It’s all fun and games until AI hallucinates a tumor in your brain.
@tunguz
-
Trusting vibes of trusted people over model propaganda
By
–
The only thing that I trust these days when it comes to *really* trusting how good models are is a sample of vibes/opinions of a handful of trusted people on this site. Everything else is pure propaganda.
-
Frustration with explaining ML for tabular data
By
–
Every time someone tries to explain to me ML for tabular data. @trainxgb @tabul_ai
-
100+ prompts for agents with Codex /goal checking on Fable launch
By
–
I've created a list of 100+ prompts for the equal number of different agents and I have Codex with /goal checking on Fable launching every 5 minutes so that I can unleash them when it happens.
-
Future brief access to top AI models for everyone
By
–
In the future everyone will have access to the SOTA models for five minutes.
-
User notices Mistral’s La Chaton Fat seems weaker lately
By
–
Is it just me or do others feel like Mistral’s La Chaton Fat has been getting weaker lately?
-
Underhyped public models are the future of AI
By
–
The future of AI belongs to underhyped and overcapable public models.
-
Deep research better with ChatGPT according to tunguz
By
–
Deep research is still better with ChatGPT IMHO.