Some people argue that a sufficiently accurate world model would make it possible to plan without reinforcement learning.

“Uh, I always hear like the current saying, well, we don't need R_L_ we just need world models, right? That's like the the similar the the uh yeah, but simply just saying if we just describe the world perfectly, then we could just use that, and we don't need to do R_L_ anymore, we can just do planning in that world model, and we we are fine. So uh we don't we shouldn't spend all our time doing R_L_”