Obayda Julien

Exploring structures, new mathematical languages, foundational ideas.

[Francesco Maggi @Frances01896069]

The moment you are on Anthropic’s or OpenAI’s payroll, you stop being a scientist or a scholar, and you become an employee, a tool in the research, development and marketing strategies of a company with billions of debt and a desperate need to become profitable (or, more realistically, to be saved by the government— we all pay for the toy). The conflict of interest is just humungous.

“At anthropic some philosophers” is just not a thing.

[Obayda Julien @TerreMath]

@evanytgrowth Comment cet escroc sévit-il encore ?

[Obayda Julien @TerreMath]

I’d go even further: every author seems focused on their own personal research agenda and on putting out their own papers, rather than on genuine collaboration and discussion. It has become a systemic problem, not only in mathematics but across society as a whole. AI is only accelerating it further, given the possibility for a single researcher to produce dozens of papers a year. Mathematics needs to change, but no one seems willing to even try!

[Emir @EmirKarsiyakali]

Bugünün internetine çıkamayan (TLS 1.0), 2007 yapımı, 8 MB hafızalı Nokia'mı Claude ile tanıştırdım. Yani artık yapay zekâsı var 🙃 2-3 gündür firmware'ine tersine mühendislik yapıyorum; amacım Java katmanını aşıp doğrudan sistemin içine girmek. "Gelecek" olduğunu düşündüğüm acayip fikirlerim var. Videoyu önden salıyorum.

[Obayda Julien @TerreMath]

For me, the real problem is that internal models seem to be more capable than the publicly released ones. My theory is that Anthropic actually has one very powerful model that they’re splitting into smaller pieces so they can maintain a publication/release roadmap over several months, at least.
And in the end, it seems that the only real breakthroughs have been achieved using internal models. The mathematicians associating themselves with this are frauds...

[Alexis Poulin @Poulin2012]

Gaza.

Il y an encore des gens qui pensent que défendre l’indéfendable suffira.

Non. C’est fini.

[Obayda Julien @TerreMath]

@Carlein90 "our" but found by gpt and internal model of google ?

[Obayda Julien @TerreMath]

@hichammouajjah C'est probablement cela qui fait peur aux entreprises IA...

[Obayda Julien @TerreMath]

@GabLattanzio Si ton métier de prof de réduit à ce que tu dis va falloir se poser des questions...

[Obayda Julien @TerreMath]

@mben_dev Se venger sur les gens parce une entreprise a vole son boulot... Pourquoi pas se venger sur l'IA ?

[Obayda Julien @TerreMath]

@RayaneRachid_ Tout dependra des usages, mais bon peut on faire pire que Astra?

[Obayda Julien @TerreMath]

@littmath @nihilsuncle Ok but at the same time what do say about the fact that IA are severely use mathematicians works for free ?

[Obayda Julien @TerreMath]

Indeed, the issue is whether such a publication has any value at all, even without taking full credit for it (although the ideas suggested to the AI are still ours). My view is that only truly groundbreaking results should be published. And of course, I keep my deeper ideas to myself: I don’t want to give them to AI systems, which ultimately belong to companies.

[Obayda Julien @TerreMath]

@rperezmarco @AnthropicAI At the sameb time none of the published Opus neither Fable are very good in mathematics so its very strange

[Obayda Julien @TerreMath]

@RimaHas Qui veut enseigner encore en France ?

[Obayda Julien @TerreMath]

Pas faux, mais, connaissant relativement bien le milieu de l’enseignement, je me demande : quels profs ont envie d’aller dans ces établissements ? Il faut nécessairement une refonte totale de l’éducation et de l’enseignement. Or, Rima Hassan fait historiquement partie de ceux qui ont permis à l’école de devenir ce monstre.

[Obayda Julien @TerreMath]

@RayaneRachid_ Pas chez moi c'était dans 4jours

[Obayda Julien @TerreMath]

@Deniz_Genlik Same for me Astra seems to compute a lot but nothing super nice at the end