The messy, secretive reality behind OpenAI’s bid to save the world

Interesting tweet sequence from Elon [1] below Karen Hao's introduction of the story on twitter:

> OpenAI should be more open imo

> All orgs developing advanced AI should be regulated, including Tesla

>> By individual governments or on a global scale, eg UN?

> Both

[1] https://twitter.com/elonmusk/status/1229544673590599681

OpenAI is clearly more secretive than the AI industry giants (internally & externally).

This doesn't diminish their work (e.g. GPT2 or CMA-ES for RL).

But this whole thing sounds like paranoia. I doubt if T1000 will suddenly emerge out of anybody's research.

I wrote a twitter thread on the key points on the article: https://twitter.com/stormtroper1721/status/12295782794320199...

What exactly is the messy part?