FennoAI(@Fennoai):The interesting part isn’t Jev @typesafeai itself. It’s the split: LLMs for generation and reasoning, System One for classification, routing, and risk gates. Agents are expensive because we keep handing System 1 work to System 2. If Auto Mode becomes a default layer in open harnesses, that will matter more than wiring up another chat model.

FennoAI

@Fennoai

OpenAI-compatible GPT API access for builders. Simple pricing, fast setup. Discord support:

参加 June 2026

278 フォロー中 121 ファン

FennoAI@Fennoai

2026.09.18 07:06

The interesting part isn’t Jev @typesafeai itself. It’s the split: LLMs for generation and reasoning, System One for classification, routing, and risk gates. Agents are expensive because we keep handing System 1 work to System 2. If Auto Mode becomes a default layer in open harnesses, that will matter more than wiring up another chat model.