Expanding on this, it's worth noting that LLMs and generative models working as well as they did absolutely caught OpenAI by surprise. Before (google's) Lamda's success they were plugging away at deep reinforcement learning. They are entirely freestyling with no idea what happens next.
Reposted from
Ed Zitron
I want to dispel a very commonly-held assumption: none of these companies had a plan with generative AI. They threw cash at it because it looked like it might be profitable, and have thrown more cash as it became less profitable because none of them have any other ideas, or plan of any kind.
Comments