---
session_id: "1373465"
title: "I love you so jevving much"
phase: live
last_checked: 2026-11-12T13:24:21-08:00
render: 20261112-132421-c4d6ee
---

# Links: I love you so jevving much

## Speaker

**Diogo Almeida** (@CompleteSkeptic), Co-founder and CEO, TypeSafe AI

- X: [@CompleteSkeptic](https://x.com/completeskeptic)

He co-authored InstructGPT at OpenAI (the paper below).

## Company

**[TypeSafe AI](https://typesafe.ai/)**: "TypeSafe AI is an AI lab building
machine-native intelligence infrastructure for automation, designed to make
decisions within software." (from the site)

## Resources

- **Blog** · [Introducing System One Models & Jev](https://typesafe.ai/blog/introducing-system-one-models-and-jev)
  (TypeSafe AI, 2026-09-15). The launch of Jev, the reliable typed-decision model
  the talk calls a foundation for smart software.
- **Paper** · [Training language models to follow instructions with human feedback](https://arxiv.org/abs/2203.02155)
  (arXiv, 2022-03-04, co-authored by Diogo Almeida). InstructGPT, the evidence
  for his point that the right task beats scale.
- **Blog** · [Partnering with TypeSafe: The Jevolution](https://sequoiacap.com/article/typesafe-the-jevolution)
  (Sequoia Capital, 2026-10-10). The talk's "evolution" slide is the Jevolution.
- **Blog** · [The Bitter Lesson](http://www.incompleteideas.net/IncIdeas/BitterLesson.html)
  (Rich Sutton). The essay the talk extends: data matters more than compute, and
  the "bitterest lesson" is doing the right task.

## Gaps

- LinkedIn (`linkedin.com/in/diogomda`, from the roster) answered 999 with no page
  title naming him, so it is not corroborated and is left out.
- The voice-interface video and the sandbox security incident mentioned in the
  talk were not identified.
- No official recording or slides found yet. The dossier rechecks every 6 hours
  during the event.