If you want to experience supersonic sound, start with some Sellotape. CC-licensed photo by Megan Young 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. Unravelling. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.
The DJI Romo robovac had security so poor, this man remotely accessed thousands of them • The Verge
Sean Hollister:
»
Sammy Azdoufal claims he wasn’t trying to hack every robot vacuum in the world. He just wanted to remote control his brand-new DJI Romo vacuum with a PS5 gamepad, he tells The Verge, because it sounded fun.
But when his homegrown remote control app started talking to DJI’s servers, it wasn’t just one vacuum cleaner that replied. Roughly 7,000 of them, all around the world, began treating Azdoufal like their boss.
He could remotely control them, and look and listen through their live camera feeds, he tells me, saying he tested that out with a friend. He could watch them map out each room of a house, generating a complete 2D floor plan. He could use any robot’s IP address to find its rough location.
“I found my device was just one in an ocean of devices,” he says.
On Tuesday, when he showed me his level of access in a live demo, I couldn’t believe my eyes. Ten, hundreds, thousands of robots reporting for duty, each phoning home MQTT data packets every three seconds to say: their serial number, which rooms they’re cleaning, what they’ve seen, how far they’ve traveled, when they’re returning to the charger, and the obstacles they encountered along the way.
I watched each of these robots slowly pop into existence on a map of the world. Nine minutes after we began, Azdoufal’s laptop had already cataloged 6,700 DJI devices across 24 different countries and collected over 100,000 of their messages. If you add the company’s DJI Power portable power stations, which also phone home to these same servers, Azdoufal had access to over 10,000 devices.
…I asked my colleague Thomas Ricker, who just finished reviewing the DJI Romo, to pass us its serial number.
With nothing more than that 14-digit number, Azdoufal could not only pull up our robot, he could correctly see it was cleaning the living room and had 80% battery life remaining. Within minutes, I watched the robot generate and transmit an accurate floor plan of my colleague’s house, with the correct shape and size of each room, just by typing some digits into a laptop located in a different country.
«
The loophole was quickly fixed, but it’s telling: yet another Chinese company with absolutely appalling security. DJI has been around long enough that it’s hard to think this is just an oversight.
unique link to this extract
Scientists crack the case of “screeching” Scotch tape: it’s supersonic • Ars Technica
Jennifer Ouellette:
»
Scotch tape has been a household mainstay for nearly a century, but it still holds some scientific surprises. Researchers have discovered that the screeching sound emitted when one rapidly peels Scotch tape—akin to the screech of fingernails on a chalkboard—is the result of shock waves produced by micro-cracks propagating along the tape at supersonic speeds, according to a new paper published in the journal Physical Review E.
It was a 3M engineer named Richard Drew who developed the first transparent sticky tape in 1930. The impetus came from car manufacturing, specifically two-color designs, where the adhesives used were so sticky they often removed the paint when peeled off; the paint then needed to be manually touched up. Drew found a sandpaper adhesive with just the right amount of stickiness and used it to coat a roll of cellophane tape. (Fun fact: Drew also co-invented the snail-style dispenser for the tape with his 3M colleague, John Borden.) The tape was hugely popular during the Great Depression; consumers used it to repair everyday items rather than replace them. That popularity has never waned.
Scotch tape has also generated considerable interest among physicists. …In 1953, Russian scientists peeling Scotch tape in a vacuum reported detecting electrons with sufficient energy to emit X-rays. Other scientists were skeptical, but this phenomenon was finally confirmed in 2008, when UCLA physicists produced X-rays while unwinding a roll of Scotch tape in a vacuum chamber.
…Thoroddsen et al. wondered whether the sound was directly generated by a crack’s rapidly moving tip, which would also produce the distinctive discrete sound wave pulses associated with peeling Scotch tape. The authors experimentally tested their hypothesis by conducting simultaneous high-speed imaging of the propagating fractures and the sound waves traveling in the air.
…Their results showed that the screeching arises from a train of weak shocks that culminate when the transverse cracks reach the edge of the tape. The supersonic speed at which they travel, relative to the surrounding air, is crucial to the generation of those shockwaves.
«
Must be an early contender for an IgNobel prize – for work that cannot or should not be repeated. Unless we’re going to have Sellotape aircraft travelling at supersonic speeds.
unique link to this extract
Reddit and Discord in trouble over controversial age verification • 9 to 5 Mac
Ben Lovejoy:
»
Both Reddit and Discord have got themselves into trouble over the use of a controversial third-party age verification service.
Reddit has been fined £14.5m ($19.5m) for the unlawful use of children’s personal information, while Discord has experienced a user backlash.
The UK’s Online Safety Act (OSA) requires online platforms to verify the age of its users and ensure that children are not exposed to harmful content. BBC News reports that an investigation by the UK’s privacy watchdog found that Reddit failed to adequately carry out these checks. The company used an outside firm called Persona which checks either an uploaded selfie or a photo of government ID. Reddit said it did so because it didn’t want to have first-hand knowledge of its users’ identities.
However, the Information Commissioner’s Office (ICO) said that these checks were inadequate, with many children being declared adults, and that the company was therefore illegally processing data from children.Discord used the same company but quickly faced anger from users who pointed out that Persona appears to access government records.
…There have been growing calls for the app stores of Apple and Google to be made legally responsible for age verification and then to apply age gating to app downloads in response. This is already the case in some US states.
«
Certainly makes sense to put the onus up on to the ownership of the phone, which Apple or Google would know about.
unique link to this extract
The AI bubble: hidden risks and opportunities • Man Group
Paul Kedrosky, Sumant Wahi and Alex Preston:
»
Hyperscalers can now push liabilities outward onto special purpose vehicles, infrastructure funds, private credit, insurance company balance sheets and even, indirectly, into data centre real estate income trust ETFs.
The off-balance-sheet financing structures being deployed contain a fundamental and dangerous expectation mismatch. Private equity (PE) and venture capital (VC) investors are placing their bets with a familiar model: hoping for one in 10 investments to deliver 100x returns. This power-law approach has served them well in software and internet investing, where marginal costs are near-zero and winner-take-all dynamics prevail.
Meanwhile, creditors funding these projects – banks, insurers, pension funds, private credit vehicles – believe they are financing long-lived infrastructure assets comparable to commercial real estate or utility-grade equipment. Their underwriting models assume seven-to-15-year useful lives, with stable cash flows and recoverable collateral values.
However, the effective economic life of GPU and ASIC chips is approximately one year. A data centre filled with NVIDIA H100s in 2024 faces severe competitive disadvantages against one with Blackwell chips in 2025 and potential obsolescence with the next architecture.This means that depreciation schedules are too long, collateral values in default are illusory, and cash flow assumptions are fragile. The VC and PE investors understand this: their short time horizons and equity-like returns compensate for high risk. But lenders do not appreciate the precipitous drop in asset values and token pricing they are exposed to. This duration and usage-risk mismatch is being masked by the financing templates and the apparent credit quality of the hyperscaler tenants. We believe it represents a ticking time bomb in credit markets.
While demand for tokens is soaring, that demand isn’t rising nearly quickly enough to make up for collapsing unit economics, as efficiency gains drive down token costs more than 70% per year (Figure 5). Merely standing still would require an increase of more than 225% in tokens demanded per year.
«
Kedrosky is definitely a voice to listen to. The authors point to what they call an obvious demand gap: “US$200bn capex versus US$12bn in revenue”. That is quite a gap.
unique link to this extract
OpenAI resets spend expectations, targets around $600bn by 2030 • CNBC
Ashley Capoot and Kate Rooney:
»
OpenAI is telling investors that it’s now targeting roughly $600bn in total compute spend by 2030, months after CEO Sam Altman touted $1.4 trillion in infrastructure commitments.
The artificial intelligence company is providing a lower number and more defined timeline for its planned spending, sources told CNBC, as broader concerns mounted that expansion ambitions were too great for the potential revenue that would follow.
OpenAI is projecting that its total revenue for 2030 will be more than $280bn, with nearly equal contributions from its consumer and enterprise businesses, said the sources, who asked not to be named because the information is private. The spending plan the company is offering is meant to more directly tie to its expected revenue growth, the people said.
In the back half of last year, OpenAI announced a flurry of multibillion-dollar infrastructure deals, partnering with leading chipmakers and cloud companies.
OpenAI is finalizing a massive funding round that could total more than $100bn, with about 90% coming from strategic investors, one person said. Nvidia is in discussions to invest up to $30bn in OpenAI as part of the round that could value the company at a $730bn pre-money valuation, CNBC has confirmed.
«
The numbers here are absolutely crazy, but what is really noticeable is how they’re going down. That topline figure has more than halved. Is this them deflating the bubble carefully?
unique link to this extract
Wearable trackers can detect depression relapse weeks before it returns, study finds • Medical Xpress
»
Could a smart watch act as an early-warning system for depression relapse? New research from McMaster University suggests that disruptions in a person’s sleep and daily activity routine, as detected through a simple wrist-worn device, can signal when there is increased risk of relapsing into major depression.
The new research highlights a simple, yet powerful way to passively monitor relapse risk in people living with major depressive disorder (MDD), often detecting the probability of a relapse weeks or months before the episode occurs. Approximately 60% of people with MDD relapse within five years, even with treatment.
“Advances in digital technology and AI algorithms have a great potential for relapse prevention in mental health. Imagine a future where a smartwatch can warn people with depression: ‘A new episode of depression is very likely coming within the next four weeks. How about seeing your health-care provider?'” says Benicio Frey, professor in the Department of Psychiatry and Behavioral Neurosciences at McMaster.
The study, published in JAMA Psychiatry, followed 93 adults across Canada who had previously recovered from depression and wore a research-grade actigraphy device, similar to a Fitbit or Apple Watch. Participants wore the wearable device for one to two years, generating over 32,000 days of sleep and activity data.
…This research underscores the untapped potential of wearable technology for people recovering from MDD because it collects data passively and could provide continuous insight between clinical appointments. Current monitoring relies heavily on symptoms, which usually appear later than what can be detected through a wearable device.
Researchers say there are opportunities for health system innovation, where wearable-derived alerts could help clinicians target care to those most at risk, improving outcomes and reducing the burden of recurrent episodes.
«
Machines to watch over us.
unique link to this extract
Apple quietly removes environmental metrics from executive pay • Los Angeles Times
Ben Elgin and Jeff Green:
»
Two years ago, many of America’s largest companies began stripping diversity targets out of executive pay packages. Now, environmental measures — including goals tied to climate emissions — are beginning to face a similar fate.
Apple quietly dropped a so-called “ESG modifier” from its 2025 pay packages for chief executive Tim Cook and other top executives, according to a corporate filing last month. The provision, in place since 2021, had allowed Apple’s board to adjust annual bonuses up or down by as much as 10% depending on the company’s performance on a variety of measures, including greenhouse gas reductions and renewable energy use among suppliers.
Apple’s move follows similar decisions at dozens of companies, including Starbucks, Salesforce, Mastercard. and Procter & Gamble, which have recently weakened or severed ties between environmental performance and the size of their executives’ paychecks.
The shift from environmental, social and governance measures is beginning to show up in the numbers. The share of S&P 500 companies tying executive compensation to environmental metrics fell to 46.7% in 2025, down from a peak of 52.6% two years earlier, according to figures from the Conference Board and analytics firm Esgauge.
The decline is modest compared with the stampede away from diversity metrics, which appeared in nearly three-quarters of S&P 500 pay plans in 2023 before plummeting to 34% last year. Still, recent high-profile retreats from climate-linked pay could embolden other companies to follow suit, according to Brian Bueno, sustainability practice leader at Farient Advisors, an executive compensation consulting firm.
«
The reason is easy to guess at: Apple Intelligence is going to consume so much energy that it’s not going to be possible to stay inside those targets. This was predicted by John Siracusa on the ATP podcast (if that’s not tautological?) a few weeks ago: AI is not environment-friendly.
unique link to this extract
How can infinity come in many sizes? • Quanta Magazine
Mark Belan and Jordana Cepelewicz:
»
nfinity invites resistance. Aristotle rejected the existence of the infinite entirely; to him, infinity was simply a limit that could never be reached, not a true mathematical entity. In the early 17th century, Galileo wrote that typical ways of thinking about sets and numbers held no meaning in the realm of the infinite, and that mathematicians would only find paradoxes if they tried to apply their usual tool kit to it. And when, 200 years later, Georg Cantor formalized the idea that infinity comes in many sizes, he was met with anger and fear. His colleagues dismissed his work as that of a madman.
But in time, Cantor’s work on sets and infinity would form the bedrock of modern mathematics. As David Hilbert, another mathematical great, later wrote: “No one shall expel us from the paradise that Cantor has created for us.”
So how can infinity have different sizes?
Welcome to Cantor’s paradise.
«
Nice visual tour through a few of the infinities available to us.
unique link to this extract
How AI helps break the cost barrier to COBOL modernization • Claude
»
COBOL is everywhere. It handles an estimated 95% of ATM transactions in the US. Hundreds of billions of lines of COBOL run in production every day, powering critical systems in finance, airlines, and government.
Despite that, the number of people who understand it shrinks every year.
The developers who built these systems retired years ago, and the institutional knowledge they carried left with them. Production code has been modified repeatedly over decades, but the documentation hasn’t kept up. Meanwhile, we aren’t exactly minting replacements—COBOL is taught at only a handful of universities, and finding engineers who can read it gets harder every quarter.
Given these roadblocks, how can organizations modernize their systems without losing the reliability, availability, and data they’ve accumulated over decades? And without breaking anything?
COBOL modernization differs fundamentally from typical legacy code refactoring. You aren’t just updating familiar code to use better patterns, you’re reverse engineering business logic from systems built when Nixon was president. You’re untangling dependencies that evolved over decades, and translating institutional knowledge that now exists only in the code itself.
Modernizing a COBOL system once required armies of consultants spending years mapping workflows. This resulted in large timelines and high costs that few were willing to take on.
AI changes this.
Tools like Claude Code can automate the exploration and analysis phases that consume most of the effort in COBOL modernization. These tools can:
1: Map dependencies across thousands of lines of code
2: Document workflows that nobody remembers
3: Identify risks that would take human analysts months to surface
4: Provide teams with the deep insights they need to make informed decisionsWith AI, teams can modernize their COBOL codebase in quarters instead of years.
«
I’d really want them to do a lot of regression testing on that refactored code, because if ATMs stop working correctly, there’ll be hell to pay. And yet: this blogpost made IBM’s stock drop by 10% when it appeared, and it hasn’t recovered since; it’s now at a 12-month low.
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 