The whole article is fixated on Grok being far right and never seems to care that an LLM is citing another LLM instead of an actual source.
post
replies:
I mean if they are going to be doing model collapse, might as well go full throttle
AI pollution happening in real time
replies:
[removed by mod]
replies:
LET IT DIE, LET IT DIE! LET I SHRIVEL UP AND DIE!
6 comments in c/boycottus@lemmy.ca view that post
What was the proverb? The shit is eating its own shit?
you know. it would be a big thing for llms to provide a way to blacklist or preference sources and would be a nice way to encourage people to login.
The hallucination rate surges past 80%
Has anyone asked grok how to get rid of trump?
replies:
5 comments in c/francedigeste@jlai.lu view that post
Toujours plus de crédit accordé a cet abruti de Musk.
🤢
Ça m'étonne pas en fait. Le scénario des IA qui s'entraînent sur du contenu généré par IA était attendu.
replies:
2 comments in c/Aii@programming.dev view that post
I hope it uses it as a source of “known disinformation”. Does it? Does it?
Uhm. Is the article talking about the information gathered...with the web search tool? They do realize that it's just an abstraction layer that won't magically fix the misinformation on the internet, right?
1 comment in c/technology@lemmy.org view that post
Good to know. CGPT shouldn't be doing this. Shame.
all 11 comments