AI and Existential Risk (Robert Wright & Connor Leahy)
Robert Wright's Nonzero
00:00
Why I Created an Open Source LLM
I don't think a model of like, you know, six or 20 billion parameters is super dangerous. This was a bet. And I would say this was a reasonable bet to make at the time. In retrospect, maybe not. As he said, I did expect things to go as fast. I didn't expect to be able to care. You could argue whether it even had much of an impact in retrospect. Like some, interpretability researchers used my work and spoke negatively about me. It's like hugging face again. Of course they're going to come back for more money. But and see them all money, you have to give them credit. He's at least saying to Congress
Transcript
Play full episode
