3 ms·
> Claude answers from its own extensive knowledge first for stable information. For time-sensitive topics or when users explicitly need current information, sea
by layer8 1y ago
> Claude answers from its own extensive knowledge first for stable information. For time-sensitive topics or when users explicitly need current information, search immediately.
It’s still curious that things like these needs prompting, instead of having an awareness mechanism from which this would be obvious to the LLM (given that the LLM knows its knowledge cutoff, in the above case).
- Nevermark 1y agoI could imagine that training and reinforcement with heavy searching would require a lot more computing time. And if a successful bias toward searching more can be added with just a prompt, that might be the most efficient way to implement that. Of course, I can imagine many things.
- layer8 1y agoIt might be more efficient for any particular case, but it’s adding special-casing to compensate for a general gap in the awareness capabilities of LLMs. And the latter is what I think needs to be solved for LLMs to become universally more reliable.