I really don't get it, maybe I tuned it badly but got a similar kind of Pulse today too – a few vaguely relevant but not very interesting responses from simple prompts from what I can tell just vanilla GPT-5. Their benchmark should be LLM generated feed that would compete with… https://t.co/PabJaMaomL

— Peter Gostev (@petergostev) 27 septembre 2025

I really don't get it, maybe I tuned it badly but got a similar kind of Pulse today too – a few vaguely relevant but not very interesting responses from simple prompts from what I can tell just vanilla GPT-5. Their benchmark should be LLM generated feed that would compete with

→ View original post on X — @petergostev,