.

In second place is Exponential View, an AI newsletter by tech entrepreneur Azeem Azhar which passed 100,000 free and paid subscribers two years ago after nine years of publishing (plus a further 200,000 who received it via Linkedin).

Third is Comment is Freed, which has more than 90,000 free and paid subscribers, by Institute for Government senior fellow Sam Freedman and his father Lawrence Freedman (emeritus professor of war studies at King’s College London).

Substack already published a “bestseller” list in the US, and has now started a separate list in the UK. It counts only paying subscribers, meaning major newsletters without paywalls may be overtaken by smaller brands with higher conversion rates.

Former Guardian media editor Jim Waterson’s local newsletter London Centric, which he started in September 2024 using a voluntary redundancy payment from the national newsbrand, has the eighth most paying subscribers in the UK on Substack.

Waterson said in May that London Centric had about 5,000 paying subscribers and was bringing in enough revenue to support two additional staff members.

«

Excellent news about London Centric, which might get a flywheel effect going as it hires more staff. And Comment Is Freed, which charges the lowest possible amount (£35 per year), is surely generating a nice bit of income for the Freedman household(s). For all the hate aimed at Substack by some people, it has enabled new models of successful journalism that really couldn’t work before.
unique link to this extract

Grok’s traffic heavily driven by NSFW content, report says • Forbes

Mary Whitfill Roeloffs:

»

Elon Musk’s xAI is reportedly leaning into explicit content generation as a core driver of its Grok chatbot traffic and adult content now accounts for the majority of the platform’s activity, according to a Wednesday report from The Information.

The report claims xAI is actively doubling down on its explicit video and image-generation tools and that adult-content dominance extends into Grok’s coding model, which The Information reports frequently receives requests for pornographic material.

Well over half of Grok’s overall traffic is driven by pornographic images and videos, adult role-play chats or other such activity, according to the report, which Vital Knowledge analyst Adam Crisafulli called “a desperate attempt for relevancy.”

The move comes as xAI has “fallen further behind” competitor chatbots from Anthropic, OpenAI, Google and Meta, per Crisafulli, and Grok recorded the largest drop in web traffic of any single AI model this year, Similarweb data shows.

22%. That’s the drop in Grok web traffic between January and May, according to Similarweb, more than any other major chatbot. Similarweb’s data doesn’t include interactions with Grok through X, Musk’s social media site.

«

Unsurprising that Musk thinks there’s better money to be had hiring the servers powering Grok out to literally anyone else who wants them.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Does sunscreen cause skin cancer? The fallacious suggestion it does has caught hold online; but it’s like saying wet streets cause rain. CC-licensed photo by Godverbs on Flickr.

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

A selection of 9 links for you. Slip, slap, no slop. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.

Apple ‘Hide My Email’ vulnerability reveals peoples’ real email addresses • 404 Media

Joseph Cox:

»

A vulnerability in Apple’s “Hide My Email” tool lets almost anyone discover a person’s real email address that is supposed to be hidden by the feature, and Apple has failed to fix it for more than a year, according to a security researcher and 404 Media’s own tests.

404 Media is not revealing the exact details of the vulnerability because it can still be exploited as of Monday, when 404 Media verified the issue with one of our own hidden email addresses.

”Apple Hide My Email is leaking email addresses that are supposed to be hidden. We reported the issue and replication instructions to Apple over a year ago. We don’t know why it hasn’t been fixed, but we don’t feel comfortable waiting any longer. Hide My Email users deserve to know that it may be possible for attackers to discover their hidden email addresses,” Tyler Murphy, the co-founder of EasyOptOuts, which discovered and reported the issue to Apple, told 404 Media.

“Free, publicly accessible people-search sites make it easy to link an email address to other personal details, so people relying on Hide My Email for safety may be at risk,” Murphy added.

…To test the issue I generated a new Hide My Email address and provided it to Murphy. Around five minutes later, he replied with my real email address linked to my Apple account which was supposed to be hidden.

“We don’t know the full scope of the issue, but in our limited tests with volunteers, 100% of Hide My Email addresses were exploitable,” Murphy said.

«

As it happens, Apple said at the end of May that it was going to tweak the system. But it still hasn’t.
unique link to this extract

Xbox testing disc-to-digital feature that digitizes a physical game collection • The Verge

Tom Warren:

»

Microsoft will likely soon follow Sony and stop the production of physical discs for Xbox games. But instead of leaving physical discs behind entirely, sources familiar with Microsoft’s plans tell me the company has quietly been working on a disc-to-digital feature that will allow Xbox owners to digitize their existing physical game collections.

Xbox employees recently started testing this new feature, after references to “enable Disc2Digital” appeared in the Xbox PC app code in May. I’m told that Microsoft’s disc-to-digital feature will work on Xbox One and Xbox Series X discs only, and not those for the Xbox 360 or original Xbox console.

Getting a digital copy of a game works simply by inserting a compatible disc and installing and playing the game. This will require a Microsoft account on an Xbox console and will grant a digital entitlement for physical games. This digital entitlement is tied to the specific disc, and it will move from account to account if you swap the physical game with a friend or log in to a different Xbox profile and try to play a disc-based game.

«

OK, but how about a digital-to-disc feature, so that you’re not reliant on being online or having tons of (increasingly expensive) storage to store your games?
unique link to this extract

Sony will kill PlayStation games on discs in 2028 and offer digital downloads only • AFP via The Guardian

»

Sony said on Wednesday that it would stop releasing new video games for the PlayStation console on disc in January 2028 following a shift in consumer preferences.

“Following this date, new games will be available on PlayStation Store and at retailers in digital formats only,” the company said on its official PlayStation blog.

In practice, that means gamers will have to download directly from Sony’s PlayStation store or obtain a download code when purchasing a title from a retailer.

The announcement comes as the upcoming exclusively digital release of Grand Theft Auto VI, which is predicted to become one of the biggest-selling cultural products of all time, has caused some consternation among gamers.

There was grumbling on social media that the lack of a physical disc would eliminate any secondhand market for the title. Sony said the upcoming shift “has no impact on games that already released, or will be releasing, prior to January 2028 in disc format”.

Sony began its move towards digital downloads in 2020 with the release of the latest console, PlayStation 5, which had a version without a disk drive.

“This is a natural direction for Sony Interactive Entertainment to adapt to consumer trends as the general preference for digital media significantly outpaces physical discs,” the company said. “We remain committed to delivering a world-class gaming experience to our fans.”

«

Along with the coming Xboxcalypse, this feels like a hinge moment. Though Steam has in effect been doing this for years.
unique link to this extract

Fact-checkers reckon with shrinking budgets and growing AI threats at GlobalFact 2026 • Nieman Journalism Lab

Andrew Deck:

»

In 2024, the Trump administration shut down the U.S. Agency for International Development (USAID), which had given grants to fact-checking organizations around the world. Last year, Meta ended its third-party fact-checking program in the U.S. — another significant revenue stream for IFCN signatories — claiming it was an effort to reduce “censorship.”

While Meta has yet to shutter its international third-party fact-checking program entirely, it has made clear its intentions to follow in the footsteps of X and disinvest in professional fact-checks in favor of crowdsourced “Community Notes.”

One signal of big tech’s waning support for fact-checkers was how few major tech companies were in the room at GlobalFact. While Meta and Google have historically sent representatives annually, only TikTok dispatched a proper delegation this year.

“The political wind shifted, and their support shifted with it. To the technology platforms not in this room today, rejoin us in the work of making high-quality, accurate information accessible to everyone,” Angie Drobnic Holan, director of IFCN and a 2023 Nieman Fellow, said in her own opening remarks. “Fact-checking is not censorship. It is not partisan. It never was.”

…“I’m not super at math, but I’m good enough to know you need more funds coming in than going out,” Holan said in a panel about alternative funding models. “Philanthropy has told me there is not enough philanthropic money to support your entire sector. There’s just not. But there’s also evidence to show that there’s a substantial pool of audience revenue.”

In April, IFCN’s annual State of Fact-Checkers report found that 62% of the 141 organizations surveyed had grown their audiences in 2025, but just 22% described their financial position as “sustainable.”

«

Depressing. (Thanks Gregory B for the link.)
unique link to this extract

Getty Images scraps $3.7bn merger with Shutterstock over UK scrutiny • Reuters

Jaspreet Singh:

»

Getty Images said on Tuesday it has called off its planned merger with Shutterstock due to the UK competition regulator’s requirement to ​sell Shutterstock’s editorial business as a condition for approval.

Two of the largest ​players in the licensed visual content industry announced the deal in January 2025 to create a $3.7bn stock-image powerhouse geared for the AI era.

The collapse ​of the merger comes as both companies face growing competition from AI image generators ​that offer a cheaper and easier way to create visuals.

“We are not convinced that scale would have done more than stave off competitive pressures for a little while longer, but without the ​scale that the merger would bring, the outlook for each looks even more ​difficult,” said Luke Stillman, a managing director at trend advisory firm Madison and Wall.

Shares of Getty were up about 1.1% at $0.87 in volatile extended trading, while those of Shutterstock plunged about 29% to $9.95.

Britain’s Competition and Markets Authority in May conditionally approved the merger, requiring Shutterstock to sell its editorial arm to address concerns over the supply of news content ​in the country. The regulator’s ​independent inquiry group had found that the editorial business, if not sold, would reduce choice for UK media outlets and could ultimately raise prices ​for customers, as Shutterstock is one of the “few meaningful” rivals ​to Getty.

…Getty, which competes ​with Reuters and the Associated Press in providing ​photos and videos for editorial use, said its board also plans to engage a financial adviser to explore strategic ​financing options for the company.

«

The specific editorial businesses to be divested were Rex Features, Splash News and Backgrid. It’s a little unclear whether Shutterstock would have to sell the picture libraries too: presumably, yes.
unique link to this extract

No, sunscreen doesn’t cause skin cancer • Full Fact

Leo Benedictus:

»

A number of social media posts are wrongly claiming that using sunscreen raises people’s risk of skin cancer. These claims are based on a classic misunderstanding.

The posts, which we’ve seen on Facebook and on Instagram, all cite an article on a website called the People’s Voice that we have fact checked many times before. This article describes a study from 2023 that used data from the UK Biobank to compare skin cancer rates in different groups of people.

The study in question found that people who were frequent or very frequent users of sunscreen were also more likely to develop skin cancer. From this, the website article and social media posts conclude that “using sunscreen massively increases the risk of three major types of skin cancer”.

But this doesn’t follow at all—and it isn’t true. Instead it’s a clear case of what’s sometimes called “confounding by indication”.

For example, people who are about to visit malaria zones are more likely to take anti-malaria tablets—and also more likely to catch malaria. People experiencing hair-loss are more likely to take hair-loss treatments—and to lose more hair. And of course people who are about to go in the sun are more likely to use sunscreen—and also more likely to develop skin cancer.

This doesn’t mean that using sunscreen causes skin cancer. It means that something else—sun exposure—leads to both. So people who use more sunscreen are probably exposed to more sunlight, on the whole, which puts them at higher risk of skin cancer that the sunscreen doesn’t completely prevent.

«

It’s a “wet streets cause rain” sort of mistake.
unique link to this extract

America at 300: Imagining the next half-century of change • The Washington Post

Joel Achenbach:

»

In April 1976, three months before the American bicentennial, two guys no one had ever heard of formed a little company called Apple.

They were building what were called “microcomputers.” The traditional tech companies of the time built large mainframe computers. Small computers were of interest to hobbyists. They were a bit like toys. But the two nobodies, Steve Jobs and Steve Wozniak, who were working out of a house in suburban California, believed they could make microcomputers that nontechnical people would find useful in everyday life.

“We didn’t get the flying cars that ‘The Jetsons’ promised us,” said Margaret O’Mara, a historian at the University of Washington whose book “The Code” describes the history of Silicon Valley. “But we walk around with supercomputers in our pocket, and these supercomputers were invented by two long-haired, vegetarian college dropouts who didn’t bathe very often and worked out of a garage.”

The future was hurtling toward us at the bicentennial in forms we didn’t see coming. That’s the nature of the future: It’s sneaky, disorderly and can’t be tamed.

Here’s a smart bet: The next 50 years of science and technology will be even wilder than the past 50.

Among the things we might see are permanent bases on the moon and Mars, data centres orbiting Earth, fusion reactors feeding the grid, quantum computers doing calculations in minutes that would take classical computers years, “social robots” with hands as soft as our own, people playing pickleball well into their second century, and workers commuting to the office via things you could plausibly call flying cars (finally!).

But futurists don’t actually make predictions. They just craft scenarios. Some are sunny. Some are gloomy. Some are … pitch-black.

«

• Permanent bases on the Moon and Mars: no. (At least, not with humans.)
• Data centres in orbit: yes.
• Fusion reactors: I’ll go with no (it’s been correct for the past 50 years when everyone was sure it was 10 years away.)
• Quantum computers: maybe? But see fusion reactors.
• Social robots: yes.
• Pickleball centenarians: probably already exist in Florida.
• Flying cars: no. Have you seen how people drive?

There’s more sense in the main article, but that preamble is annoying. Not AI helpers everywhere? That’s a certainty.
unique link to this extract

Silenced • How To Survive the Broligarchy

Carole Cadwalladr:

»

Facebook’s use of an arbitration court to silence Sarah Wynn Williams [author of Careless People, an account of working for Facebook from 2011-2017] is not an accident. There’s no shortage of courts in either the US or UK, where Wynn Williams now lives. Criminal courts to try criminal cases and civil courts to settle business and other disputes.

But arbitration courts are an entirely parallel system. It’s commonplace for Silicon Valley companies to force them as a procedure for settling disputes not just in severance agreements but in joining agreements. Almost everyone who goes to work for a social media platform signs a contract on joining that prevents them from ever disclosing confidential information about what happens inside the company before they’ve even started their jobs.

It’s one reason why vanishingly few people have ever spilled the beans. The other is that they’ve drunk the kool aid/had their mouths have stuffed with gold/been so deeply compromised they’re in a state of denial or in Nick Clegg’s case all three.

It’s an ongoing source of pain to me that the former deputy prime minister and latterly Facebook’s head of policy and spin, is considered a plausible voice on anything to do with Silicon Valley.

…Tech companies use arbitration courts because they can. Just as they write the terms of service that we the users have no choice but to accept, they also write the contracts that employees must sign to take a job. It’s a legitimised form of corporate bullying, a non-state parallel justice system. And now one that Sarah Wynn Williams finds herself trapped in.

And it maps onto other networks, one used by criminals and gangsters. I found myself investigating arbitration courts, some years ago, when I was trying to understand why and how Cambridge Analytica was seeking to set one up in St Kitts and Nevis.

I never uncovered a definitive answer to that but I did speak to a lawyer who’d been involved in the scheme. What are some of the dodgier possible use cases of arbitration courts, I asked him. “Well, money laundering for one,” he said. It’s a simple matter to set up a fake arbitration to settle a fake dispute. Party A seeks resolution to a dispute with Party B, a judge finds in Party B’s favour and Party A has to pay them a load of cash. The money is legitimately washed through another company’s accounts.

Another scenario is where individuals use these courts to seize assets by illegitimate means. Nice company you keep, Facebook.

«

Recommended Wynn Williams’s book to a friend, who bought it. A little poke in the eye for Facebook.
unique link to this extract

Inside Apple’s chipflation dilemma • Culpium

Tim Culpan:

»

Apple had tried for months to save money, my sources tell me. But skyrocketing DRAM costs have driven the price of memory from $30 per phone to over $130 apiece, bringing it above 30% of the cost of building an iPhone, according to my discussions with sources.

In reality, Apple’s decision to add $100 to a MacBook Neo, $300 to a MacBook Pro, and $500 to a Mac Studio shows Apple’s waning position in the global supply chain for technology hardware.

…Apple’s price hike last week was not sudden, it was not rash, and it wasn’t Tim Cook displaying some fit of pique.

It was calculated, planned, and executed over a period of months. Apple could see it coming as far back as six months ago, my sources tell me, and has been preparing the groundwork ever since.

And the messaging was carefully stage managed. First was the timing. Cook, as CEO, or his CFO Kevan Parekh could have signaled the price rise back in its April 30 earnings call, or they could have done so in the next event around a month from now. But they chose the end of June, when the message wouldn’t get lost.

Then they chose the medium. In this case, an interview with The Wall Street Journal’s Rolfe Winkler, the same reporter who got exclusive tours of Apple supplier facilities including TSMC, Foxconn, and GlobalWafers back in February.

…Memory chips aren’t the only thing that’s become more expensive over the past year. We already know that logic chips have gone up in price, which includes not only core A-series and M-series processors but also microcontrollers, power-management, networking, and display drivers. There’s also upward pressure on magnets, exterior casings, and even batteries.

But Cook chose to blame memory. Memory takes the rap for two reasons. It’s the component he has the least control over. More importantly, according to my discussions across the supply chain, memory’s contribution to Apple’s total components costs will climb to as much as one-third of the bill of materials for many devices during calendar 2026. And Cook could do nothing to stop it.

«

unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Two papers by Max Planck, a pioneer of quantum mechanics, have been removed from their online journal. There is uncertainty about why. CC-licensed photo by Julia Tulke on Flickr.

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

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

“Easily the biggest leak in Apple’s history”: iPhone 18 Pro final design may have just been revealed in a stolen drop test video • Tom’s Guide

Jeff Parsons:

»

Apple is obsessed with preserving as much secrecy as it can before officially revealing the rumoured iPhone 18 Pro, iPhone 18 Pro Max and iPhone Ultra/Fold at the Apple Event in September. But the company has just been dealt a huge blow by a data breach at one of its suppliers.

More than 200,000 files are believed to have been posted to the dark web following the ‘cybersecurity incident’ Tata reported last week.

Photos, videos and component lists of the upcoming devices have been circulating on social media after one of Apple’s India-based suppliers, Tata Electronics, suffered a data breach last week. More than 200,000 files are believed to have been posted to the dark web following the ‘cybersecurity incident’ Tata reported last week.

These files include documentation on other products made by Tata, but according to Reuters, drawings of the iPhone 18 Pro circuit board, A20 chip, and supplier lists for components are also among them.

They also include drop test photos, which is part of customary testing for durability. While it’s not been confirmed whether or not any videos are among the leak, one such video is being circulated on social media. Tom’s Guide hasn’t been able to verify if this video is accurate or merely AI-generated. If genuine, it would give us some huge clues about what to expect in September.

«

The phone looks like.. an iPhone. Not the least surprise about that. But Apple will be seething about two points: that the leak has come from Tata, one of its new partners outside China; and that the component list has been leaked. In April 2021 a ransomware attack against Apple’s laptop maker, Quanta, led to the leak of forthcoming designs.

and in January Apple supplier Luxshare was hit, leading to the leak of lots of internal data.
unique link to this extract

The Dogs of San Francisco: 51,379 dogs

Ryan McEntush and Luke Eigel:

»

Every licensed dog, month by month. After a long slide from 2017, the registry rebounded to a record 11,200 in 2025. Licenses still spike each spring, peaking in May 2025.

«

Didn’t know that dogs had to be licensed in San Francisco, but it turns out that means you can generate a database showing all sorts of fun details, particularly about breeds. Large dogs turn out to be surprisingly popular, though maybe the weather is cold enough there to be tolerable for them. A fun project, harking back to the halcyon days of Web 2.0.
unique link to this extract

An entire Herculaneum scroll has been read for the first time • Vesuvius Challenge

»

For almost 2,000 years, the carbonized library of Herculaneum has kept a cruel bargain: its scrolls survived the eruption of Mount Vesuvius, but only by becoming too fragile to open. To read one was to destroy it. Hundreds of rolls have therefore remained sealed, their contents preserved yet unreachable.

Today that changes. We have completely virtually unwrapped and read PHerc. 1667 — the scroll the Vesuvius Challenge community knows as Scroll 4 — without ever touching its pages. It is the first Herculaneum papyrus to be digitally unrolled and read in full, end to end, and made available for sustained scholarly study.

PHerc. 1667 began as a blackened, rolled mass of carbonized papyrus. To read it, we never unrolled it physically. Instead, we scanned it with high-resolution X-rays, reconstructed the wound sheet inside the volume, flattened it into a readable surface, and used machine learning to bring out the faint traces of ancient ink.

…PHerc. 1667 is what survives of a larger roll: earlier attempts to open it by hand — in the 19th century, and again in 1969 and the 1980s — destroyed its outer layers and left only the compact inner core, about 8 cm of an original height of 19–24 cm. From that surviving portion we have now recovered and read the text in full — the lower parts of some 22 columns, transcribed and reviewed by papyrologists. It is the first time the preserved text of a rolled Herculaneum scroll has been read continuously, end to end, rather than in isolated words or patches.

The recovered text is a philosophical treatise on ethics, and the evidence points to a Stoic work: it turns on human nature, impulse, and the moral progress of human beings, and its final preserved column names Aristocreon — nephew and disciple of the great Stoic Chrysippus — which, together with the language and themes of the text, places it in a Stoic context and dates it to the 2nd century BC.

Because the papyrus is damaged, the readings are fragmentary, with gaps where the surface is lost. Even so, several passages can be read clearly for the first time in two thousand years.

…The scans were acquired with high-resolution phase-contrast X-ray microtomography on the BM18 beamline at the European Synchrotron Radiation Facility (ESRF) in Grenoble — an instrument able to resolve the wafer-thin, densely packed layers of a Herculaneum roll. The work was carried out in collaboration with the National Library of Naples “Vittorio Emanuele III”, which safeguards the Herculaneum papyri. From those volumes, the team reconstructed the scroll’s geometry, traced and flattened its surface into a readable sheet, and trained machine-learning models to detect ink that is almost indistinguishable from the carbonized papyrus beneath it. Each reading was then examined and transcribed by papyrologists.

«

unique link to this extract

About those “hackquisitions”… • Spyglass

MG Siegler:

»

The news that Noam Shazeer is (once again) leaving Google seems like a big deal. The news that he’s joining OpenAI, which turned the transformer paper he helped write into a product that he couldn’t launch (in his first stint) at Google seems like an even bigger deal. Bigger still may be the fact that he had rejoined to help the Gemini product take on ChatGPT, which was seemingly working, at least to some degree. But actually, the biggest deal has to be the actual deal that brought him back to Google. Because it wasn’t even two years ago when Google paid $2.7B to bring Shazeer back.

And like that – poof – he’s gone.

To be fair, there were others on the Character.ai team that Google seemingly wanted too. The non-exclusive licensing rights for Character? Probably less so. If anything, that aspect of the deal has ranged from a headache to a nightmare.1 But clearly it was a deal structure in such a way to get Shazeer back with an offer he couldn’t refuse. And he didn’t. Until he did. Again.

That deal structure, of course, was one of the early “hackquisitions” – a deal to bring on a company’s key talent without acquiring the company itself. Because that clearly would have been messy from a regulatory perspective for any of Big Tech. If nothing else, such deals would be bogged down for months while they’re scrutinized. A “hackquisition”, by contrast, could be done almost instantly.

«

Siegler goes into quite some details about all the hackquisitions that are going on among all the AI companies. It’s a lot, and there seems to be an absolute revolving door between all the different companies. But also: it’s hard to see whether AI progress is really reliant on any particular one of these researchers. Is it like a football team, where individuals matter but it’s the team that makes it, or like tennis players, where the individual is what counts?
unique link to this extract

When cybercriminals hire burglars: inside an alleged Russian effort to infiltrate multibillion-dollar US law firms • CNN Politics

Sean Lyngaas:

»

When an executive at a US law firm’s phone rang in April, the voice on the other end was urgent: A computer virus was spreading through the firm.

The caller said they were from IT support and needed physical access to the lawyer’s computer because remote fixes to stop the attack weren’t working. The lawyer told his purported colleague to swing by his desk at the law firm’s office in New Jersey.

The next day, the firm’s receptionist called: The lawyer had a visitor from IT at the front desk.

“That’s when an alarm bell went off: Why would an IT person need to check in with reception?” said Leeann Nicolo, who handles incident response for cybersecurity insurance firm Coalition, which the law firm hired to investigate the incident.

The visitor ran out of the building when the lawyer approached the front desk, according to Nicolo.

It’s one of several incidents at law firms across the country in the last year in which, the FBI and private investigators suspect, the Russian-speaking Silent Ransom Group has hired people in the US to show up in-person and plug thumb drives into law firms’ computers. The physical access could help bypass anti-virus protections that the hackers run up against from afar.

The group’s millions of dollars in returns contrasts with its modest investments: In a private Telegram channel, the group is offering $500 to people to visit law firms and plug in USB sticks, one cybersecurity professional familiar with the incidents told CNN.

The hired hands are “cannon fodder” for the Russian-speaking cybercriminals — expendable assets in a much larger cybercrime war, the source said.

«

The scheme is: get data about the law firms’ clients, leak it if the companies won’t pay a ransom after they’re hit by ransomware.
unique link to this extract

Why have papers by one of history’s most famous physicists been retracted? • Science

Sam Kean:

»

In early May, Yves Gingras, a historian of physics at the University of Quebec (UQ) at Montreal, was browsing Retraction Watch, a website that catalogs fraud, data manipulation, and other scientific sins. He noticed a link that read “Retractions by Nobel Prize winners.” Were there really Nobel laureates whose papers had been withdrawn from the scientific literature?

After clicking, Gingras froze. “That’s impossible,” he recalls thinking. The fourth name on the list, with two retracted papers, was Max Planck—a legendary pioneer of quantum mechanics and the 1918 Nobel laureate in physics. Gingras had never heard a whiff of scandal about Planck, who was almost as widely revered for his character as his physics. In 1933, for example, he bravely confronted Adolf Hitler over Nazi Germany’s discriminatory laws against Jews.

Gingras called up Mahdi Khelfaoui, a fellow historian of science at UQ Trois-Rivières. “There’s something fishy,” Gingras said. The papers, both quietly retracted in 2011, originally appeared in the early 1940s in Naturwissenschaften, a German journal now owned by publishing giant Springer Nature. After some sleuthing, Khelfaoui determined one of the Planck pieces, a philosophical essay from 1942 titled “Sinn und Grenzen der exakten Wissenschaft” (“Meaning and Limits of Exact Science”), about how to achieve certainty in scientific knowledge, had also appeared in two other journals and been reprinted twice in books.

Repackaging the same work multiple times is considered “self-plagiarism” and frowned upon today—the practice produces copyright conflicts and inflates scholars’ publication records. The Naturwissenschaften site gives “copyright violation” as the reason for the retraction.

Yet publishing identical material in multiple journals was widespread before the internet. “Science was more fragmented” then, Khelfaoui says. “You wanted different audiences …  to have access to your work.” The practice was especially common for luminaries like Planck. Albert Einstein did the same (but escaped retractions).

Springer Nature’s “anachronistic” application of modern standards to a 1942 paper “distort[s] the historical record,” Gingras and Khelfaoui argue in a preprint posted last month on arXiv.

«

Seems that a bot thought it was copying somewhere else, and removed it. But:

»

Springer Nature was nevertheless still selling the empty PDF for $39.95 until this story was published.

«

Never change, academic publishers.
unique link to this extract

Spurious copyright claim sees second Press Gazette story removed from Google search • Press Gazette

Dominic Ponsford:

»

A Press Gazette article exposing the dubious ‘parasite SEO’ tactics of online marketing company Clickout Media has been removed from Google search results after a spurious anonymous legal challenge.

A mysterious entity called DRF Corp wrote to Google stating Press Gazette had “willfully violated copyright law by copying our entire content word for word, including all images, which are solely owned by our company” even though the content allegedly copied was on an unrelated subject.

It is the second time a fake copyright claim has been used to get Press Gazette reporting about Clickout Media removed from search engine results this year.

According to the Lumen database, the complaint was filed under the US Digital Millennium Copyright Act.

It alleged that the original article was a month-old (now removed) Reddit post headlined: “Casinos Not Gamstop in 2026: The Brutally Honest Truth Before You Deposit.”

The Press Gazette article removed by Google from search results, which was published last week, details how Clickout Media has bought three reputable UK sports news websites and introduced AI-generated reporters whose stories are littered with errors and fabrications. It is headlined: “AI reporters churn out error-strewn stories for football websites.”

This article no longer appears in Google search results. Any search that would have previously surfaced the story now includes the following disclaimer from Google: “In response to a complaint that we received under the US Digital Millennium Copyright Act we have removed results from this page.”

«

Odd how the Reddit post has been removed, which makes it harder to verify whether it is indeed exactly the same as the Press Gazette story. (It isn’t.) This adds more detail to the story yesterday about Clickout Media. There’s more to come about this company and its partners. Press Gazette might just be annoyed enough to go and find it.
unique link to this extract

Pollen tried to remove my article about CEO Callum Negus-Fancey and CTO Bradley Wright, and Google is assisting with it • The Pragmatic Engineer

Gergely Orosz:

»

In 2022, I wrote about the damning fall of events tech company Pollen. The short of it:

»

Pollen seemed to have pulled off the improbable feat of building a business in the notoriously low margin industry of events, surviving Covid-19, and building a solid software engineering organization. In April this year, the company announced it had raised another $150M in fresh funding.

But just three weeks later, Pollen laid off about 200 people, a third of staff. Leadership assured employees all was well. However, from that point on, things got worse. Leadership later pulled the plug on Slack, employees were not paid wages, pension contributions went missing, and vendors were not paid. Some vendors took matters into their own hands; on 9 August 2022, JIRA was suspended when Atlassian tired of the company’s failure to pay.

On 10 August 2022, Pollen went bankrupt, collapsing into administration.

«

The article looked bad on Pollen’s founder, Callum Negus-Fancey. He was ultimately responsible for lying to staff, not paying salaries, the missing pension contributions, and the unpaid health insurance for US employees. The story was so bad that the BBC created a documentary titled Crashed: $800M Festival Fail. 

And then there was the $3.2m double charge for customers, manually initiated by CTO Bradley Wright, detailed extensively in the documentary Crashed: $800m Festival Fail. That double charge would have been trivial to reverse, but the reversal never happened, customers never got their money back, and the postmortem of the incident was never released to staff.

Four years later, Pollen and Callum Negus-Fancey are attempting to erase this shameful story from the public record. The article is my original writing, and thus I am the copyright holder of it. So imagine my surprise when I was notified that Google removed the article from its search results thanks to a copyright infringement claim it received.

«

Guess what? Google removed his article from its search results, based on a copyright claim made by an unknown owner (from an uninhabited country) against an unspecified source. Orosz eventually found the complaint, which asserts that it’s a copy of a New York Post article. It isn’t (that article has a URL ending “band-leader-hits-winning-chord”). Google might want to look at how it’s being abused.
unique link to this extract

Bracing for layoffs, unionized Xbox developers hold press conference to make their point • IGN

Cade Onder:

»

Unionized Xbox employees are pushing back against the company’s looming layoffs and have outlined various demands.

Earlier today, the CWA (Communications Workers of America) held a press conference in which various unionized Xbox employees spoke out against Microsoft. The conference was held ahead of reported layoffs at Xbox, which insiders have stated will be a “bloodbath.”

It’s also a painful reminder of last year’s layoffs at Microsoft, where 9,000 people lost their jobs across the entire company (not just Xbox) and resulted in multiple projects being cancelled, including the long-awaited reboot of Perfect Dark. The studio behind that game, The Initative, was also shuttered without having ever released a game.

The upcoming layoffs reportedly puts more studios, such as Double Fine and Ninja Theory, at risk of closure. South of Midnight developer Complusion Games is also reportedly at risk of shutting down, despite winning a Peabody Award earlier this year. New Xbox boss Asha Sharma celebrated the win on her socials, months before the studio’s reported demise: “A well-deserved recognition for storytelling that truly matters!”

…It remains to be seen what will come of this, but Activision QA tester Andrew Snell and [Diablo senior environment artist Mahreen] Fatima both made it clear that Microsoft’s actions don’t just impact workers, but also the players: “Workers and players are on the same side of this and we’re done paying for executives’ failures,” said Snell.

“We, the developers, demand that you respect our labor and our games,” added Fatima. “Together we’ve built a huge community and touched the lives of millions of gamers everywhere. Don’t disrespect the developers. Don’t disrespect the gamers.”

«

This really does look like it will be bloody. The quarter ended on Tuesday, so now Microsoft can announce any cuts it feels like making in Xbox. They’re expected to be hefty.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified