Does it stand to reason that an Oura ring can track all sorts of details about your sleep? A new lawsuit claims not. CC-licensed photo by Timo on Flickr.
You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.
A selection of 10 links for you. Wide awake. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.
Iranian hackers shut down UK power plant • The Telegraph
Tony Diver, Rozina Sabur and Matt Oliver:
»
Iran shut down a British power plant for four days in an unprecedented cyber attack, The Telegraph can disclose.
It is thought to be the first time that hackers affiliated to the Iranian regime have succeeded in closing down such a facility in the UK, and is believed to be the most successful cyber attack of its kind.
The Telegraph understands that the incident took place at the same time as a series of attacks on US water infrastructure last month, which affected 12 states and caused concern in the White House.
British officials refused to be drawn on which power plant had been affected, citing security concerns, but it is understood that the facility was disabled for four days while staff fought to bring it back online.
It is understood that the plant affected was relatively small and that the outage had no impact on the UK’s wider power supply or energy generation.
«
“Relatively small” is an odd phrase to use about a power plant, but everyone is insisting that this didn’t create any risk to the wider grid. Even so: it’s a statement of intent and ability by Iran, which has long been recognised as a cyber adversary worth taking seriously.
unique link to this extract
Oura faces lawsuit accusing it of misleading consumers about sleep-tracking accuracy • TechCrunch
Aisha Malik:
»
A proposed class action lawsuit is accusing smart ring maker Oura of deceiving consumers about the accuracy of its sleep tracking features.
The lawsuit, filed on Thursday by Clarkson Law Firm in San Francisco, alleges that Oura rings are unable to measure any of the physiological signals needed to assess sleep quality or determine sleep stages, and that they instead rely on AI-generated estimates that have “a coin flip’s chance of being correct,” according to the complaint.
The lawsuit follows years of Oura users taking to the internet to share their frustration with the smart ring’s sleep accuracy, with some saying they felt they slept poorly but were told their sleep had been optimal. Some have also expressed doubts about the device’s ability to monitor and detect sleep cycles.
“To capitalize on consumers’ desire for a tracker that could actually track sleep and monitor their sleep cycles, Oura sold expensive tracking rings, priced at $300 and up, advertising exactly that: that Oura rings are capable of seeing what only a hospital sleep lab can see, including the four different stages of sleep,” the complaint reads. “Oura did not just claim to measure a heartbeat or a temperature, but the exact stage of sleep the wearer is in—which in reality requires electrodes in the scalp and sensors on the eyes, as only a hospital or other clinical setting can do.”
Despite this, the complaint alleges that Oura marketed its ring as being “built for accuracy” and offering “unparalleled accuracy.” It says Oura told customers its smart rings could achieve 79% accuracy in their measurements, and more recently claimed that the rings offered 95% sleep-staging accuracy compared with clinical sleep labs.
«
Sleep scores are daft anyway. (One person recently commented on X “I put years of my sleep data into Claude. After working for hours it concluded that when I went to bed late, I got less sleep.”)
unique link to this extract
Would even an AI disaster on the scale of Hiroshima be enough to make humankind protect itself? I fear not • The Guardian
Timothy Garton Ash:
»
Competition has been a vital engine of human evolution, yet now two of the fiercest kinds of contemporary competition threaten to prevent us from taking the collective action required to safeguard our species. That’s the commercial competition (for profit) between the corporations developing these agents, and the geopolitical competition (for power) between the US and China.
On the first, compare and contrast the development of nuclear weapons, which was tightly controlled by a handful of states – first the US, with its Manhattan Project, then the Soviet Union, followed by Britain, France and China. Even then, we came close to the brink several times – notably in the Cuban missile crisis of 1962. And even with the compelling logic of “mutually assured destruction”, it took a quarter-century to get from the horrors of Hiroshima and Nagasaki to the nuclear non-proliferation treaty that came into force only in 1970. India, Pakistan, Israel and North Korea have subsequently ignored that treaty, but still no nuclear weapons have been deliberately used in anger for more than 80 years. The Hiroshima taboo has (just) held.
Now, with AI, it’s as if there were 10 separate Manhattan Projects run by fiercely competing corporations. Unlike with nuclear weapons, there’s also no obvious logic of “mutually assured destruction” to tame the fierce geopolitical competition between the US and China. And whatever shape the first major AI-related disaster takes, it won’t impact all countries, companies and people equally.
It therefore seems unlikely that just one disaster will be enough to bring us humans to our senses, acting together to control the superhuman power we are creating. I really hate to say this, but … an AI Hiroshima? We should be so lucky.
«
The concern is that unlike the Cuban missile crisis, even the smartest people might not know that the disaster was happening – as with the agents which swarmed to hack into a solutions database.
unique link to this extract
AI research frontiers: mathematics, physics & human biology
»
A source-based record of research advances made with decisive help from AI—separating established results from fast-moving claims still under review.
An established entry needs a strong public source and meaningful verification. Promotion from “under review” depends on the quality of checking—not time passed or headline size.
«
This offers three levels of importance: 1 – Narrow specialist result; 3 – Important within its field; 5 – Historic major advance. Though presumably with something in the middle.
One hopes that chemistry and biochemistry will join them soon.
unique link to this extract
Watching Roku’s AI channel is like eating from a trough • The Verge
Charles Pulliam-Moore:
»
Roku [ghas] added four new channels to its library of streamable programming. Along with dedicated feeds for old episodes of Mad TV, Whose Line Is It Anyway?, and a variety of Black sitcoms, the platform also debuted a 24/7 stream filled with projects from Colin Petrie-Norris’ AI startup, Fairground. Partnering with Roku is a significant step up for Fairground, which launched in 2025 and is reportedly working on its own AI-generated adaptations of Robin Hood and Dracula. But watching the Fairground channel, you get the distinct sense that Roku hasn’t really gotten all that much out of the deal aside from the ability to say that it’s getting in on the gen AI craze.
Like all FAST [free ad-supported streaming television] channels, Fairground AI Creator TV is filled with a curated selection of programming that you’re meant to watch somewhat passively. While you can scrub forward and backward a few seconds, there’s no way to choose exactly what kinds of videos the channel serves up. That might add a fun sense of discovery to the channel if it featured series and films that were actually riveting. But everything that I saw while watching for about an hour was more or less the same kind of short-form slop that text-to-video models have become known for.
Aside from being produced with AI, there’s no consistent theme or genre to any of Fairground’s programming — all of which is made by a selection of content creators who have partnered with the company. Some videos are clearly generated by models trained on footage ripped from professionally produced anime, while others just look like terrible “lifelike” CGI. Surprisingly, there were semi-cohesive narratives to some of what I watched. I could follow what was going on in a “live-action” horror about people destroying a demon by making fun of it, and I got the sense that a fantasy about Greek heroes might have been inspired by The Odyssey.
…the overall shoddiness of Fairground’s programming becomes especially apparent whenever the stream cuts to an ad — this is a FAST channel — for one of the traditionally produced movies and shows Roku offers.
«
People are going to consume this, though, or have it on in the background. And that’s just fine by Roku. Cost next to nothing and carry adverts. Basically a slow money generator.
unique link to this extract
Astronomers discover a brand-new type of astrophysical object: A black hole star • MIT News
Jennifer Chu:
»
Astronomers at MIT and elsewhere have spotted an extremely bright red spot in the early universe. The object resembles an enormous star, spanning the size of our solar system. But it also is putting out 100 billion times more energy than any known star can physically produce. In fact, such energies are closer to what a black hole might generate.
The curious combination suggests that the red spot is an entirely new type of astrophysical source. The astronomers are calling it a “black hole star.”
In a paper appearing today in the journal Nature, the team presents their analysis of the new object, which they discovered using NASA’s James Webb Space Telescope (JWST). The telescope spotted the bright red dot in the very early universe, just a few hundred million years after the Big Bang.
The scientists conclude that the most likely explanation for the strange red dot is that it is a mashup of a black hole and a star — a combination that has never been observed until now. The object is likely a hugely dense cloud of gas, powered not by standard nuclear fusion, but by a central black hole.
…If the bright red dot is indeed a black hole star, it would help to solve the identity of other mysterious “little red dots” that have appeared in nearly every deep space image JWST has taken to date.
“These little red dots seem to be everywhere in the early universe but essentially disappear by the present day,” [lead author Rohan] Naidu says. “What exactly these objects are has been one of the most debated topics of the JWST era.”
«
More black holes around than we thought, seems like.
unique link to this extract
Amazon hiked prices on Echo, Fire TV, Kindle, eero to offset ‘significant increases’ in memory costs | Fortune
Sebastian Herrera:
»
Amazon has raised prices for several of its first-party devices, including its Echo smart speakers and Fire TV line, making it the latest big tech company to increase consumer costs as memory chip shortages pressure several industries.
The company overnight raised prices across its hardware product line, according to a review by Fortune. This includes raising the cost of its base Echo Dot from $49.99 to $79.99 and its Echo Show 11 from $219.99 to $249.99. It also raised its 16GB Kindle from $109.99 to $149.99 and its 16GB Kindle Paperwhite from $159.99 to $199.99.
Its Fire TV Stick HD model increased from $34.99 to $39.99, and its Fire TV Stick 4K Max went from $59.99 to $84.99. The Amazon eero 7 wireless mesh networking system increased from $349.99 to $399.99, and the eero Pro 7 rose from $699.99 to $799.99. Notably, Amazon did not increase prices for its Ring products. The website Pocket-lint earlier reported on the Fire TV increases.
An Amazon spokeswoman, who confirmed the price increases, said in a statement that the consumer electronics industry is “facing significant increases in memory and storage component costs. After absorbing these increases for as long as we could, we recently adjusted pricing across our product lines.”
«
Nobody’s safe from RAMageddon.
unique link to this extract
Tiny-Net 2D Embedding Visualizer
Robert Graham:
»
I’ve been obsessing over trivial technical details of LLMs, spending way too much time grokking trivial issues.
One of them is how ‘tokens’ are extracted from text and sent through the neural-nets. It’s done by creating a random vector for each token, such as converting the letter ‘x’ into coordinates [1.37, 2.98]. During “training”, these “parameters” are “adjusted”. What that means is that when plotted, they’ll move.
So I created a webpage that animates this, watching what happens to this randomized parameters during training.
We see:
* vowels and punctuation create their own clusters away from others
* The letter ‘v’ moves away by itself
* The ‘-‘ character stays at the center, moving toward the unpopulated quadrant.«
This gives a glimpse into how LLMs categorise content (the bigger ones work with words, not characters, and use multi-dimensional matrices to locate those words). There’s more explanation on a linked page.
unique link to this extract
Exclusive: China is building AI models of American voters. Why? • Natalie Winters
Natalie Winters:
»
China is building increasingly detailed models of the American electorate—who voters are, what they believe, why they support Donald Trump or Joe Biden, and how they might respond to political messages.
A review of Chinese-language research, university projects, datasets, and government-linked think-tank publications reveals that Chinese institutions have moved far beyond simply following American polls. Researchers are now constructing artificial versions of US voters, assigning political and demographic labels to more than one million X users, simulating state-by-state presidential elections, and conducting experiments on synthetic voters in Pennsylvania.
Another Chinese study, funded by the country’s National Social Science Fund, used American polling data to infer “America First” attitudes toward tariffs, alliances, international organizations, and democracy promotion. Its author explicitly argued that the research could help China evaluate the future direction of US foreign policy and formulate a response.
Chinese universities and government think tanks are developing increasingly granular, interactive models of the American electorate that can be used to forecast political behavior, test messages, interview simulated constituencies, and identify the domestic forces shaping US policy toward China.
Their own documents raise a simple question: Why?
The most expansive project uncovered in this investigation began with an enormous collection of posts from X, formerly Twitter.
In a paper titled “ElectionSim: Massive Population Election Simulation Powered by Large Language Model Driven Agents,” a Fudan University-led research team said it collected 171,210,066 posts from 9,596,198 X users during the 2020 election period.
«
Not sure that figuring out if they back Trump or Biden is useful, nor slurping up a million accounts on X. And good luck trying to work out Trump’s policy towards China, which still has another two years to run. A fun project to do, one supposes.
unique link to this extract
US citizen who deleted phone’s data says his prosecution puts privacy at risk • The New York Times
Adeel Hassan continues the story of Samuel Tunick, whose phone runs Graphene OS which wiped itself when US border staff tried to access it with the wrong code:
»
A spokeswoman for US Customs and Border Protection said in a statement that the agency had the authority to search the electronic devices of anyone entering or leaving the United States, regardless of citizenship, to enforce laws addressing terrorism, child exploitation, drug- and human-smuggling, visa fraud and national security threats.
“The border search will only include an examination of information that is present on the device at the time it is presented for inspection,” the spokeswoman said, adding that it searched the electronic devices of fewer than 0.01% of all arriving international travelers in the last fiscal year.
In an interview, which has been condensed and edited for clarity, Mr. Tunick explained how he ended up in this situation.
NYT: Why did you decide to replace your phone’s default software with Graphene OS, which can delete the phone’s contents?
Knowing the kind of surveillance that activists have been subjected to, it just seems like a good idea to protect my private data that I don’t necessarily want corporations or governments to have access to, just out of the principle of privacy.
NYT: Some people think that if you have that software, you’re trying to hide something.
If someone you don’t know, who’s actively hostile to you, is trying to access your private data, your pictures, your messages or notes to self, that may be not something that you’d like.
NYT: Your lawyers have said that the government was investigating you over your association with the movement against “Cop City,” a planned police and fire training center in Atlanta. What are your views?
I was a participant, along with hundreds of thousands of other people in the movement. I heard that people were trying to fight it, and I have a background in environmental studies. That’s what I got my bachelor’s degree in. I’m about to go back to school for geography, so the environmental justice issue struck a chord for me, but especially so when it was combined with this concerning trend that we’ve seen over the last several years of expanding and militarizing the police at the exact time when tens of millions of Americans are demanding the exact opposite.
«
| • 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
The Overspill: when there's more that I want to say 