RSI is about turning that extra computation into acceleration - that is applying the compute to the (useful_work / compute) factor - this is the recursive part.
It says much more than that. It says that all custom architectures and approaches are become irrelevant if compute increases, which was proven again by the large latent models recently solving RL tasks better than any classical RL technique.
I don't get it. TBL is an observation, not a law or guarantee. It says (Compute / $) is increasing. In efficiency terms:
useful_work / $ = (useful_work / compute) × (compute / $)
But the left side only improves if the method can turn that extra computation into useful output.
RSI is about turning that extra computation into acceleration - that is applying the compute to the (useful_work / compute) factor - this is the recursive part.
plausible, but unclear
It says much more than that. It says that all custom architectures and approaches are become irrelevant if compute increases, which was proven again by the large latent models recently solving RL tasks better than any classical RL technique.
RSI is also known under more mundane words such as "Learning".