michfrapp (@michfrapp) header
michfrapp (@michfrapp) profile picture
Joined July 2007
“In praise of folly”. Over 500 years old, but so timely relevant
1
1
3
3,777
« Bonjour Londres » pour @PauseAI
They used to say America and China working together on AI was impossible. Well, the impossible is starting to happen. But it is not enough. Anything short of an international treaty to @PauseAI development is not enough. My speech today at the March Against AI Extinction in London.
3
It had to be a French leading the anti AI movement. We’re always a step ahead… in the wrong direction.
Le patron mondial de @PauseAI est un Français @FournesMaxime Il a été interviewé par @Alex_Tsico le jeune spécialiste des agents
8
Not sure it will stop market from going down
BREAKING: Taiwan announces new restrictions on short selling after its stock market falls -18% in one week, its worst weekly decline in 35 years. Effective Monday, regulators will ban short sales below the previous closing price and raise margin requirements to 150%. Taiwan’s $15 billion stock stabilization fund has also been buying shares to support the market. Authorities are intervening to contain selling pressure.
2
Congrats @claudeai! What is the next frontier you’ll reach?
The Lean FRO and Mathlib Initiative have released a joint statement on the Lean verification of the Riemann Hypothesis. On November 20, Levent Alpöge released a proof of the Riemann Hypothesis as a Lean library of 17 million lines of code. Over the past day, the Lean and Mathlib communities worked together to verify it. The statement certifies the proof as correct and describes how it meets all necessary verification criteria. Congratulations to Levent on this achievement, and thank you to everyone in the community who contributed to the verification. Full statement: leanprover-community.github.…
6
J’avoue que dans cette histoire c’est ce qui est le plus fascinant: les jeunes n’ont pas compris que cela allait contre leur intérêt de garder l’âge actuel de la retraite (et pire encore de le rabaisser).
Je le répéterai aussi à l’infini , avec beaucoup d’autres: les retraités les plus aisés vivent aux crochets de leurs enfants et petits-enfants et de tous les autres jeunes français . La solidarité des grandes fortunes est nécessaire; elle ne suffira pas. Et pourtant la majorité des jeunes Français s’apprêtent à voter pour deux partis qui veulent les spolier davantage, en abaissant l’âge de la retraite: RN et LFI sont les ennemis des jeunes.
5
And it’s done. AI = 1 / Humans = 0 After mathematics what will it master?
“We have full confidence that Alpöge's proof of the Riemann Hypothesis is correct, and we congratulate him on this achievement.” The Lean community has spoken. GRH has fallen. What a historic moment. Full read: math.mit.edu/~drew/grh/
5
Looks like the pointed jokes on his French accent haven’t mended the relationship (NB: it is true he has a strong accent, which is rather endearing :) )
Mistral’s Arthur Mensch on Jensen a few days ago: “I’d be very glad to see him come back…I would love to see him come back.”
28
And it continues
The FBI is investigating an apparent hack of its nationwide criminal information database. It's the second breach of the agency in recent months. cnn.it/4j7jPmM
6
Well there’s once again clear signs there’s not just a copper cables shortage going on.
BREAKING: The US Department of Agriculture was hit by a cyberattack this week, raising concerns that hackers accessed "sensitive information" at the agency, per NBC News. The USDA informed employees on Thursday that a system the agency uses to facilitate payments to farmers was breached. They warned that accessing the payments system could provide hackers with a backdoor to reaching more sensitive material. It is not yet known who is responsible for the hack.
9
I think we picture here the future of humanity. It’s truly stunning and life changing technology. But not sure it’s the best application of it one could do.
Playing GTA VI with a Neuralink
5
Bon la bonne nouvelle c’est qu’il y a le choix entre la crise… ou la crise. Me voilà pleinement rassuré sur le futur, proche, très proche.
En réalité, ce qu’on ne dit pas c’est qu’il n’y a probablement que deux issues possibles et les deux seront source, pour des raisons opposées, d’une crise économique majeure : 1. La valorisation de l’IA est une bulle et cette bulle éclate et entraîne tout le monde dans sa chute ; 2. La valorisation de l’IA est conforme aux gains de productivité qui justifieront le prix qu’il faudra payer pour amortir les infrastructures, mais alors pour payer ce prix les entreprises licencieront ou gèleront leurs embauches au minimum pendant toute la période où cela restera un avantage compétitif, créant du chômage massif chez les cols blancs qui ont le plus fort pouvoir d’achat. En gros on va soit vers une crise financière qui déclenchera une crise sociale, soit vers une crise sociale qui déclenchera une crise économique. J’ai du mal à entrevoir un scénario alternatif où l’on évite toute crise, ce qui est sans doute consubstantiel à toute révolution technologique.
7
Protecting the walled garden
As Apple develops its own computer-use agent, it’s making it harder for rival agents to control Macs. In macOS 27.3, enabling ChatGPT’s third-party computer-use support now requires 14 distinct steps, up from four. Apple cites security; Some users see anti-competitiveness.
2
Nope after India, we’re back to the US.
🇺🇸 A swarm of AI hackers broke into a company and locked down its whole system in 14 minutes. "This stuff is happening, and it's happening quicker than we can typically respond to it." Former NSA hacker David Kennedy says more than 30 AI agents hit at once, slipped past two-factor security, and walked off with the data. The company didn't even notice for 4 and a half hours. 14 minutes to break in and hours before anyone noticed. That's the scary part. 🥶 Writer: Daniyal
4
Ah for once it’s not Europe or the US. Good to see different news for once and talk about the rest of the World. Today it’s India cybersecurity day. Tomorrow is Brazil?
9
Comme quoi ça paye bien d’être âgé
Dans le Financial Times de ce matin:
9
We all need a Trumpy Nappy Machine.
Do you want the same deep, untroubled sleep the president gets in his meetings? Try the Trumpy Nappy White Noise Machine
5
Is there some hidden causality right there…
10
Entièrement d’accord
C'était ironique. Je répète depuis 4 ans qu'apprendre à prompter ne servirait à rien car l'IA le ferait elle même (c'est fait) Et se former à l'IA non plus : d'abord les entreprises de formation IA n'en savent pas plus que vous (lancées par opportunisme en 2022 sans aucune connaissance) et les formations sont écrites par Chatgpt/Claude.. Si vous ne savez pas utiliser l'IA, demandez lui elle vous expliquera. Sans parler du fait que les agents réalisent directement vos tâches... Mais surtout : l'IA sera de plus en plus autonome. C'est la première fois dans l'histoire qu'une technologie transforme tout et qu'on n'a pas besoin d'y être formés. C'est littéralement le but de l'IA.
4
The symbol of today’s World
BREAKING: Rockstar Games says Grand Theft Auto VI sold $3.79 billion in 24 hours, making it the biggest entertainment launch of all time. Pre-sales of the game reached 32.5 million units
8
Always mesmerising to listen to @POTUS Lucky us that he resolved that ATC antiquated systems issue, working actively during the 6 years he’s been in office… cough… cough… cough… oups sorry. Well at least we have a (half built) ballroom and a (somewhat) reflective pool.
🇺🇸Trump had some profoundly weird shit to say about ATC: “IT’S ANCIENT! In fact, you know, before I was here, I had a plane. I had a very nice life. And my pilot would use the air traffic control from a totally different country. This was years ago, before I was in office. I don't want to say the country, but he found one country better than ours.” What does that even mean? Can you just use a foreign country’s ATC in the U.S.? We’re going to need a pilot to break this all down! Writer: Jamie
8