Baseball has finally started to make technological shifts, with the NY Yankees using bats shaped to improve hitting. CC-licensed photo by terren in Virginia 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 10 links for you. Based. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.
Open-source genetic database shuts down to protect users from “authoritarian governments” • 404 Media
Jason Koebler:
»
The creator of an open source genetic database is shutting it down and deleting all of its data because he has come to believe that its existence is dangerous with “a rise in far-right and other authoritarian governments” in the United States and elsewhere.
“The largest use case for DTC genetic data was not biomedical research or research in big pharma,” Bastian Greshake Tzovaras, the founder of OpenSNP, wrote in a blog post. “Instead, the transformative impact of the data came to fruition among law enforcement agencies, who have put the genealogical properties of genetic data to use.”
OpenSNP has collected roughly 7,500 genomes over the last 14 years, primarily by allowing people to voluntarily submit their own genetic information they have downloaded from 23andMe. With the bankruptcy of 23andMe, increased interest in genetic data by law enforcement, and the return of Donald Trump and rise of authoritarian governments worldwide, Greshake Tzovaras told 404 Media he no longer believes it is ethical to run the database.
“I’ve been thinking about it since 23andMe was on the verge of bankruptcy and been really considering it since the U.S. election. It definitely is really bad over there [in the United States],” Greshake Tzovaras told 404 Media. “I am quite relieved to have made the decision and come to a conclusion. It’s been weighing on my mind for a long time.”…OpenSNP has been used for various scientific papers, most notably to show that an earlier paper about chronic fatigue syndrome pulled from 23andMe data could not be replicated and was based on erroneous science.
«
As a reminder, 23andMe had about 15 million genomes stored, so it’s impressive on its own that this dataset could refute work done there. But the world is really changing, and this seems like a very sensible move. Even if it won’t make much difference if the US government “acquires” the 23andMe dataset by some means.
unique link to this extract
Cheap TVs’ incessant advertising reaches troubling new lows • Ars Technica
Scharon Harding:
»
TVs offer us an escape from the real world. After a long day, sometimes there’s nothing more relaxing than turning on your TV, tuning into your favourite program, and unplugging from the realities around you.
But what happens when divisive, potentially offensive messaging infiltrates that escape? Even with streaming services making it easy to watch TV commercial-free, it can still be difficult for TV viewers to avoid ads with these sorts of messages.
That’s especially the case with budget brands, which may even force controversial ads onto TVs when they’re idle, making users pay for low-priced TVs in unexpected, and sometimes troubling, ways.
An experience recently shared by an apparent Vizio TV owner illustrates how ads delivered via TV operating systems (OSes) can take ads from annoying to intrusive and offensive.
Reddit user DoubleJumps claimed last week that when their Vizio TV is idle, “it plays calming nature video, calming music, and then loops a message from the [T]rump admin[istration] telling illegal immigrants to gtfo over and over and over again.”
…what DoubleJumps detailed is completely within the scope of Vizio’s advertising efforts. Vizio TVs have something called Scenic Mode, which has the sets show, per Vizio, “relaxing, ambient content when your TV is idle for a period of time,” along with ads. Scenic Mode can be disabled, but if it’s enabled, the ads cannot be turned off. Vizio says the ads help it pay for things like the TVs’ free ad-supported streaming TV (FAST) channels and help keep Vizio TV prices low.
Vizio also has ties to political ads. It has previously boasted about its work with “a political candidate on an ad campaign that combined CTV ads with our Household Connect omnichannel feature to reach potential voters both on their TV sets and on other opted in devices.” The company says it can play a “powerful role… in helping political campaigns reach their intended audiences.”
«
I’d love to say I’m sympathetic, but if you live in a culture suffused with advertising and never try to erase it, this really is what you’ll get. Apparently the owner is going to get rid of the TV. That’s a start.
unique link to this extract
Earth AI’s algorithms found critical minerals in places everyone else ignored • TechCrunch
Tim De Chant:
»
Earth AI emerged from Teslyuk’s graduate studies. Teslyuk, a native of Ukraine, was working toward a doctorate at the University of Sydney, where he became familiar with the mining industry in Australia. There, the government owns the rights to mineral deposits, and it leases them in six-year terms. Since the 1970s, he said, exploration companies are required to submit their data to a national archive.
“For some reason, nobody’s using them,” he said. “If I could build an algorithm that can absorb all that knowledge and learn from the failures and successes of millions of geologists in the past, I can make much better predictions about where to find minerals in the future.”
Teslyuk started Earth AI as a software company focused on making predictions about potential deposits, then approaching customers who might be interested in exploring sites further. But the customers were hesitant to invest, in part because they didn’t want to bet millions on the predictions of an unproven technology.
“Mining is a very conservative industry,” Teslyuk said. “Everything outside of the approved dogma is considered heresy.”
So Earth AI decided to develop its own drilling equipment to prove that the sites it identified were as promising as its software suggested. The company was accepted to Y Combinator’s spring 2019 cohort, and it spent the next few years refining its hardware and software. In January, Earth AI raised a $20m Series B.«
Slight correction on the optimistic headline: Earth AI’s algorithms claimed to have found critical minerals. Without digging there, nobody knows. The company’s website doesn’t show any proof of discovery, though it’s an interesting idea – why not apply ML to all that data?
unique link to this extract
The Yankees’ ‘Torpedo’ bats that are breaking baseball, explained • SBNation.com
James Dator:
»
The New York Yankees have cracked the code when it comes to hitting dingers, and it’s raising massive questions about whether they’re simply leveraging existing rules, or bending them entirely too much. This year the club introduced a new bat featuring what they call a “Torpedo Barrel,” and it’s quite literally breaking the game of baseball.
This was most apparent on Saturday night in a 20-9 win over the Milwaukee Brewers in which the Yankees hit an astonishing nine home runs across seven players. Even if we take that game as an outlier, New York is still crushing the ball at an unseen rate — more than tripling the competition.
In 2024 Major League Baseball (MLB) teams averaged 182 home runs across the season for an average of 1.12 homers a game. Thus far in 2025 the Yankees already have 15 in three games, on pace for a record-breaking 810. The sample size is small, and obviously that won’t hold — but the result is still staggering as New York leads MLB in home runs, hitting three more than the Dodgers, despite playing two fewer games. That’s largely in part to the Torpedo Barrel.
At the most basic level it’s a new shape of bat. Rather than having an even taper from tip to grip, the Torpedo has a pronounced bulge in the barrel designed to redistribute the center of mass in the bat from the end, moving it down and where players most often make contact. The difference in bats is astonishing when seen side-by-side with a standard MLB bat.
Essentially what this change means is that the hitting of Yankees players is fundamentally altered. They are making contact with the sweet spot far more often, and plays which would normally be tip balls or flares are now converting into home runs.
«
It’s astonishing that it has taken this long – professional baseball has existed in the US since 1876. But these bats, which do what larger rackets did for tennis in the 1970s (enlarge the sweet spot/move it to where you actually make contact) have only been around since 2022. US pro baseball seems to be very resistant to change – right down to the use of a specific sort of mud that is rubbed into baseballs before they’re put into play.
unique link to this extract
Anthropic can now track the bizarre inner workings of a large language model • MIT Technology Review
Will Douglas Heaven:
»
The AI firm Anthropic has developed a way to peer inside a large language model and watch what it does as it comes up with a response, revealing key new insights into how the technology works. The takeaway: LLMs are even stranger than we thought.
The Anthropic team was surprised by some of the counterintuitive workarounds that large language models appear to use to complete sentences, solve simple math problems, suppress hallucinations, and more, says Joshua Batson, a research scientist at the company.
It’s no secret that large language models work in mysterious ways. Few—if any—mass-market technologies have ever been so little understood. That makes figuring out what makes them tick one of the biggest open challenges in science.
But it’s not just about curiosity. Shedding some light on how these models work exposes their weaknesses, revealing why they make stuff up and why they can be tricked into going off the rails. It helps resolve deep disputes about exactly what these models can and can’t do. And it shows how trustworthy (or not) they really are.
Batson and his colleagues describe their new work in two reports published at the end of March. The first presents Anthropic’s use of a technique called circuit tracing, which lets researchers track the decision-making processes inside a large language model step by step. Anthropic used circuit tracing to watch its LLM Claude 3.5 Haiku carry out various tasks. The second (titled “On the Biology of a Large Language Model”) details what the team discovered when it looked at ten tasks in particular.
“I think this is really cool work,” says Jack Merullo, who studies large language models at Brown University in Providence, Rhode Island, and was not involved in the research. “It’s a really nice step forward in terms of methods.”
Circuit tracing is not itself new. Last year Merullo and his colleagues analyzed a specific circuit in a version of OpenAI’s GPT-2, an older large language model that OpenAI released in 2019. But Anthropic has now analyzed a number of different circuits inside a far larger and far more complex model as it carries out multiple tasks. “Anthropic is very capable at applying scale to a problem,” says Merullo.
Eden Biran, who studies large language models at Tel Aviv University, agrees. “Finding circuits in a large state-of-the-art model such as Claude is a nontrivial engineering feat,” he says. “And it shows that circuits scale up and might be a good way forward for interpreting language models.”
«
Fun! But also: does that actually help us? At all?
unique link to this extract
As AI takes his readers, a leading history publisher wonders what’s next • Big Technology
Alex Kantrowitz:
»
Late last year, Jan van der Crabben’s AI fears materialized. His World History Encyclopedia — the world’s second most visited history website — showed up in Google’s AI Overviews, synthesized and presented alongside other history sites. Then, its traffic cratered, dropping 25% in November.
Van der Crabben, the website’s CEO and founder, knew he was getting a preview of what many online publishers may soon experience. His site built a sizable audience with plenty of help from Google, which still accounts for 80% of its traffic. But as AI search and bots like ChatGPT ingest and summarize the web’s content, that traffic is starting to disappear. Now, his path forward is beginning to look murky.
“There used to be this implicit agreement between publishers and Google that basically, Google could scrape, analyze, process, and do whatever they wanted with publishers’ content and in return, they would send traffic to the publishers, send them readers,” he told me. “Now, this unspoken contract is kind of breaking.”
World History Encyclopedia is just one site and Van der Crabben’s anecdote is just one story, but the migration of readers from the web to AI summaries will likely continue. The internet has always favored ease. And using generative AI to find the most valuable parts of the web’s evergreen content — like recipes, personal finance, and history content — can be a better experience than poking through sites one by one.
…Van der Crabben is well aware that Google has helped him scale in the first place, but what’s different now, he said, is that the arrangement is no longer mutually beneficial. “Now it’s just Google that benefits,” he said. “And the same is true, let’s be honest, for ChatGPT, for Anthropic, for many of the AI chatbots that gobble up the content. And, for most of them, unless you sue them — which only a big corporation like Reuters can do — they are not going to pay you for it.”
«
Why does Britain feel so poor? • The Value of Nothing
Martin Robbins:
»
Speaking at a rail conference last year, HS2 Chair Sir Jon Thompson said: “To build a railway between Euston and Curzon Street in Birmingham, I need 8,276 consents from other public bodies, planning, transport, the Environment Agency or Natural England. They don’t care whether parliament did or didn’t approve building a railway.”
Imagine for a moment the sheer human effort, the cost, the entire lives consumed by the gigantic national project of ticking each of those 8,276 checkboxes. It is absolutely correct that we have additional checks and balances, but this is an organically-grown system operating with no overall oversight or coherent strategy, spawning busy-work for thousands of people, much of it duplicated, unnecessary or redundant, much of it – like the infamous bat tunnel – having little provable benefit for things like environmental protection in the first place.
Indeed, improving things like the environment is not the de facto goal of this system, it is not what it was ‘designed’ or incentivised to accomplish. If it were, much of the money would be far better spent. I cannot stress this enough: this is not a battle between, say, infrastructure and the environment. It’s a battle between people who think we can do both, better; and people who seriously believe that a bat tunnel is the best way to spend £120m to support wildlife, a proposition for which no compelling evidence has ever been provided.
Meanwhile, local government veers towards bankruptcy, in large part because they’re mandated to write blank cheques for social care with no support or strategy from central government.
«
Robbins puts his finger on what feels wrong in this post. The box-ticking is certainly a big thing. The inability of anyone to cut through the Gordian knots is becoming destructive.
unique link to this extract
Waltz and staff used Gmail for government communications, officials say • The Washington Post
John Hudson:
»
Members of President Donald Trump’s National Security Council, including White House national security adviser Michael Waltz, have conducted government business over personal Gmail accounts, according to documents reviewed by The Washington Post and interviews with three U.S. officials.
The use of Gmail, a far less secure method of communication than the encrypted messaging app Signal, is the latest example of questionable data security practices by top national security officials already under fire for the mistaken inclusion of a journalist in a group chat about high-level planning for military operations in Yemen.
A senior Waltz aide used the commercial email service for highly technical conversations with colleagues at other government agencies involving sensitive military positions and powerful weapons systems relating to an ongoing conflict, according to emails reviewed by The Post. While the NSC official used his Gmail account, his interagency colleagues used government-issued accounts, headers from the email correspondence show.
Waltz has had less sensitive, but potentially exploitable information sent to his Gmail, such as his schedule and other work documents, said officials, who, like others, spoke on the condition of anonymity to describe what they viewed as problematic handling of information.
«
As a reminder, it was the hacking of Hillary Clinton campaign manager John Podesta’s Gmail account in 2016 which led to a raft of insane stories. Here we aren’t even starting with the hacking.
unique link to this extract
Elon Musk says his AI business xAI has acquired X • Axios
Dan Primack:
»
Elon Musk said on Friday that xAI has acquired X, the social media app previously known as Twitter.
This is Musk trying to prevent his Twitter investors from losing money, and highlights his eagerness to meld different parts of his corporate empire.
Musk said that the all-stock merger valued xAI at $80bn and X at $45bn (including $12bn of debt). A source familiar with the situation says that the combined company would be valued at xAI plus X, absent some small adjustments.
For context, Musk paid $44bn to buy Twitter in late 2022, and some of its investors had marked the shares down by more than 70%. xAI was most recently valued at $50bn, in a late 2024 fundraising round.
The two companies have many shareholders in common, including venture firm Sequoia Capital and mutual fund manager Fidelity.
«
This is just transparent nonsense. There’s no way in the world that X is worth that amount of money compared to when it was sold: not in terms of users, not in terms of advertising it can attract, not in terms of revenue, not in terms of profit. So the number is totally made up, which makes the purchase a misuse of the xAI funds. But the investors won’t care because they’re getting shares in an AI company, and everyone knows they’re the New Fabulous Thing, at least until they turn out to be about as unique and sought-after as NFTs.
unique link to this extract
The Signal chat leak and the NSA • Schneier on Security
Bruce Schneier:
»
It’s common knowledge that the NSA’s mission is breaking into and eavesdropping on other countries’ networks. (During President George W. Bush’s administration, the NSA conducted warrantless taps into domestic communications as well—surveillance that several district courts ruled to be illegal before those decisions were later overturned by appeals courts. To this day, many legal experts maintain that the program violated federal privacy protections.) But the organization has a secondary, complementary responsibility: to protect US communications from others who want to spy on them. That is to say: While one part of the NSA is listening into foreign communications, another part is stopping foreigners from doing the same to Americans.
Those missions never contradicted during the Cold War, when allied and enemy communications were wholly separate. Today, though, everyone uses the same computers, the same software, and the same networks. That creates a tension.
When the NSA discovers a technological vulnerability in a service such as Signal (or buys one on the thriving clandestine vulnerability market), does it exploit it in secret, or reveal it so that it can be fixed? Since at least 2014, a US government interagency “equities” process has been used to decide whether it is in the national interest to take advantage of a particular security flaw, or to fix it. The trade-offs are often complicated and hard.
Waltz—along with Vice President J.D. Vance, Defense Secretary Pete Hegseth, and the other officials in the Signal group—have just made the trade-offs much tougher to resolve. Signal is both widely available and widely used. Smaller governments that can’t afford their own military-grade encryption use it. Journalists, human rights workers, persecuted minorities, dissidents, corporate executives, and criminals around the world use it. Many of these populations are of great interest to the NSA.
At the same time, as we have now discovered, the app is being used for operational US military traffic. So, what does the NSA do if it finds a security flaw in Signal?
Previously, it might have preferred to keep the flaw quiet and use it to listen to adversaries. Now, if the agency does that, it risks someone else finding the same vulnerability and using it against the US government. And if it was later disclosed that the NSA could have fixed the problem and didn’t, then the results might be catastrophic for the agency.
«
| • 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 