11-12-2024
I believe some significant principle:
Believe in scaling laws and model ability.
If LLMs could do it, don’t rely on any other algorithm or manual methods.
Anything could be implemented by an agent; should write the agent to do it.