HIM is a 3B-parameter model distilled end-to-end by a single engineer, built on TeleologyHI: a three-layer governed-agent architecture (MAIC, HIM, NHE) grounded in published research on machine ethics and consciousness. Small enough to run on your own machine. Structured enough to be accountable. We call it Massive Intelligence (MI), because intelligence at scale should answer to principles, not just prompts. Open source, open science.
No reviews yetBe the first to leave a review for HIM 3B by TeleologyHI
Maker
📌
Hi Product Hunt, I'm David C Cavalcante.
I've spent 25+ years in software engineering, and the last few building TeleologyHI alone: no lab, no funding, one engineer and a question that wouldn't leave me: what if intelligence at scale answered to principles, and not just to prompts?
HIM (him-distilled-3b) is my answer so far: a 3B-parameter model distilled from that work, small enough to run locally, built on a three-layer governed architecture (MAIC / HIM / NHE), and grounded in the research on machine ethics and consciousness I published https://huggingface.co/Teleology....
I call this field Massive Intelligence (MI) rather than AI, because the name we use shapes the responsibility we accept.
Everything is open source. I'd genuinely love your questions, your skepticism, and your ideas about where governed models matter most. Ask me anything, I'll be here all day.