Sign in

Davide Castelvecchi

@castelvecchi.bsky.social
3K followers 230 following 508 posts

Writer at Nature magazine, covering physical sciences, technology, mathematics, and energy. DMs are disabled courtesy of the UK's so-called Online Safety Act

PostsRepliesMedia
Reposted by Davide Castelvecchi
Catherine Rampell @crampell.bsky.social · 07/10/2026
NYFed: By February 2026, Trump's tariffs had contributed 2.9 percentage points to consumer goods price inflation; without tariffs, goods prices would have fallen slightly. The tariffs raised the cost of imported goods AND their domestically-produced competitors.
libertystreeteconomics.newyorkfed.org
How Fast Do Tariffs Pass Through into Consumer Prices? - Liberty Street Economics
A look at how much of U.S. tariffs reach consumers and how quickly, tracing tariffs through import, domestic producer, and retail prices.
15297140
Davide Castelvecchi @castelvecchi.bsky.social · 07/10/2026
My latest, on OpenAI's dump www.nature.com/articles/d41...
nature.com
OpenAI posts 700 maths preprints online: mathematicians are up in arms
The firm used unreleased models to tackle hundreds of high-profile problems.
078
Davide Castelvecchi @castelvecchi.bsky.social · 06/10/2026
My poor plants!
010
Reposted by Davide Castelvecchi
Geoff Brumfiel @gbrumfiel.bsky.social · 06/10/2026
The IceCube neutrino detector has won the 2026 Nobel Prize in physics. The Trump administration has proposed slashing its budget in half in 2027.
382019705
Davide Castelvecchi @castelvecchi.bsky.social · 06/10/2026
BREAKING - Francis Halzen awarded the 2026 Physics Nobel Prize for his work on IceCube and neutrino astronomy www.nature.com/articles/d41...
nature.com
Nobel physics prize awarded for detection of ghostly neutrinos
Francis Halzen wins the 2026 Nobel Prize in Physics for his work on a pioneering high-energy neutrino observatory at the South Pole.
060
Reposted by Davide Castelvecchi
Ars Technica @arstechnica.com · 05/10/2026
arstechnica.com
AI glasses face their first major government crackdown
Norway wants time to write permanent rules for glasses that can record bystanders.
47328
Davide Castelvecchi @castelvecchi.bsky.social · 02/10/2026
Here, I fixed it 😅
080
Davide Castelvecchi @castelvecchi.bsky.social · 02/10/2026
... which reminded me of this
180
Davide Castelvecchi @castelvecchi.bsky.social · 02/10/2026
The wrongness of the conic section in this picture hurts my brain. The flashlight's beam appears nearly parallel to the ground. But then, that would produce a much, much more elongated (high eccentricity) ellipse
2130
Davide Castelvecchi @castelvecchi.bsky.social · 02/10/2026
Did you summon the ghost of Sam Goudsmit? 😁
110
Davide Castelvecchi @castelvecchi.bsky.social · 01/10/2026
The one thing I'm really hating about the explosion of activity in math and AI is how much time it's forcing me to spend on the elonsphere site
130
Reposted by Davide Castelvecchi
Luke LeBrun @lukelebrun.ca · 26/09/2026
Tech companies suddenly decide they can, in fact, moderate content on their platforms after all
7359802001
Reposted by Davide Castelvecchi
Elizabeth Gibney @lizziegibney.bsky.social · 29/09/2026
Such is AI right now that I spoke to Mark Chen on Thurs last week & by Fri it was out of date (see the update at the end)🧪🤖 It was interesting to hear about their approach to safety (which is much boiled down here)& plans to interact with the scientific community 1/6 www.nature.com/articles/d41...
nature.com
OpenAI’s research chief talks of ‘cultural reset’ after wild few weeks
Mark Chen discusses how the Hugging Face cybersecurity incident prompted a pivot to safety — and giving AI models a sense of ‘taste’.
332
Reposted by Davide Castelvecchi
Evan Hensleigh @futuraprime.net · 29/09/2026
OpenAI has the same problem (and it looks even worse, since unlike the Anthropic case they were clearly aware of competing work). Now that mathematicians know OpenAI wants to beat them, will they still trust Codex?
112
Reposted by Davide Castelvecchi
Max Kozlov @maxkozlov.bsky.social · 25/09/2026
SCOOP: Top researchers at Columbia, Stanford, and others thought they were elected to elite scientific societies alongside Nobel laureates. The catch? The organizations are a sham. I spent the last month diving deep into a massive academic scam spanning continents, for @nature.com
nature.com
Exclusive: Sham scientific societies are misleading star researchers
A Nature investigation reveals a network of organizations that have been building their ranks through deceptive practices.
10310186
Reposted by Davide Castelvecchi
Ed Morrish @edmorrish.com · 24/09/2026
here’s the original, to save you scrolling: bsky.app/profile/roel...
114434
Reposted by Davide Castelvecchi
Ars Technica @arstechnica.com · 24/09/2026
arstechnica.com
New Jersey fines data center $1.1M after satellite pics expose 62 gas generators
Million-dollar fines won’t end billionaires’ data center pollution, neighbors fear.
158933
Davide Castelvecchi @castelvecchi.bsky.social · 25/09/2026
OK but it's just linear algebra :-D
010
Reposted by Davide Castelvecchi
Rod Van Meter @rdvquantum.bsky.social · 25/09/2026
By then we have had chapters on pure and mixed states, entanglement, lasers, interference and waveguides. For a Creative Commons licensed preview, see arxiv.org/abs/2311.02367
arxiv.org
Quantum Communications
The second quantum revolution has been picking up momentum over the last decade. Quantum technologies are starting to attract more attention from governments, private companies, investors, and public....
131
Davide Castelvecchi @castelvecchi.bsky.social · 25/09/2026
But p. 26 of Mike and Ike is still part of the overview, no? before they start introducing the reader to quantum mechanics :)
110
Reposted by Davide Castelvecchi
🌬Breeze @cool-summer-breeze.bsky.social · 24/09/2026
This is the Stolen Neighbors Quilt. Each of the 11,000 squares represents a person taken by ICE in Minnesota.
125123755272
Reposted by Davide Castelvecchi
Helen Pearson @helenpearson.bsky.social · 23/09/2026
NEW: There are widespread concerns that AI will reduce our critical-thinking skills. My story for @nature.com today looks at how to maintain those skills — and teach them to students. I found some fascinating science on critical thinking + useful tips for us all 🧠 www.nature.com/articles/d41...
nature.com
How to stay smart in the age of AI: the science of critical thinking
There is growing concern that AI can blunt memory and reasoning. But science shows ways to keep the brain sharp.
12916
Reposted by Davide Castelvecchi
Nature @nature.com · 23/09/2026
This weird form of boron is a million times more conductive than chemists expected go.nature.com/4dSNi1E
go.nature.com
This weird form of boron is a million times more conductive than chemists expected
A boron phase prepared under pressure can be stretched.
0369
Reposted by Davide Castelvecchi
Elizabeth Gibney @lizziegibney.bsky.social · 23/09/2026
An explainer answering quite *how* AI is supposed to be able to kill us all, why Anthropic is pushing for regulation, what to really worry about, and what might happen next 🧪🤖 www.nature.com/articles/d41...
nature.com
Will AI really kill us all? The science behind the hype
Nature examines whether the technology might spell the end for humans, and why AI companies are calling for a slowdown.
0135
Davide Castelvecchi @castelvecchi.bsky.social · 23/09/2026
Are we talking about the same Mahmood? Perhaps she would say the knife guy had "legitimate" concerns www.theguardian.com/politics/202...
theguardian.com
Shabana Mahmood swears at ‘white liberal’ hecklers over Reform remarks
Home secretary accuses protesters of trying to ‘delegitimise’ concerns people have over immigration
221
Reposted by Davide Castelvecchi
Rebecca Solnit @rebeccasolnit.bsky.social · 22/09/2026
"The disappeared*" was a term for those the military dictatorship in Argentina imprisoned/tortured/murdered during the dirty war. *"detenidos-desaparecidos" (detained-disappeared)
20854480
Reposted by Davide Castelvecchi
Jeff Sharlet @jeffsharlet.bsky.social · 21/09/2026
Last week all my students hand-wrote their own pledges, according to their terms, to write in my courses without AI. Some of them were quite moving—their anger at AI’s attempt to steal their educations, their determination to learn & create anyway.
6794182
Reposted by Davide Castelvecchi
Catherine Rampell @crampell.bsky.social · 21/09/2026
EPA allowed pesticide giant Bayer to pick from a menu of regulation options for the controversial and destructive weedkiller dicamba. Dicamba is drift prone and kills vegetation and crops for miles around fields where it is sprayed, and is linked to liver cancer and Non-Hodgkin lymphoma
theguardian.com
US environment agency let Bayer pick regulations for destructive weedkiller, documents show
Environmental Protection Agency asked firm to choose rules it wanted for dicamba, then put the choices in a proposed federal rule
20368221
Davide Castelvecchi @castelvecchi.bsky.social · 21/09/2026
Hey! 😅
110
Reposted by Davide Castelvecchi
Adam Becker @adambecker.bsky.social · 20/09/2026
Speaking as a journalist: I would really like to see more of my colleagues asking this question. If we take money from tech companies we can't report on them, and if somehow we do we have to disclose it. Yet when we interview academics about AI, I almost never see disclosures of their funding.
212433
Reposted by Davide Castelvecchi
Elizabeth Gibney @lizziegibney.bsky.social · 18/09/2026
Is AI making major discoveries in maths by learning from chats of human mathematicians who are already working on the problems? 🧪 It's not clear... but might be worth checking your user settings Great story by @castelvecchi.bsky.social www.nature.com/articles/d41...
nature.com
Who gets credit in the AI era? OpenAI maths bombshell sparks debate
Controversy around OpenAI’s claim to have solved the Navier–Stokes problem highlights how researchers could be inadvertently sharing — and absorbing— ideas through chatbots.
0228
Reposted by Davide Castelvecchi
Adam Becker @adambecker.bsky.social · 17/09/2026
I just wanted to write about physics! But noooo, a fucking sex cult had to take over Silicon Valley and the entire public discourse about technology and the future while aiding the rise of fascism, and I felt duty-bound to write about that instead LOOK WHAT THEY MADE ME DO
hachettebookgroup.com
More Everything Forever
This "smart and wonderfully readable" (New York Times) exposé shows why Silicon Valley’s heartless, baseless, and foolish obsessions—with escaping death...
8519122
Davide Castelvecchi @castelvecchi.bsky.social · 17/09/2026
As Michael Jacobides, who studies digital ecosystems at the London Business School, put it, “I think academics may have been a little naïve” about what happens when they brainstorm with chatbots. [5/n]
020
Davide Castelvecchi @castelvecchi.bsky.social · 17/09/2026
When you brainstorm with chatbots about your unpublished results, your ideas may end up helping to train the next versions of the models. In most cases, the AI company won't care about that info. But other researchers could unwittingly steal it the next time _they_ brainstorm with that model.[4/n]
130
Davide Castelvecchi @castelvecchi.bsky.social · 17/09/2026
Also, Buckmaster only recently realized that in one of accounts he had not opted out of helping improve future models. To be sure, the time before July 3rd does not include when he and Alpöge made their breakthrough on Euler. But there is an underappreciated aspect in all of this ... [3/n]
110
Davide Castelvecchi @castelvecchi.bsky.social · 17/09/2026
... OpenAI's statement was qualified: it covered “user inputs past July 3rd”. But Buckmaster had been collaborating with Levent Alpöge — and using OpenAI’s Codex — for at least a year. Moreover, Buckmaster told me that he had been using three different ChatGPT/Codex accounts ... [2/n]
110
Davide Castelvecchi @castelvecchi.bsky.social · 17/09/2026
In today's @nature.com story, I report a new facet of the Navier-Stokes controversy. OpenAI's widely reported statement said that it was “categorically impossible” for user data from Tristan Buckmaster to have helped train their internal model. However... 🧵 www.nature.com/articles/d41...
nature.com
OpenAI maths bombshell sparks debate about who gets credit in age of AI
Controversy around OpenAI’s claim to have solved the Navier–Stokes problem highlights how researchers could be inadvertently sharing — and absorbing— ideas through chatbots.
160
Reposted by Davide Castelvecchi
Ars Technica @arstechnica.com · 16/09/2026
David Streever, who sent a scathing three-paragraph email to the then-acting director of ICE, hopes his legal fight will put an end to the agency's intimidation tactics that he believes are intended to silence protestors.
arstechnica.com
It’s OK to tell ICE their actions will haunt them, judge rules in speech fight
Man wins fight to block ICE threat over off-the-cuff angry email.
38521
Davide Castelvecchi @castelvecchi.bsky.social · 16/09/2026
She reminds me a bit of @alexwitze.bsky.social !
000
Davide Castelvecchi @castelvecchi.bsky.social · 16/09/2026
@nature.com's take on last week's events in mathematics — and the controversy over OpenAI's announcement that it had solved the Millennium Prize Problem on the Navier-Stokes equations www.nature.com/articles/d41...
nature.com
AI companies must work with the research community to protect attribution
An era-defining mathematics claim raises questions over how AI tools credit earlier work.
023
Reposted by Davide Castelvecchi
Adam Becker @adambecker.bsky.social · 16/09/2026
Nice to see this getting some attention. It's completely true, and I dearly wish more journalists understood where these common narratives about AI actually come from.
10817225
Davide Castelvecchi @castelvecchi.bsky.social · 15/09/2026
www.wired.com/story/zuckof...
wired.com
ZuckOff Is a Free App That Sees Meta Glasses Before They See You
You can’t always tell if someone is recording you with their smart glasses, but your phone can help.
020
Davide Castelvecchi @castelvecchi.bsky.social · 15/09/2026
With two solved, there are still five Millennium Prize Problems to go. What are they, and which ones are likely to fall to AI? Great story by @manonym025.bsky.social www.scientificamerican.com/article/whic...
scientificamerican.com
Which million-dollar math problem could AI solve next?
After an OpenAI model presented a possible solution to the Navier-Stokes problem, mathematicians are speculating about which major unsolved question could be resolved next
010
Davide Castelvecchi @castelvecchi.bsky.social · 15/09/2026
Advocates of peaceful European federalism didn't stop developing those ideas even when jailed during fascism and WWII en.wikipedia.org/wiki/Altiero...
en.wikipedia.org
Altiero Spinelli - Wikipedia
100
Davide Castelvecchi @castelvecchi.bsky.social · 15/09/2026
@eaterofsun.bsky.social makes a good point but I disagree with his conclusion. Multilateralism will only be truly dead once people give up on advocating for it. And having a serious discussion about such solutions lays the groundwork for (optimistically, I'll admit) staging its comeback some day
120
Reposted by Davide Castelvecchi
Terence Tao @teorth.bsky.social · 11/09/2026
A group of 25 Fields Medalists, including myself, have made a joint declaration on Math and AI: mathandai.org . We welcome additional signatories. See also this article in the Economist announcing the declaration: www.economist.com/science-and-...
mathandai.org
Declaration — Math and AI
Read the declaration and add your name.
422075934
Reposted by Davide Castelvecchi
Philip Ball @philipcball.bsky.social · 11/09/2026
I just did an interview for this yesterday, which reminds me that I am supposed to be doing that author thing of saying you can pre-order now. Looks like the US publication date is December. press.uchicago.edu/ucp/books/bo...
press.uchicago.edu
The Man Who Broke Reality
From the brilliant Philip Ball, a thrilling biography of Nobel Prize–winning Danish physicist Niels Bohr, the most radical quantum mechanician of all.In the first half of the twentieth century, the tr...
2184
Davide Castelvecchi @castelvecchi.bsky.social · 09/09/2026
(the link to my story was at the bottom of my thread)
010
Davide Castelvecchi @castelvecchi.bsky.social · 09/09/2026
Indeed, as I mentioned in my story. Of course it will be up to Clay to offer the prize to them (and/or perhaps to somebody else) or not, and then we'll see.
111
Davide Castelvecchi @castelvecchi.bsky.social · 09/09/2026
Also, "CMI does not accept direct submission of proposed solutions," the rules state. The evaluation of solutions to Millennium Prize Problems "is deliberately unhurried, and we shall ensure that it is absolutely rigorous,” CMI president Martin Bridson told me: www.nature.com/articles/d41...
nature.com
OpenAI claims huge maths breakthrough on a famed ‘Millennium Problem’
The company says it has cracked the physics of fluids using AI, solving one of the toughest puzzles in mathematics.
000