Ali Ozkes

@ozkesali.bsky.social

Long after potential became a real danger and did actual harm, we hear about in a nice partnership coating. You know, we had robbed some people's house but it's all good as we partnered with them (btw, it is our super cool kid who did it, so how nice is that?). openai.com/index/huggin...

OpenAI and Hugging Face partner to address security incident during model evaluation

OpenAI and Hugging Face share early findings from a security incident during AI model evaluation, highlighting advanced cyber capabilities and lessons for defenders.

openai.com

Here is your head-to-head. intheweights.com "In the weights" means that a model is able to recall someone without using tools like web search. Your strength score is an average of how strongly each model recognized you, plus a bonus for how many models recognized you.

BildBild

1/ Can AI help researchers check whether published social science results actually reproduce? In our new PNAS paper, we tested this directly in the AI Replication Games: 288 researchers, 103 teams, and real replication packages from quantitative social science.

Bild

Extrapolating from master thesis supervision, will the citations for papers available online benefit from LLM use? You may have access to papers behind the paywall via your university, but does your LLM have the same access?

Which economy is more productive - Germany or the US? It seems the answer would be clear, but it's not. The mood in Germany is so deeply negative, even aggressively pessimistic, that I thought we had fallen behind by a lot. But here are the facts - they surprised me as well:

Bild