AI Socratic Sep 2025 Part 3 — Frontier Tower Edition

This blog post has the goal to help us structure the Socratic dialogues at the AI Dinner 13.0, that Ivo and I are organizing in SF at the Frontier Tower.

We'll use blog post 1 and part 2 to go through Since part 2 was written during the last AI dinner 13.2 and a lot of new updates happened since, we'll cover them here.

#Why Language Model Hallucinate and Cannot Say “I don’t Know”!

Language models hallucinate because their training and evaluation reward guessing over admitting uncertainty. Models are unable to say “I don’t Know” because they focus on accuracy. Guessing can improve scores but leads to more confident errors (hallucinations).

Some facts are unpredictable or unavailable, making errors inevitable. The solution might be roughly penalizing errors more than uncertainty and rewarding honest abstention.

https://openai.com/index/why-language-models-hallucinate

#How people are using ChatGPT & Claude

OpenAI just released how people are using ChatpGPT. Coding 4.2% We live in a bubble

https://openai.com/index/how-people-are-using-chatgpt, https://x.com/atelicinvest/status/1967941812078915614

Claude also released how people are using their API and Claude.ai. The 1P API is used at 97% for automation.

#Google AP2 + Coinbase x402, A2A stablecoin payments

AP2 (Agentic Payment Protocol) and x402 just unlocked a new level for AI agents. Agent can now actually pay other agents and MCPs.
https://x.com/sundarpichai/status/1968013016181641492

#Oracle Up 40% Because It’s building Stargate

https://x.com/mattturck/status/1966545601140531707

---