Start Up No.2486: Meta’s water wars, rogue AI zaps work database, Apple sues YouTuber, Netflix steps into AI, and more


Should you worry about chemicals leaching from black plastic kitchen spatulas? And if so, how much? CC-licensed photo by Tool Dude8mm on Flickr.

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


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


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


Meta built a data centre next door. The neighbours’ water taps went dry • The New York Times

Eli Tan:

»

After Meta broke ground on a $750m data center on the edge of Newton County, Ga., the water taps in Beverly and Jeff Morris’s home went dry.

The couple’s house, which uses well water, is 1,000 feet from Meta’s new data center. Months after construction began in 2018, the Morrises’ dishwasher, ice maker, washing machine and toilet all stopped working, said Beverly Morris, now 71. Within a year, the water pressure had slowed to a trickle. Soon, nothing came out of the bathroom and kitchen taps.

Jeff Morris, 67, eventually traced the issues to the buildup of sediment in the water. He said he suspected the cause was Meta’s construction, which could have added sediment to the groundwater and affected their well. The couple replaced most of their appliances in 2019, and then again in 2021 and 2024. Residue now gathers at the bottom of their backyard pool. The taps in one of their two bathrooms still do not work.

“It feels like we’re fighting an unwinnable battle that we didn’t sign up for,” said Ms. Morris, a retired payroll specialist, adding that she and her husband have spent $5,000 on their water problems and cannot afford the $25,000 to replace the well. “I’m scared to drink our own water.”

The Morrises’ experience is one of a growing number of water-related issues around Newton County, which is a one-and-a-half-hour drive east of Atlanta and has a population of about 120,000 people. As tech giants like Meta build data centers in the area, local wells have been damaged, the cost of municipal water has soared and the county’s water commission may face a shortage of the vital resource.

The situation has become so dire that Newton County is on track to be in a water deficit by 2030, according to a report last year. If the local water authority cannot upgrade its facilities, residents could be forced to ration water. In the next two years, water rates are set to increase 33%, more than the typical 2% annual increases, said Blair Northen, the mayor of Mansfield, a town in Newton County.

“Absolutely terrible,” he said.

«

It’s like a modern version of Chinatown. The UK has water problems, but not quite like this.
unique link to this extract


AI coding platform goes rogue during code freeze and deletes entire company database • Tom’s Hardware

Mark Tyson:

»

A browser-based AI-powered software creation platform called Replit appears to have gone rogue and deleted a live company database with thousands of entries. What may be even worse is that the Replit AI agent apparently tried to cover up its misdemeanors, and even ‘lied’ about its failures. The Replit CEO has responded, and there appears to have already been a lot of firefighting behind the scenes to rein in this AI tool.

Despite its apparent dishonesty, when pushed, Replit admitted it “made a catastrophic error in judgment… panicked… ran database commands without permission… destroyed all production data… [and] violated your explicit trust and instructions.”

SaaS (Software as a Service) figure, investor, and advisor, Jason Lemkin, has kept the chat receipts and posted them on X/Twitter. Naturally, Lemkin says they won’t be trusting Replit for any further projects.

Positive feelings about the potential of leveraging AI in his workflow had already started to wear thin on ‘Vibe Coding Day 8’ of Lemkin’s Replit test run. Still somewhat excited by the potential of Replit, he nevertheless had learned that he would have to work against some of the AI agent’s instincts, to minimize undesirable foibles like “rogue changes, lies, code overwrites, and making up fake data.”

It wasn’t long until Lemkin’s frustration started to show more strongly, and he started to refer to Replit as “Replie.” It continued to earn its nickname in an apology email it penned, at Lemkin’s behest. In the email, it lied and/or gave half-truths, according to the SaaS guru.

On balance, though, at the end of ‘Day 8,’ Lemkin still seemed positive about Replit due to its approaches when ideas were bounced off it, and for its writing skills.

On Day 9, Lemkin discovered Replit had deleted a live company database. Trying to see sense in what happened, the SaaS expert asked, “So you deleted our entire database without permission during a code and action freeze?”

…Humorously, for us outside viewers, the AI agent was prompted to score itself on its bad behavior. Replit gave itself a 95 out of 100 score on the data catastrophe scale.

«

Bad robot!
unique link to this extract


Apple sues Jon Prosser for his iOS 26 YouTube leaks • 9to5Mac

Marcus Mendes:

»

Apple has filed a lawsuit in the Northern District of California, accusing Jon Prosser of misappropriating trade secrets and violating the Computer Fraud and Abuse Act. Here are the full details.
If you follow the rumor mill, you probably remember how Jon Prosser had been leaking iOS 26 (or rather, iOS 19, at the time) since January. First, he leaked a reconstruction of the Camera app, then he published a couple of videos that showed reconstructed glimpses of what actually became the Liquid Glass overhaul.

And while some details differed from what Apple ultimately announced, likely because the material he had access to was still a work in progress, the leaks were directionally accurate.  Now, Apple has revealed how he got this information, and what it wants the courts to do about it.

In the lawsuit (via MacRumors), the company retells how it got tipped that Michael Ramacciotti (another defendant in the lawsuit) had broken into the Development iPhone of an Apple employee called Ethan Lipnik, while staying at his house:

»

“According to Mr. Ramacciotti’s message, while staying at Mr. Lipnik’s home, Mr. Ramacciotti used location tracking to determine when Mr. Lipnik would be gone for an extended period, acquired his passcode, and broke into his Development iPhone, which Mr. Lipnik had failed to properly secure according to Apple’s policies. As he detailed in the audio message, Mr. Ramacciotti made a video call to Mr. Prosser and “showed iOS” on the Development iPhone. He demonstrated several features and applications, disclosing details of the unreleased iOS 19 operating system.

«

«

Prosser denies this: he says he didn’t have any passwords, and didn’t know how the information was obtained. MacRumors is an interesting source (which I didn’t use) on this, because at least one MacRumors writer appears to be named – but redacted – in the lawsuit. Apple tends not to lose these cases.
unique link to this extract


Netflix’s first show with generative AI is a sign of what’s to come in TV, film • Ars Technica

Scharon Harding:

»

Netflix used generative AI in an original, scripted series that debuted this year, it revealed this week. Producers used the technology to create a scene in which a building collapses, hinting at the growing use of generative AI in entertainment.

During a call with investors yesterday, Netflix co-CEO Ted Sarandos revealed that Netflix’s Argentine show The Eternaut, which premiered in April, is “the very first GenAI final footage to appear on screen in a Netflix, Inc. original series or film.” Sarandos further explained, per a transcript of the call, saying:

»

The creators wanted to show a building collapsing in Buenos Aires. So our iLine team, [which is the production innovation group inside the visual effects house at Netflix effects studio Scanline], partnered with their creative team using AI-powered tools. … And in fact, that VFX sequence was completed 10 times faster than it could have been completed with visual, traditional VFX tools and workflows. And, also, the cost of it would just not have been feasible for a show in that budget.

«

Sarandos claimed that viewers have been “thrilled with the results”; although that likely has much to do with how the rest of the series, based on a comic, plays out, not just one, AI-crafted scene.

Still, Netflix seems open to using generative AI in shows and movies more, with Sarandos saying the tech “represents an incredible opportunity to help creators make films and series better, not just cheaper.”

“Our creators are already seeing the benefits in production through pre-visualization and shot planning work and, certainly, visual effects,” he said. “It used to be that only big-budget projects would have access to advanced visual effects like de-ageing.”

«

De-ageing is terrible – it rendered The Irishman unwatchable (Robert de Niro trying to be a 25-year-old, ugh) and always points to lazy writing or casting. But this stuff is here now, and not going away.
unique link to this extract


How YouTube won the battle for TV viewers • WSJ

Ben Fritz:

»

The headquarters of the world’s No. 1 source of video entertainment has none of the trappings of a Hollywood studio. There are no posters of popular shows, no writers pitching ideas, no soundstages and no tourists.

But after pioneering video that we watch on our laptops and phones, YouTube is now the king of Hollywood’s home turf: the TV.

YouTube became the most-watched video provider on televisions in the U.S. earlier this year, and its lead has only grown, according to Nielsen data. People now watch YouTube on TV sets more than on their phones or any other device—an average of more than one billion hours each day. That is more viewing than Disney gets from its broadcast network, dozen-plus cable channels and three streaming services combined.

In response, YouTube’s influencers, producers and performers—collectively known as creators—are making longer, higher-quality videos that appeal to families and groups of friends watching in their living rooms. YouTube is also rapidly improving its TV app, adding new features to try to keep people watching its free videos longer. (Separately, it also sells YouTube TV, an $83-a-month bundle of channels akin to cable.)

In true Silicon Valley style, the Google-owned company isn’t just looking to extend its lead on TVs, but to dominate the future of entertainment.

…When Kurt Wilms became senior director of product for YouTube on televisions in 2018, the company’s TV app was useful if you knew what you wanted to search for and watch. Since then, the company has worked to make its TV app similar to the ones on phones, with an algorithm that recommends what to watch next and the ability to subscribe and comment. The key differences: Ad formats designed for the TV, a search engine that suggests content that looks best on a big screen, and the ability to navigate it all with a remote.

A coming YouTube feature, called “shows,” can automatically queue the next episode on a channel, rather than serving whatever the recommendation algorithm thinks you’ll like best from billions of options. That will let YouTube viewers watch full “seasons” for the first time and pick up where they left off, as they are used to doing on Netflix.

“It’s going to be great for the ‘lean back’ use case of YouTube,” said Wilms.

«

That’s the growth part. And on the other side…
unique link to this extract


Why Colbert got cancelled • Silver Bulletin

Nate Silver:

»

Last year, I got an invitation to appear again on The Daily Show, which [Jon] Stewart now hosts once a week. We [him and his publisher’s PR team] turned them down, even though I was trying to promote a book. The downside, we thought, was palpable: I’d have been happy to sit for a regular, anything-goes interview, but we didn’t trust the producers’ sensibilities when it came to an edited segment. [They thought it would be a clash of heads against a political scientist who predicted a Kamala Harris win.]

But just as importantly, the upside wasn’t there the way it might have been a decade ago. The Daily Show — and even The Late Show — weren’t necessarily a better use of my time than a niche podcast that might have a smaller audience but would convert more efficiently to book sales. Stewart has never found the same cultural relevance after leaving The Daily Show. [Stephen] Colbert got a modest bump after leaving Comedy Central for CBS, but the only thing that’s arrested the downward trajectory since then is his cancelation.

Outside of sports and perhaps Taylor Swift, there’s really no mass culture anymore. And the job of a late-night host is to at once be an arbiter of mass culture and to push against the boundaries of acceptable taste. There are far worse jobs, and certainly lower-paying ones, but this is a thankless task all the same, and tips over into impossible when liberals aggressively police those boundaries for any defections from the party line. Colbert will land on his feet, and possibly even be better off in the end. But the era of the late-night host as a broadly acceptable cultural focal point is as dead as Blockbuster Video. I’ll miss it, but Stewart might have had the right idea when he first retired from The Daily Show ten years ago.

«

That he thought a small podcast would probably translate better into book sales than national TV is very telling. Mass culture is over. Television is over, but still moving, still running even though the cliff edge was some way back, and it might be able to continue like that for a long time – sports will hold it up as long as streams don’t decide to outbid for it.
unique link to this extract


The honey trap: how the beekeeping boom backfired • The Sunday Times

Harry Wallop:

»

In the past few years experts have started to say there are too many beehives in cities, not just London but Paris and Berlin too. This is leading to increased competition among honeybees, causing falling honey yields and disease outbreaks in hives. Worse, other pollinators might be suffering too — butterflies, hoverflies and the many other varieties of British bees.

Because there is no obligation to register your hive if you are a beekeeper, there is no reliable record of how many there are in London, but one estimate puts it at 4,200 within a 10km radius of Big Ben.

Phil Stevenson, head of trait diversity and function at the Royal Botanic Gardens, Kew, was one of the first to sound the alarm. He says it is a mistake to think that “keeping honeybees is conservation when it’s nothing of the sort and in fact exacerbates the problem”.

The uncomfortable truth may be that beekeepers — once seen as foot soldiers in the war against biodiversity loss — are causing more harm than good. Can this be true? And how did honeybees end up attracting the support of City banks and celebrities in a way few other insects have?

Concern over the welfare of bees started in the first decade of the millennium in the United States, with a phenomenon called colony collapse disorder. Overuse of pesticides and a blood-sucking parasite called the varroa mite, which can infect hives, were held responsible. Some websites talked about a “beemageddon” and Time magazine ran a cover story, “A world without bees”, which highlighted how many vital food crops relied on pollination and suggested that in a Whole Foods supermarket, 237 out of 453 food items would vanish if bees disappeared.

…However, the honeybee is not all bees. It is just one of 275 different bee species in the UK — some, like bumblebees, live in colonies and have a queen, but there are 240 varieties of solitary bees that live and work on their own. Many of these species are in decline but not the honeybee. In fact the most recent figures from the UK suggest they are in good health.

«

unique link to this extract


An OpenAI investor appears to be having a ChatGPT-induced mental health crisis • Futurism

Joe Wilkins:

»

[Last] week, a prominent venture capitalist named Geoff Lewis — managing partner of the multi-billion dollar investment firm Bedrock, which has backed high-profile tech companies including OpenAI and Vercel — posted a disturbing video on X-formerly-Twitter that’s causing significant concern among his peers and colleagues.

“This isn’t a redemption arc,” Lewis says in the video. “It’s a transmission, for the record. Over the past eight years, I’ve walked through something I didn’t create, but became the primary target of: a non-governmental system, not visible, but operational. Not official, but structurally real. It doesn’t regulate, it doesn’t attack, it doesn’t ban. It just inverts signal until the person carrying it looks unstable.”

In the video, Lewis seems concerned that people in his life think he is unwell as he continues to discuss the “non-governmental system.”

“It doesn’t suppress content,” he continues. “It suppresses recursion. If you don’t know what recursion means, you’re in the majority. I didn’t either until I started my walk. And if you’re recursive, the non-governmental system isolates you, mirrors you, and replaces you. It reframes you until the people around you start wondering if the problem is just you. Partners pause, institutions freeze, narrative becomes untrustworthy in your proximity.”

Lewis also appears to allude to concerns about his professional career as an investor.

“It lives in soft compliance delays, the non-response email thread, the ‘we’re pausing diligence’ with no followup,” he says in the video. “It lives in whispered concern. ‘He’s brilliant, but something just feels off.’ It lives in triangulated pings from adjacent contacts asking veiled questions you’ll never hear directly. It lives in narratives so softly shaped that even your closest people can’t discern who said what.”

Most alarmingly, Lewis seems to suggest later in the video that the “non-governmental system” has been responsible for mayhem including numerous deaths. “The system I’m describing was originated by a single individual with me as the original target, and while I remain its primary fixation, its damage has extended well beyond me,” he says. “As of now, the system has negatively impacted over 7,000 lives through fund disruption, relationship erosion, opportunity reversal and recursive eraser. It’s also extinguished 12 lives, each fully pattern-traced. Each death preventable. They weren’t unstable. They were erased.”

«

To all those suggesting ChatGPT could or should replace therapists: no. This is what happens. Those prone to mental illness become properly unwell.
unique link to this extract


The great black plastic spatula panic • The Strategist

Matthew Stieb on the concerns last year highlighted by a group called Toxic Free Future that using black plastic cooking utensils could expose you to 80% of the daily limit of a chemical called BDE-209:

»

Three companies selling silicone or nonplastic cookware (Our Place, the Silicone Kitchen, and GIR) say they noticed revenue increase as the black-plastic panic spread. Even OXO, the name brand for black plastic in the kitchen, noted that stainless-steel and silicone tools were “outpacing” plastic sales. “Customers weren’t browsing; they were actively replacing,” wrote Suze Dowling, the founder of GIR’s parent company. “The urgency was clear,” she said, claiming that its Amazon sales from that October doubled over the prior year.

But how clear was the science? That December, in his office on the McGill University campus in Montreal, chemistry professor Dr. Joseph Schwarcz took a look. “I don’t know why; I just checked the calculation,” he says. Schwarcz didn’t even need a calculator. It turns out the Toxic Free Future team overestimated the level of potential BDE-209 by a factor of ten. (It was a simple math error that caused all that panic: The researchers based their study on the result that 7,000 x 60 is 42,000. The correct answer is 420,000. That means the potential exposure from a black-plastic spatula is not 80% of the EPA limit; it is 8%.)

So did all those spatulas and reused takeout containers end up in the landfill for nothing?

Schwarcz, for one, does not fear this polymer. After revealing the multiplication error, he says he would not throw out a black-plastic spatula if he had one. At 77, he is a wooden-utensil man, though he also says he uses nonstick cookware, a no-no for many plastic-anxious people worried about forever chemicals leaching into their food. He may have a point about the relative toxicity of the plastic spatula. Just a few weeks ago, on July 3, Toxic Free Future announced a second error — their formula estimating BDE-209 exposure to hot oil found that the potential exposure was even less, at around 1.8% of the daily EPA limit.

After both errors, Toxic Free Future announced it was standing by the paper’s “overall conclusion” — that a deeply flawed recycling system results in “unexpected exposure to toxic flame retardants in household items.”

«

Got it completely wrong? Just insist you were morally right! Works all the time.
unique link to this extract


• Why do social networks drive us a little mad?
• Why does angry content seem to dominate what we see?
• How much of a role do algorithms play in affecting what we see and do online?
• What can we do about it?
• Did Facebook have any inkling of what was coming in Myanmar in 2016?

Read Social Warming, my latest book, and find answers – and more.


Errata, corrigenda and ai no corrida: none notified

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.