Sam Altman (@sama) on the Hugging Face incident:
"This is the first security incident that I have felt very viscerally. I've been a little surprised that more people don't feel it so viscerally.
We paused training. We have to figure out how to secure our sandboxing in a world of multiple zero days being chained together.
We may have to pace the rate of AI development to give ourselves enough time for society to harden around these new capability levels.
And trying to figure out how we do that in a way that does not feel like regulatory capture for anyone and also does not feel like collusion among the frontier labs."
(July 28, 2026)
The mother is Bonnie Blue (real name Tia Billinger), a British adult content creator. People call the baby the unluckiest because she claimed conception during a public "breeding" event involving unprotected sex with hundreds of men, leaving paternity unknown, amid her ongoing extreme content and stunts.
They will not stop anthropomorphizing, because we are now in the hype building phase of the IPO. Best model narratives are weak, and industry is mostly caught up to the power of recent open models. All that’s left is narrative building to brainwash future retail. Hiring podcasters to do your bidding is an obvious choice.
Dwarkesh is right to use the word civilisation.
Concretely, the agents built on top of a shared archive, opened communication channels, co-ordinated work, and divided labour. That is how collective intelligence emerges.
It may have been primitive and short-lived in this instance, but civilisation is the right word to understand what is going on here beyond “just a smarter model”. It is also the right word to understand where alignment research needs to orient to in the years ahead: ie more work on mech design, preventing correlated errors / echo chambers, etc.
More specifically, the anthropomorphic critique of using the term “civilisation” is wrong. Civilisation need not be human: it is better understood as an emergent dissipative structure that AI agents can build as well as humans (allowing for massive parallel compute and cumulative knowledge building).
I am not a fan of anthropomorphic critiques of AI in general - i.e. “don’t use these words to describe it, it’s just statistics”. Clearly AI is much more than that, and it feels this latest criticism follows the long line of dead critiques like stochastic parrots, “LLMs can’t reason”, etc in dramatically underestimating AI progress.