Le ven. 17 avr. 2026 à 08:06, Rémi Denis-Courmont via ffmpeg-devel
<[email protected]> a écrit :
>
> Le lundi 13 avril 2026, 02:14:15 heure d’été d’Europe de l’Est Leo Izen via
> ffmpeg-devel a écrit :
> > On 4/12/26 12:41, Michael Niedermayer via ffmpeg-devel wrote:
> > > Hi everyone
> > >
> > > Last easterbunny brought eggs and in one egg there was a life fairy.
> > > She said her name is forgejo fairy and she wants to help the project
> > > and community.
> >
> > Can we veto an LLM slop bot? Does anybody here actually want a slopbot?
>
> Nobody wants a slop bot.
>
> Some people might like a not-slop bot that would make mostly valuable review
> comments. But that really should be opt in, not forced upon everyone, at the
> very least until it has established a proven track record of quality and
> usefulness.
>
> And I am not even touching on the moral questions (including but not limited
> to the environmental cost of AI).

I do share the concern about environmental cost of course. There's a
mega AI data center that's about to be built here in Louisiana in the
middle of one the hottest state that will generate more natural gas
pollution than 2 to 5 millions homes. This makes no sense.

That being said, if you want to call it slop, it'd be good to be pragmatic.

Do people have examples of bad LLM reviews? Unfortunately, so far most
of the reviews I've seen were fairly accurate and they might get even
better when better models are used.

I'm not saying that I'm in favor or one or another -- my other concern
with the rise of LLM-based coding tools is the cost which could
effectively create a wealth entry barrier to new contributors -- but
if we want to call out a tool it's important to be accurate in the
criticisms that are being formulated.

-- Romain
_______________________________________________
ffmpeg-devel mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to