Re: Project-wide LLM budget for helping people
On 1/11/25 5:25 PM, M. Zhou wrote:
> Opinion against this post will include something about hallucination.
> In the case LLM write something that does not compile at all, or write
> some non-existent API, a human is intelligent enough to easily notice
> that build failure or lintian error and tell whether it is hallucination
> or not. I personally believe LLMs, at the current stage, is useful
> as long as used and interpreted properly.
LLMs ingest documentation. If we stop writing or overhauling
documentation, what is the LLM going to suggest? Even hallucination
aside, how can it be the least bit accurate if it is not fed up-to-date
information about how things work?
> BTW, I was in the middle of evaluation LLMs for the nm-template. I did lots
> of procrastinations towards finishing the evaluation, but the first
> several questions were answered perfectly.
> https://salsa.debian.org/lumin/ai-noises/-/tree/main/nm-templates?ref_type=heads
> If anybody is interested in seeing the LLM evaluation against nm-templates,
> please let me know and your message will be significantly useful for me
> to conquer my procrastination on it.
If anyone is attempting to answer these using LLMs, I'd expect them to
be excluded from the process. It's one thing to generate documentation
by reviewing LLM output for accuracy and potentially publishing that to
be helpful. It's another thing to try to lie yourself through the
process in order to gain the project's trust.
(In job interviews candidates already regularly use LLMs in the
background to answer the questions. There I think it's still noticable
when people claim knowledge that they do not have. In offline
communication all bets are off.)
Kind regards
Philipp Kern
Reply to: