Audiophiles are discovering cheap Raspberry Pi devices inside hugely expensive streaming devices. Are they being ripped off? CC-licensed photo by osde8info 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. All ears. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.
“IG is a drug”: internal messages may doom Meta at social media addiction trial • Ars Technica
Ashley Belanger:
»
Anxiety, depression, eating disorders, and death. These can be the consequences for vulnerable kids who get addicted to social media, according to more than 1,000 personal injury lawsuits that seek to punish Meta and other platforms for allegedly prioritizing profits while downplaying child safety risks for years.
Social media companies have faced scrutiny before, with congressional hearings forcing CEOs to apologize, but until now, they’ve never had to convince a jury that they aren’t liable for harming kids.
This week, the first high-profile lawsuit—considered a “bellwether” case that could set meaningful precedent in the hundreds of other complaints—goes to trial. That lawsuit documents the case of a 19-year-old, K.G.M, who hopes the jury will agree that Meta and YouTube caused psychological harm by designing features like infinite scroll and autoplay to push her down a path that she alleged triggered depression, anxiety, self-harm, and suicidality.
TikTok and Snapchat were also targeted by the lawsuit, but both have settled. The Snapchat settlement came last week, while TikTok settled on Tuesday just hours before the trial started, Bloomberg reported.
For now, YouTube and Meta remain in the fight. K.G.M. allegedly started watching YouTube when she was six years old and joined Instagram by age 11. She’s fighting to claim untold damages—including potentially punitive damages—to help her family recoup losses from her pain and suffering and to punish social media companies and deter them from promoting harmful features to kids. She also wants the court to require prominent safety warnings on platforms to help parents be aware of the risks.
To avoid that, platforms have alleged that other factors caused K.G.M.’s psychological harm—like school bullies and family troubles—while insisting that Section 230 and the First Amendment protect platforms from being blamed for any harmful content targeted to K.G.M.
They also argued that K.G.M.’s mom never read the terms of service and, therefore, supposedly would not have benefited from posted warnings. And ByteDance, before settling, seemingly tried to pass the buck by claiming that K.G.M. “already suffered mental health harms before she began using TikTok.”
But the judge, Carolyn B. Kuhl, wrote in a ruling denying all platforms’ motions for summary judgment that K.G.M. showed enough evidence that her claims don’t stem from content to go to trial.
«
No clue on when this might come to trial, if ever. Nor is it clear whether TikTok and Snapchat settling will have any impact. The suspicion is that the other companies will settle too, which means that nothing comes to trial, which means that nothing becomes precedent or case law.
unique link to this extract
Moltbot, the AI agent that “actually does things”, is tech’s new obsession • The Verge
Emma Roth:
»
Federico Viticci at MacStories highlighted how he installed Moltbot on his M4 Mac Mini and transformed it into a tool that delivers daily audio recaps based on his activity in his calendar, Notion, and Todoist apps. Another person prompted Moltbot to give itself an animated face, and said it added a sleep animation without prompting.
Moltbot routes your request through the AI provider of your choice, such as OpenAI, Anthropic, or Google. Like many of the AI agents we’ve seen so far, Moltbot can fill out forms inside your browser, send emails for you, and manage your calendar — but it does so a lot more efficiently, at least according to some of the people using the tool.
There are some caveats, though; you can also give Moltbot permission to access your entire computer system, allowing it to read and write files, run shell commands, and execute scripts. Combining admin-level access to your device and your app credentials could pose major security risks if you’re not careful.
“If your autonomous AI Agent (like MoltBot) has admin access to your computer and I can interact with it by DMing you on social media, well now I can attempt to hijack your computer in a simple direct message,” Rachel Tobac, the CEO of SocialProof Security, says in an email to The Verge. “When we grant admin access to autonomous AI agents, they can be hijacked through prompt injection, a well-documented and not yet solved vulnerability.” A prompt injection attack occurs when a bad actor manipulates AI using malicious prompts, which they can either pose to a chatbot directly or embed inside a file, email, or webpage fed to a large language model.
«
People are doing deeply foolish things with these AI agents. Meredith Whittaker, CEO of the messaging app Signal, pointed to a post where someone claims to have created a “skill” for Claude which would in fact backdoor your computer. It’s like Windows XP and the virus explosion all over again.
unique link to this extract
Audiophiles keep finding a $40 computer board inside hi-fi streamers selling for thousands • Headphonesty
Alexandra Plesa:
»
If you browse audio forums, you might be familiar with this pattern. Someone opens a premium streamer and discovers that a Raspberry Pi sits inside. Occasionally, posters might frame these discoveries as proof that network streamers are unnecessarily expensive.
Bryston offers the clearest examples of this design approach. The BDP-π, which sold for $1,295 before being discontinued, used a Raspberry Pi with a HiFiBerry Digi+ HAT (a hardware accessory board). It delivered bit-perfect S/PDIF output and support up to 24-bit/192 kHz.
At the top of the lineup, Bryston’s $6,795 BR-20 includes a Raspberry Pi 4–based streaming module. That unit is part of a full DAC and preamp supporting high-resolution PCM and DSD playback.
Other manufacturers take a similar route:
• Orchard Audio’s $550 PecanPi Streamer runs on a Raspberry Pi 3B using Volumio
• Pi2Design’s $400-$450 Mercury V3 DAC is a Raspberry Pi 4–based streamer/DAC with AES, S/PDIF, and other digital outputs, plus an onboard DAC for analog output.Raspberry Pi prices start at $35 for home use, so the difference might suggest that manufacturers are hiding something. After all, the same board used in DIY projects shows up in products that cost many times more.
Still, Bryston did not attempt to disguise the component, even going as far as naming a product the BDP-π. That openness shifts the discussion away from deception and leads to more questions. What does the Raspberry Pi do in a streamer, and what justifies the rest of the cost?
«
The answer turns out to be pretty simple:
»
In a modern system, the Raspberry Pi acts as a networked transport, so it’s not responsible for shaping the sound.
«
Expensive gas still biggest driver of high UK electricity bills, says UKERC • Carbon Brief
Simon Evans:
»
High gas prices are responsible for two-thirds of the rise in household electricity bills since before the global energy crisis, says the UK Energy Research Centre (UKERC).
The new analysis, from one of the UK’s foremost research bodies on energy, flatly contradicts widespread media and political narratives that misleadingly seek to blame climate policies for high bills.
Kaylen Camacho McCluskey, research assistant at UKERC, tells Carbon Brief that despite “misleading claims” about policy costs, gas prices are the main driver of high bills. She says: “While the story of what has driven up GB consumer electricity bills is often largely attributed to policy costs, our analysis shows that this is not the case. Volatile, gas-linked market prices – not green policies, as some misleading claims have suggested – dominate the real-terms increase in bills since 2021.”
In its 2025 review of UK energy policy, published today, UKERC says that annual electricity bills for typical households have risen by £166 since 2021.
It says that, after adjusting for inflation, some two-thirds of this increase (£112) is due to higher wholesale gas prices…UKERC estimates that, despite only supplying a third of the country’s electricity, gas-fired generators set the wholesale price of power around 90% of the time in 2025.
«
It’s hard though to see how this can be changed. If gas provides the marginal filler, and you can’t find a way to remove it from supply, then its cost will always determine electricity prices. The UKERC has some suggestions in the article, but they seem handwavy.
unique link to this extract
Global population living with extreme heat to double by 2050 • University of Oxford
»
A new University of Oxford study finds that almost half of the global population (3.79 billion) will be living with extreme heat by 2050 if the world reaches 2.0°C of global warming above pre-industrial levels – a scenario that climate scientists see as increasingly likely.
Most of the impacts will be felt early on as the world passes the 1.5°C target set by the Paris Agreement, the authors warn. In 2010, 23% of the world’s population lived with extreme heat, and this is set to grow to 41% over the next decades.
Published in Nature Sustainability, the findings have grave implications for humanity. The Central African Republic, Nigeria, South Sudan, Laos, and Brazil are predicted to see the most significant increases in dangerously hot temperatures, while the largest affected populations will be in India, Nigeria, Indonesia, Bangladesh, Pakistan, and the Philippines.
Countries with colder climates will see a much larger relative change in uncomfortably hot days, more than doubling in some cases.
Compared with the 2006–2016 period, when the global mean temperature increase reached 1°C over pre-industrial levels, the study finds that warming to 2°C would lead to a doubling in Austria and Canada, 150% in the UK, Sweden, Finland, 200% in Norway, and a 230% increase in Ireland.
«
To consider: 2050 isn’t actually that far away. It’s 24 years, and I’d expect the majority of readers here still to be alive then.
unique link to this extract
Apple and its suppliers are counting the cost of iPhone Air failure • Culpium
Tim Culpan is a former Bloomberg technology reporter in Taiwan with many excellent contacts among its suppliers:
»
Apple and its suppliers are now stuck with components for up to 1.5 million units of iPhone Air, my sources tell me, even after the order came down in October to cut back production. What’s worse, some of that cannot be repurposed and instead may need to be scrapped, I am told. To be clear, that doesn’t mean 1.5 million iPhones will be scrapped, merely some of the components specific to the iPhone Air.
Apple will release earnings this week, but it’s unlikely Tim Cook and his team will discuss the issue. Apple’s CEO dodged the question of Air’s poor sales last quarter, and instead chose to talk up the iPhone 17 Pro and iPhone 17 — a tacit admission that Air was a dud.
My own analysis, based on what cannot be re-used, puts the write-off into the low hundreds of millions of dollars. Frankly, that’s not a big deal: iPhone sales were $49bn in the three months to Sept. 27, and are expected to be around $80bn for the December quarter.
The bigger issue is how Apple failed to understand its own customers, and how this mess will ripple through the supply chain. I am told that while some vendors will be stuck with a bill, Apple itself will soak up most of the cost.
…The OLED “Super Retina XDR” screen is basically the same across all models, but the Air’s 6.5-inch size is mid-way between the 6.9-inch and 6.3-inch versions. I am told that displays which have already been cut, framed and put onto modules will need to be scrapped, though some of that will also be crushed, separated, and recycled.
Possibly the biggest hurt could be with the chips. Apple uses the same A19 Pro CPU in the Air as it does with the iPhone 17 Pro. But the Air has only 5 GPU cores — as does the base iPhone 17 — while the iPhone 17 Pro has 6 GPU cores. (To be blunt, this is merely chip binning, not a new chip).
As a result, the unused Air chips cannot be put in the the lower-end base iPhone 17 nor in the higher-end iPhone 17 Pro. They cannot be repurposed. Even worse, the Air has 12GB of DRAM while the baseline iPhone 17 has just 8GB, according to TrendForce. So, any processor modules which have already had their DRAM fused onto the CPU would also result in wasted DRAM — unless Apple and TSMC find some magical way to “unfuse” the memory from the base die.
«
Those millions of modules of DRAM that have gone to waste will really stick in the craw of Cook, who of course abhors waste. Speaking of whom…
unique link to this extract
Aside from that, Mr. Cook, what did you think of the movie? • Spyglass
MG Siegler:
»
Tim Cook is captured. There is simply no other explanation for his actions over the past year or so. But it perhaps culminated this weekend when Cook went to a special private showing of the documentary Melania [I think that should be “documentary” – Overspill Ed] at the White House. Yes, that Melania. That in and of itself would have probably been fine. I mean, it’s potentially problematic for a host of reasons that I’ll get to, but such is our world right now. Then one shot – a gunshot – turned attending that movie screening into a statement…
While Cook was enjoying his popcorn and champagne with the likes of Mike Tyson, Tony Robbins, and other “VIPs”, it was complete and utter chaos on the streets of Minnesota. Just hours earlier, Alex Pretti, a 37-year-old ICU nurse, was shot and killed by ICE agents. Maybe, just maybe, postpone the movie premiere?
Of course, President Trump was never going to do that because the official White House stance is that Pretti was a “domestic terrorist” and the agents were acting in self defense. And never mind that this was the second such murder in the past 17 days, the show must go on!
But it didn’t have to for Cook. He could have, and should have, backed out of the event. Obviously. The fact that he didn’t either suggests horrible judgement on his part or worse, cowardice. This is a man and leader of one of the biggest and most important businesses in the world who had long been thought to have a great moral compass.
He has lost his way.
«
Among longtime Apple fans, there is now simply a revulsion at Cook’s fealty towards Trump. The idea – which Cook put forward in an internal memo to Apple staff this week – that he can influence Trump by being there is discounted. Why should Trump listen to someone who, in effect, has no leverage over him, but over whom Trump has enormous leverage through tariffs?
unique link to this extract
How Olympic athletes stay healthy during cold and flu season • Outside Online
Alex Hutchinson:
»
Even if you’re really good at avoiding exposure, you’ll encounter some germs eventually—and when you do, you’ll hope that your immune system can deal with them. Getting a flu shot is a good start, as is taking care of basics like getting enough sleep. Jooste points out that athletes are two to three times more likely to pick up an infection when they travel across more than five time zones.
But it’s not just sleep: immune function is also suppressed by stress. In athletes, that can take the form of hard training, but more general life stress also plays a role. In a study published last fall by Sophie Harrison of the University of Bangor, for example, runners who reported higher levels of anxiety were more likely to pick up a respiratory infection in the weeks following a marathon. It’s not always easy to dial back stress in our lives, but understanding its potential consequences is a good motivator to take it seriously.
There are also a near-infinite number of supplements that claim to boost immune function. Few have much evidence behind them, but two that Jooste and his colleagues highlight are vitamin D and probiotics. I’ve generally found the evidence in favor of vitamin D as a sports supplement to be unclear, but keeping your levels above the “deficient” threshold does seem like a good idea. For probiotics, they suggest multi-strain combinations of Lactobacillus and Bifidobacterium at a dose of at least 1 billion CFU per day.
If the first two strategies—avoiding exposure and boosting immunity—fail, then your last hope is to fight an infection as soon as it takes hold. At the 2018 Olympics, the Finnish team took an aggressive approach to identifying and immediately fighting infections, and subsequently detailed the results in a journal paper. (My favorite nugget from that study was that they tracked how infections spread on the flights to South Korea, and found that sitting in business class was the best way to stay healthy and avoid passing an infection on to others.)
«
For this reason I will insist on business class seating in travelling to future commissions.
unique link to this extract
AI chatbots are infiltrating social-science surveys — and getting better at avoiding detection • Nature
Sara Phillips:
»
In November, Sean Westwood, a political scientist at Dartmouth College in Hanover, New Hampshire, demonstrated an AI chatbot that can reliably impersonate a human participant and evade most known mechanisms built into surveys to detect fake responses1.
Westwood used OpenAI’s o4-mini, an AI-based reasoning tool, to build a bot and set it loose on a survey that he designed for the purpose of testing it. In 6,700 tests, the bot managed to pass standard ‘attention check’ questions (which are designed to catch inattentive humans and simple bots) 99.8% of the time.
To help it evade detection, the bot could be programmed with a persona and use reasoning in line with that persona. For example, when it was programmed to answer as an 88-year-old woman and was asked about time spent at children’s sporting events, the bot said that it spent little time at them because its children had grown up. And it remembered its answers to previous questions.
Westwood’s bot also breezed past common questions that are placed into surveys to trip up bots by detecting capabilities that most people do not have. The bot declined to translate a sentence into Mandarin, for example, and it pretended it could not quote the US constitution verbatim.
The ease with which it evaded detection led Constantine Papas, a blogger and user-experience researcher at a big technology firm based in New York, to declare a “scientific validity crisis”. He wrote that “the foundational assumption of survey research (that a coherent response is a human response) is no longer tenable”.
«
The rest of the article is paywalled, but you get the idea already: surveys can’t be relied on any more. An obvious question is whether we can rely any longer on polling data, so much of which now comes from online surveys (because people won’t pick up the phone to pollsters).
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
The Overspill: when there's more that I want to say 