[D] Connor Leahy on EleutherAI, Replicating GPT-2/GPT-3, AI Risk and Alignment

Hey, just want to share the new Gradient interview with Connor Leahy. It has a lot of fun stories about getting into the field of AI from as a sort of hacker and how EleutherAI got going and still works as a group passion project of a bunch of hackers, so it's a fun listen!

Things discussed:

• Replicating GPT2–1.5B , GPT2, Counting Consciousness and the Curious Hacker
• The Hacker Learns to Trust
• The Pile
• GPT-Neo
• GPT-J
• Why Release …

ai eleutherai gpt gpt-3 machinelearning risk