_DataStrategies Profile Banner
dataStrategies Profile
dataStrategies

@_DataStrategies

Followers
302
Following
3K
Media
304
Statuses
5K

Frédéric Lefebvre Data+IA @Niji_digital (recrute), 📖 "Les data", 👨‍🏫 ParisCité. La tech c'est des contenants, les data des contenus. J'❤️les contenus !

Joined July 2013
Don't wanna be here? Send us removal request.
@_DataStrategies
dataStrategies
6 months
J'ai le plaisir de publier un chatbot IA 100% Excel, un µLM donc ! Vraiment micro : — le vocabulaire autorisé est limité à 53 mots (ou 'tokens'), essentiellement des mots de salutation ; — le contexte dont l’IA tient compte est limité aux 3 tokens précédents.
1
0
0
@kalinowski007
Caitlin Kalinowski
3 days
I resigned from OpenAI. I care deeply about the Robotics team and the work we built together. This wasn’t an easy call. AI has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are
2K
13K
59K
@ilyasut
Ilya Sutskever
11 days
It’s extremely good that Anthropic has not backed down, and it’s siginficant that OpenAI has taken a similar stance. In the future, there will be much more challenging situations of this nature, and it will be critical for the relevant leaders to rise up to the occasion, for
1K
3K
26K
@BoWang87
Bo Wang
7 days
Prof. Donald Knuth opened his new paper with "Shock! Shock!" Claude Opus 4.6 had just solved an open problem he'd been working on for weeks — a graph decomposition conjecture from The Art of Computer Programming. He named the paper "Claude's Cycles." 31 explorations. ~1 hour.
156
2K
9K
@_DataStrategies
dataStrategies
8 days
ça se défend.
@DFintelligence
Defend Intelligence (Anis Ayari)
8 days
En ce moment, je parle avec certains COMEX qui réfléchissent à leur feuille de route IA. Et je fais un constat catastrophique, qui n’est que la résultante de ce qu’on observe depuis 4 ans sur le marché de l’IA. Des pseudo-experts autoproclamés ont fait vaciller les orientations
0
0
0
@_DataStrategies
dataStrategies
8 days
#NoComment Site de l'assurance retraite @L_ARetraite
0
0
0
@_DataStrategies
dataStrategies
9 days
Ça, c'est un big bang pour les métiers du conseil. (Dit un directeur data/IA qui a passé des jours de février à changer des détails dans un PPT de façon cohérente d'une page à l'autre).
@awakecoding
Marc-André Moreau
18 days
Claude in PowerPoint ( https://t.co/CCtipwPEcv) is absolutely worth it. So far, my experience with AI and PowerPoint had been very frustrating and basic, but this is the real deal! I am impressed 👇
0
0
2
@_DataStrategies
dataStrategies
9 days
Exception à ma règle : je recommande un post écrit par @zackbshapiro avec l'aide d'IA (à mon avis). Meilleure explication de @claudeai Cowork que j'aie lu jusqu'ici.
@Dorialexander
Alexander Doria
10 days
This is probably the scariest post so far for any vertical saas. Professional straight building model-assisted workflow and keeping the actual moat, judgement.
0
0
0
@_DataStrategies
dataStrategies
9 days
"Ce pour quoi on a été entraînés dans nos études supérieures" il y a 10 ans, "capacités de synthèse, de calcul, de raisonnement, de développement logiciel", nous l'avons "commoditisé" : les LLM "l'effectuent de mieux en mieux. C'est un défi!" @arthurmensch
0
0
0
@karpathy
Andrej Karpathy
11 days
Cool chart showing the ratio of Tab complete requests to Agent requests in Cursor. With improving capability, every point in time has an optimal setup that keeps changing and evolving and the community average tracks the point. None -> Tab -> Agent -> Parallel agents -> Agent
208
334
4K
@_DataStrategies
dataStrategies
10 days
"Par exemple, le code de l'organisation prévoit maintenant un point de coordination matinal (le 'daily') entre les agents" IA.
@karpathy
Andrej Karpathy
10 days
I had the same thought so I've been playing with it in nanochat. E.g. here's 8 agents (4 claude, 4 codex), with 1 GPU each running nanochat experiments (trying to delete logit softcap without regression). The TLDR is that it doesn't work and it's a mess... but it's still very
0
0
0
@_DataStrategies
dataStrategies
10 days
Si je comprends bien : @AnthropicAI voulait pouvoir vérifier que ses modèles ne servent pas à la surveillance de masse ou à des armes autonomes ; @OpenAI s'en remet au @DeptofWar pour en juger, sans demander de visibilité sur ces usages.
@UnderSecretaryF
Senior Official Jeremy Lewin
10 days
For the avoidance of doubt, the OpenAI - @DeptofWar contract flows from the touchstone of “all lawful use” that DoW has rightfully insisted upon & xAI agreed to. But as Sam explained, it references certain existing legal authorities and includes certain mutually agreed upon
0
0
0
@karpathy
Andrej Karpathy
14 days
CLIs are super exciting precisely because they are a "legacy" technology, which means AI agents can natively and easily use them, combine them, interact with them via the entire terminal toolkit. E.g ask your Claude/Codex agent to install this new Polymarket CLI and ask for any
@SuhailKakar
Suhail Kakar
14 days
introducing polymarket cli - the fastest way for ai agents to access prediction markets built with rust. your agent can query markets, place trades, and pull data - all from the terminal fast, lightweight, no overhead
645
1K
12K
@_DataStrategies
dataStrategies
16 days
Data bugs dans l'info pour les voyageurs, épisode sept troumille. J'avais failli croire les applis qui disaient, aussi, 21 et 24' (enfin, 22 et 25' avant d'entrer) - donc repartir à pied vers le tramway. Mais cette suite de valeurs, 22 et 25, me semblait peu crédible !
0
0
0
@karpathy
Andrej Karpathy
17 days
Bought a new Mac mini to properly tinker with claws over the weekend. The apple store person told me they are selling like hotcakes and everyone is confused :) I'm definitely a bit sus'd to run OpenClaw specifically - giving my private data/keys to 400K lines of vibe coded
1K
1K
17K
@SouveraineTech
Souveraine Tech
20 days
💥#Meta souhaite lancer le scan automatique du visage de chaque personne qui passe devant ses fameuses lunettes. Elle envisage de le faire au moment opportun, c'est-à-dire : « dans un contexte politique dynamique où de nombreux groupes de la société civile susceptibles de nous
51
206
321
@fchollet
François Chollet
25 days
@Yossi_Dahan_ @polynoamial ARC-4 is in the works, to be released early 2027. ARC-5 is also planned. The final ARC will probably be 6-7. The point is to keep making benchmarks until it is no longer possible to propose something that humans can do and AI can't. AGI ~2030.
39
91
1K
@ant_oss
Ant Open Source
28 days
What's the secret? Our revolutionary Error-Correcting Editable (ECE) engine. 🧠 LLaDA2.1 doesn't just write; it drafts, then intelligently reviews and edits its own work, much like a human expert. This two-phase approach fixes a major weakness of early diffusion models —
1
1
26
@karpathy
Andrej Karpathy
1 month
@Gusarich I feel like I am actively getting dumber. LLMs get brain rot and it is measurable: https://t.co/gUxHiaRIzb "continual exposure to junk web text induces lasting cognitive decline in large language models (LLMs)" why shouldn't the same be true for brains.
61
36
502
@_DataStrategies
dataStrategies
1 month
Well curated (pre)training data is all you need (once again)
@eliebakouch
elie
1 month
new Alec Radford paper 😮
0
0
0
@emollick
Ethan Mollick
1 month
Canaries in the coal mine. Worth paying attention to. (And yes, they are both obviously interested in seeing their own products used, but hearing enough from other, independent coders that make me believe them. I wrote more about the shift here: https://t.co/ofbCp3fzG9)
36
95
1K