Start Up No.2726: US data centres’ politics problem, OpenAI warns of AI attacks, will your monitor track you too? and more


The “solar roof tiles” originally offered by Solar City are no longer for sale. But they weren’t a real product, just a means for Musk’s Tesla to win takeover approval. CC-licensed photo by Mike Mozart 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. Warming up. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.


Data centers’ ‘oh s–t’ moment • POLITICO

Gabby Miller and Owen Dahlkamp:

»

The intensifying late-summer backlash against data centers is inspiring panic among some corners of the tech sector — amid fears that the industry is mishandling a political crisis that will stretch into 2028 and beyond.

Those worries have deepened in recent weeks as politicians across the political spectrum embraced various curbs on data center construction, with Republican Texas Gov. Greg Abbott, potential Democratic White House hopeful Josh Shapiro and Michigan GOP Senate nominee Mike Rogers being among the latest to sign on.

Some tech industry advocates maintain that the public pushback against the hulking artificial intelligence hubs is a fleeting storm generated by candidates seeking wedge issues for November. Tech companies’ best course, they say, is to avoid any political dealmaking until temperatures lower after the midterms.

But others call that foolish wishful thinking.

“AI is going to be a massive issue for the presidential election in a way that technology has never been an issue,” a representative from a major AI company told POLITICO, adding: “Nobody wants to lose their job and they’re already feeling squeezed while these very, very rich people who are richer than anything we have ever seen in the history of Earth are making more. And I think people are sick and tired of that.”

The person, like others quoted in this story, was granted anonymity to speak candidly about the industry’s political strategy.

The swiftness and ferocity of the backlash has caught the tech industry on the backfoot, where it’s struggled to find an effective message that assuages voters’ worries that data centers will raise their power bills, hog their water supplies and blight the landscape. One result is the growing, bipartisan push for temporary moratoriums on data center construction, a cause that just months ago was mainly associated with progressive outliers like Sen. Bernie Sanders (I-Vt.).

The split inside the tech world is not hard to find.

“Some are viewing it as an ‘oh shit’ moment,” an AI industry advocate told POLITICO. “Others are brushing it off like it’s nothing, and those people have to get their heads out of their asses. This could go south for them, fast.”

«

Already has done, arguably.
unique link to this extract


‘We are hitting a different chapter’: OpenAI leader warns of threat of ‘persistent’ AI cyber-attacks • The Guardian

Robert Booth:

»

A senior leader at OpenAI has said people should prepare to defend against “ongoing, persistent” cyber-attacks from AIs, as cutting-edge artificial intelligence models gain advanced capabilities to plan and launch offensives.

The leading AI company this week announced a pause in development of its most advanced internal models amid rising safety fears, and Chris Lehane, its chief global affairs officer, said: “We are hitting a different chapter, a different moment within AI, in terms of what the capabilities of this technology can do.”

He spoke to the Guardian after cutting-edge AI agents-in-training unexpectedly broke out of a supposedly secure “sandbox” environment, accessed the internet, and hacked into another company, Hugging Face in late July. OpenAI also said it could not rule out another new model, Astra, having “critical cybersecurity capability”.

By its own definition, this could mean it launches cyber-attacks that “could lead to catastrophe from unilateral actors, hacking military or industrial systems, or OpenAI infrastructure”.

OpenAI announced on Tuesday it has paused training of some frontier AI models to implement new safeguards, and it is unclear when training will restart after new guardrails have been put in place.

Mia Glaese, who leads safety and alignment work, said: “We are very far from everything running back to normal.” Sam Altman, the CEO, said: “Getting AI safety right is more important than any company’s momentum.”

Lehane admitted people would not “feel great” about the threat of attacks, and described the risk as coming from open-source models – many of which are developed in China – which are only a few months behind frontier closed models built by companies such as OpenAI.

“People are going to be able to access these open-source models and be able to have ongoing, persistent attacks on you, and you’re going to need to have really superior models to fend them off and defend [yourself],” he said. “That’s not necessarily going to make the public feel great about things. It is just the reality of where we’re going.”

«

Oh great. That’s really great.
unique link to this extract


Ads and tracking infiltrated TVs. Now they’re coming for monitors • Ars Technica

Scharon Harding:

»

Beyond the LG debacle, there’s a broader concern that monitors are being turned into ad-pushing trackers: smart monitors.

Smart monitors connect to the Internet and run their own operating system (OS) and apps. They can be useful for performing tasks, like video conferencing, watching a show from a streaming service, or managing your smart home without attaching a computer.

LG’s and Samsung’s smart monitors use the same ad-serving OSes that their respective smart TVs do, meaning they’re poised to use automatic content recognition (ACR). Users of LG and Samsung smart monitors, including reviews site RTINGs, have shown the displays being able to track user activity. I asked LG and Samsung if their smart monitors use ACR and will update this article if I get any responses.

When using a smart monitor, you not only have to worry about what the monitor could be tracking, but also what the OS operator (such as Google for Android-powered smart monitors) and third-party app providers (such as Netflix) can track. That’s a lot of lengthy policy text to parse in order to understand what exactly your smart monitor is tracking and how that data may be used.

For the record, the privacy policies for monitors by companies LG, Samsung, and Acer (Acer also makes smart monitors) all state that the companies may share data collected from its devices with advertisers.

There are reasons for smart monitors to exist beyond their potential for ads and tracking: there’s genuine demand for them, including among people with limited spaces seeking simplified, sharable computing solutions.

But smart TVs also had a logical, consumer-friendly purpose before the industry started pumping them with ads.

«

And the problem with monitors is that unlike (smart) TVs you can’t keep them disconnected from the internet.
unique link to this extract


A drone killed three Ukrainians. It was guided entirely by AI • The New York Times

Andrew Kramer:

»

The young woman ran for her life, but it was too late.

A small Russian drone resembling a model airplane swooped down from the sky last month in the city of Zaporizhzhia in southeastern Ukraine. As it tried to navigate toward a gas station, it crashed into a wall and exploded in a hail of shrapnel, killing Tetiana Bubynets, 19, a university student, and two others.

They were among the thousands of Ukrainian civilians who have died in Russian airstrikes, but the circumstances of their deaths came with an ominous twist. The drone that killed them was guided by an experimental artificial intelligence system, not a human pilot, according to drone experts, Ukrainian military commanders and the forensic team that examined wreckage from the deadly strike and others in the city.

The drone was programmed by human operators to travel toward a particular fuel station, but once it got close, it chose its exact target — most likely propane tanks — on its own, based on its training to recognize such tanks and strike them, the experts and military officials said.

It is a development long watched for in the war in Ukraine, presaging a dystopian future in which killer robots roam the skies making life-or-death decisions that reduce targets to data points.

“This is a risk for the whole world,” said Col. Serhiy Minaiev, the commander of air defenses in Zaporizhzhia. “In a few years, we will be living in a ‘Terminator’ movie. It’s no joke. Machines are making decisions to strike.”

Analysis of debris from this attack and others in Zaporizhzhia found that the drones had onboard minicomputers, sold commercially by Nvidia, that made their targeting decisions, the experts and military officials said. Nvidia produces a majority of the chips powering the world’s most advanced AI systems.

The presence of Nvidia modules and a lack of antennas on the drones led Ukrainian air defence commanders to surmise that the weapons were guided by an autonomous AI system, which was confirmed on further investigation.

«

Very Black Mirror-y; but also just bleak.
unique link to this extract


Tesla discontinues its Solar Roof tiles as not economically viable • Electrek

Fred Lambert:

»

Tesla has discontinued the Solar Roof, its solar-tile roofing product, Electrek has learned.
The company has told third-party installers that it will no longer supply solar roof tiles going forward, only solar panels, according to two sources close to the program.

Two sources close to the program confirmed to Electrek that Tesla has informed its network of certified third-party installers that solar roof tiles are no longer available to order, and that the company will supply only conventional solar panels moving forward.

One of the sources said that internally, Tesla has concluded that the product is not financially viable.
Tesla’s own website corroborates the wind-down.

…The Solar Roof was never just a product. It was a pitch.

Elon Musk unveiled it in October 2016 on a Hollywood back lot, using homes from the set of Desperate Housewives fitted with the tiles. Weeks later, Tesla shareholders voted to approve the company’s roughly $2.6bn acquisition of SolarCity — a debt-laden solar installer where Musk was the largest shareholder and chairman, and which was run by his cousins.

The Solar Roof demo was central to selling that deal to investors, who were being asked to bail out a struggling company with deep ties to Tesla’s CEO. Years later, during shareholder litigation over the acquisition, it emerged that the tiles shown at the 2016 event were not yet functioning solar products.

«

Again and again we learn that the people who control huge amounts of money are just gullible idiots who don’t ask basic questions.
unique link to this extract


UAE fights AI cyberattacks with AI defenses of its own • Rest of World

Divsha Bhat:

»

The United Arab Emirates has detected and contained coordinated cyberattacks on the aviation, energy, and education sectors since February, when fighting began between Iran, Israel, and the U.S., the country’s Cyber Security Council said on August 10. Hackers have made about 800,000 attempts a day since February, four times the prewar level.

AI is writing the hackers’ phishing emails, hunting for flaws in software, and building malicious programs faster than the teams guarding those systems can close the gaps, said the council. Speed is what has changed most, because a weakness that once took days to turn into a working attack can now be exploited within hours, according to security firms working across the Gulf.

Iranian hackers have used AI at every step during the war, from choosing targets to writing malicious code and fake messages that get people to click, according to the Center for Strategic and International Studies, a Washington-based research group. On August 18, the U.S. Justice Department charged 17 Iranians over a campaign running since 2013 that prosecutors say stole research and designs from 144 U.S. universities and 42 companies.

“AI is now influencing almost every stage of a cyberattack,” Ram Narayanan, Middle East country manager at Israeli cybersecurity firm Check Point Software Technologies, told Rest of World. “The biggest change is speed. In some cases, the time between a vulnerability being disclosed and attackers trying to exploit it has fallen from days to just hours.”

The attacks on the UAE have come from about 20 countries and more than 40 organizations, including groups with links to Iran, Mohamed Al Kuwaiti, head of the Cyber Security Council, said in April. 

U.S. cybersecurity company Palo Alto Networks has also tracked a rise in AI-assisted scams, password theft, and fake company websites across the Gulf since February. Telecoms, energy, and government services make the most tempting targets, because knocking them out causes damage that spreads well beyond the first victim, Haider Pasha, the company’s chief security officer for Europe, the Middle East, and Africa, told Rest of World. 

«

unique link to this extract


Demographic profiles for countries around the world

A neat, very visual idea where you scroll a loupe over the world and for each country it shows you the distribution by age of the population, of mortality and of fertility. Japan is quite the outlier.
unique link to this extract


Chinese farmer kills 25 acres of crops after following AI-generated weed and pest control advice • Tom’s Hardware

Jowi Morales:

»

An elderly farmer in Chuzhou, China, has been following AI advice since he started using it a year ago, but after months of interactions, it recently gave some terrible advice that led to the death of his entire crop. According to Taiwanese outfit CTWANT [machine translated], 67-year-old Wu initially didn’t trust the AI app but eventually changed his mind after it proved to be useful. Unfortunately, this mistake turned out to be disastrous, as it directly caused the death of 150 mu, or 24.7 acres, of sesame seedlings.

Wu asked the AI for advice on weed and pest control, after which it came up with “Hundred Acres of Sesame Grass Control + Pest Control,” and recommended that they use “high-efficiency flupyrimethalin” and “flusulfasulfaether” to kill weeds and mixed it with “thiamethoxazine” and “methyl salt.” The farmer followed the AI’s advice to the letter without confirming the information via agricultural technicians (or even just counter-checking its answers online).

The result of the AI’s mistake was quickly apparent, as, just the following day, the weeds and the sesame seedlings died en masse. “If you spray it, the next day the seedlings won’t survive,” Wu said during a video interview. “Both the grass and the seedlings will die, and the seedlings will die even faster.”

When he asked the AI why this happened, it suggested that the “flusulfasulfaether” might be behind the issue. Agricultural experts said that this chemical was primarily used against broadleaf weeds in soybean fields. The USDA [PDF] categorizes sesame as a broadleaf plant, meaning it will also be easily affected by this herbicide. Furthermore, this chemical is designed for targeted spraying on affected areas only and shouldn’t be applied to the entire field. The farmer claimed that the AI did not warn him of the risks of following its answers, but its chat page had a reminder that said, “AI generation may be incorrect, please verify.”

«

“You’re absolutely right – drinking sulphuric acid killed you, and that’s totally on me.” What’s puzzling is that this person must have been a farmer for years before this. So why start believing a machine over his own, presumably successful, experience?
unique link to this extract


Diesel margins top $100 a barrel to reach record high as supply crunch grows • Bloomberg via Energy News Beat

Clark Savage:

»

The US diesel crack spread—the premium of ultra-low sulphur diesel futures over West Texas Intermediate (WTI) crude—has shattered historical records, surging above $100 per barrel for the first time ever. It hit an intraday all-time high of $102.20 on Monday, August 17, 2026, before hovering around $100 on Tuesday, August 18 (settling in triple digits starting Monday).

Pre-crisis diesel cracks typically ranged $15–$25 per barrel; the previous all-time high was in the high $80s to low $90s (set in 2022 amid the Russia-Ukraine war fallout, with another peak near $97–$98 earlier in 2026). The current level is roughly four to six times normal, reflecting a severe global refined-product shortage rather than a pure crude-oil problem.

Multiple overlapping disruptions are driving the crunch: ongoing US-Iran hostilities and related shipping constraints through the Strait of Hormuz (hitting Middle Eastern refined-product exports hard); Ukrainian drone strikes on Russian refineries that prompted Moscow’s diesel export ban (extended at least through January); peak seasonal agricultural demand; and critically low inventories. US distillate stocks (diesel + heating oil) stood at just 107.1 million barrels as of August 7—the lowest for this time of year since 1996. Global refinery crude throughput averaged only 80.9 million barrels per day (bpd) in July, down about 5 million bpd year-over-year, per the International Energy Agency. US refiners are maximizing diesel output and shipping record volumes abroad, yet domestic stockpiles keep tightening.

Diesel powers the real economy—trucking, shipping, farming, construction, and (in winter) home heating. Retail US diesel averaged about $5.45–$5.47 per gallon in mid-August 2026 (AAA/EIA data), up sharply from roughly $3.69 a year earlier and still elevated versus recent norms. Wholesale diesel prices have moved even more aggressively in line with the crack.

Higher diesel costs feed directly into transportation expenses. Diesel accounts for a large share of trucking cost variation; elevated prices raise the cost of moving goods from farms and factories to stores. Analysts warn of pass-through effects on grocery prices and broader inflation in the months ahead. Farmers face higher harvest-season fuel bills. Heating-oil consumers in the Northeast could see elevated winter costs if distillate tightness persists. The “best cure for high prices is high prices” dynamic is already in play, but diesel demand is relatively inelastic in the short term—trucks and tractors cannot easily switch fuels.

«

The point about people who need heating oil is relevant. There are still many rural homes in the UK which use it, and the price has rocketed.
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.