dataStrategies

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

[dataStrategies @_DataStrategies]

@StJu25641 @FrancoisGeerolf @Niji_Digital Il y a PLEIN de clignotants, à condition de les brancher (c’est un peu de travail) et ensuite de les regarder (c’est beaucoup de travail).

[dataStrategies @_DataStrategies]

@StJu25641 @FrancoisGeerolf @Niji_Digital Dans le second cas, elles ne sont pas intéressées pour payer ;-)
Elles jouent sur les différences de délai : « prends l’oseille et tire-toi ».

[dataStrategies @_DataStrategies]

@StJu25641 @FrancoisGeerolf @Niji_Digital La fraude passe par « des entreprises réelles avec Rib faux », des entreprises réelles créées à l’effet de frauder et vite dissoutes, etc. etc.

[Fabien @Fabien_Mikol]

Jean-Baptiste Kempf (VideoLAN, VLC, Kyber): Claude dépasse tout en code, on utilise presque que ça même si c'est cher en token (10% de ma masse salariale) ; je pense comme @mikiane que les prix baisseront contrairement à ce qu'on raconte, notamment grâce à la concurrence chinoise

[dataStrategies @_DataStrategies]

[AI slop]
La personne qui n'a pas fait l'effort d'écrire elle-même,
pourquoi voudrait-elle que je fasse l'effort de lire moi-même ?

[Zara Zhang @zarazhangrui]

If you've adopted AI at your company but haven't seen any tangible results, read this 1990 article: "The Dynamo and the Computer" by Paul David.

When electricity first arrived, factories that "adopted" it barely got faster. They just swapped the steam engine for an electric one and ran everything else exactly as before: same machine layout, same workflow, same management. Electricity in, no real gains out.

The most common mistake with any new technology is to drop it into the old organization and then declare the transformation done.

The real leap came decades later, when each machine got its own small motor. Suddenly machines no longer had to be lined up around one central drive shaft. They could be rearranged around the actual flow of work.

The productivity gains didn't come from electricity. They came from REDESIGNING THE ENTIRE FACTORY around it.

AI is the same. Bolting it onto your existing process gets you a faster steam engine. The payoff comes when you redesign the work itself.

(link to paper in comments)

[Taelin @VictorTaelin]

this is my personal singularity moment

this post may sound like a paid ad. I only wish. I'm concerned, more so than happy. the world is changing, and, among the scenarios where AI goes terribly wrong, inequality is the most realistic, yet, the one Anthropic seems to be the least concerned about. I'm glad OpenAI is taking the opposite stance: *personal AGI for everyone*. I think this is a commendable position in the times we live. but who am I in the queue of the bread?

anyway, Fable is here, so I'll just report my first-hour experience

first of all, all my pet prompts are solved.
→ λ-calculus puzzles
→ bug questions
→ one-shot apps
all are trivial to it.

I don't have anything harder other than my
ongoing work

so, in the last several days, I've been toying with HVM5, a new interaction net evaluator with a faster loop.

after writing the first version, I left 32 GPT-5 agents working for ~20 hours each. this resulted in up to 2x speedups, but the file size increased by 2-fold and quality decreased significantly.

I then simplified the whole thing into an even simpler core, and left Opus 4.8 and GPT 5.5 optimizing it for 8 hours. Opus got a legit 6% - 34% speedup in most benches. GPT got better results, but, sadly, an unusable file.

I then asked Fable to optimize it.

2 hours later, it landed a 1770% speedup in one case, 100%+ in other 4, and 22% in average. yes, in 2 hours it outperformed me, opus 4.8 and a swarm of gpt 5.5 agents, by one order of magnitude.

that could not possibly be legit. "it must be hardcoding the benchmarks" (GPT trauma). so I read its explanation and what it did was, indeed, the most high impact optimization one could try first. seems like HVM5 was wasting a lot of time garbage-collecting unused branches of pattern-match nodes. I had optimized that for static mats, but not for dynamic mats. skill issue. Fable figured how to do it for these, resulting in a massive speedup in some benches

but wait, is that *correct*? I'm not sure yet, it is credible, but this is the kind of thing that is very easy to get wrong on interaction nets. the problem is, when I was ready to start auditing Fable's solution so I could tell whether it was buggy or legit, it interrupted me to tell me it had found a massive bug on the code *I* had written.

... wait, what?

so... for garbage collection purposes, I stored a bit on lambda term pointers that meant "the variable bound by this lambda has been freed, so, its lambda must free whatever argument it is applied to". that's fine. yet, on duplicator nodes, I also used the same bit to mean "one of the duplicated variables was freed, so, treat this dup as a passthrough no-op". so, if a lambda entered a duplicator, it would mistake the lambda's collection bit for its own, resulting in corrupted interaction!

that's a mouthful, why I'm writing this?

just so you can appreciate the sheer absurdity of what just happened. I didn't ask it to find bugs. I asked it for an optimization. and even if I did ask it to find bugs, this bug is so astonishingly subtle and specific, identifying it takes mastering the domain to an extent that it beyond even me. I'd easily need hours or days to fix it, *if* I ever came across it. chances are it would just go unnoticed. and Fable found it and fixed it like it was nothing, while it was busy adding a 17x speedup to a file that neither I, nor Opus 4.8, nor a fleet of GPT 5.5 managed to barely make 2x faster.

oh and there is also another tab where it is also ripping through Bend's codebase and finishing everything I had to do

I don't know what to say anymore

this isn't about Anthropic or OpenAI, this is about our collective future as a species. the world is changing, and we need to be aware of it, and discuss how to handle this change.

receipt below . . .

[dataStrategies @_DataStrategies]

P.S. L'abstract ci-dessus contient une erreur et une phrase vide de sens. Le travail avec l'IA commence quand elle a produit son résultat (enfin, son premier résultat).

Ici ça doit être mieux :

[dataStrategies @_DataStrategies]

Eh bien j'aurai #vibemathé avant de vibecoder.

Je sais : ce 8 pages en style arXiv vaut à peu près autant qu'une chanson Suno. C'est de l'AI-slop. Mais *pour un non-matheux* comme moi, ça fait la blague. Et la qualité du "harnais mathématique" de Cowork (avec Opus 4.8) est 🎩.

[dataStrategies @_DataStrategies]

C'est un truc que j'avais tenté d'étudier au lycée, fin 1980/début 81 :

que se passe-t-il si on lit
l'écriture classique d'une #permutation,
par exemple "(1 2 3) -> (3 1 2)",
comme l'écriture cyclique d'une autre permutation ?
(3 -> 1, 1 -> 2, 2 -> 3)
…et qu'on recommence ?

[dataStrategies @_DataStrategies]

@karpathy Thank you!
But, the scorecard looks like misleading re Fable 5, the one we ordinary humans can use. The sentence in very small fonts says that these are not Fable 5's scores - they are just not written.

[Ethan Mollick @emollick]

The Matrix idea of keeping humans as batteries is obviously weird... we would be more useful as dice.

LLMs default to very similar kinds of arguments & structure, and even different LLMs seem to collapse to similar concepts. Humans provide a lot more variation in their own work.

[dataStrategies @_DataStrategies]

@Xudong07452910 Sur un article scientifique en rédaction avec Claude Cowork, l'état de la recherche en cours .md de Claude est plus utile que l'article rédigé en Latex : plus utile et lisible par les agents… *et* plus utile et lisible par les humains.

[dataStrategies @_DataStrategies]

@FrancoisGeerolf @Niji_Digital Confirmation : le chiffrage de la Cour des comptes ne porte que sur *une* des configurations de fraude, sans doute la plus simple. ("Le rapport ne s’appuie toutefois que sur l’étude du taux d’encadrement des apprentis.") https://t.co/95UfYGngFx

[dataStrategies @_DataStrategies]

@FrancoisGeerolf @Niji_Digital P.S. : j’ouvre l’article après avoir répondu du tac au tac, oups. Et je ris jaune : à mon avis (personnel et professionnel, mais sans avoir de fichiers de données), c’est entre 25 et 40% des dépenses qui sont détournées ; pas « jusqu’à 10% ».

[dataStrategies @_DataStrategies]

@_fracapuano Looks like a reference to Leonard Cohen's comment on the great mysteries unraveled, after "the Tower of Song", here in London at 5:45 https://t.co/OTiXTD5ejF

[dataStrategies @_DataStrategies]

@Dan_Jeffries1 @Sams_Antics (Hi — I don't know any of you, nor the bet, but)

this "shrimp" reminds me of the "shrimp/turtle case": the WTO ruling (in 1998) provides a legal means of fighting greenhouse gas emissions at international level. (That's for welfare!). But no country has dared to use it 🥵