# Quentin Feuillade-Montixi: Why AI Needs Its Own Language | Intelligence at the Frontier 2026

- Channel: [Funding the Commons](https://streameth.org/funding_the_commons)
- Date: 2026-05-09
- Duration: 12:01
- Topics: Quentin Feuillade-Montixi, WeaveMind, AI safety evaluation, AI systems design, AI abstraction, behavior elicitation, AI alignment 2026, AI safety 2026, AI agent frameworks, AI interpretability, AI development tools, AI safety research, Intelligence at the Frontier, Funding the Commons, tech for good, AI infrastructure, agentic AI 2026, AI governance 2026, AI tooling
- Watch: https://streameth.org/watch/yt-a4FMgj7zvQ0
- YouTube: https://www.youtube.com/watch?v=a4FMgj7zvQ0

## Description

Building AI systems today means choosing between two bad options: the rigidity of traditional code or the unpredictability of language models. Quentin Feuillade-Montixi thinks neither is good enough, and that we need a new abstraction layer between them.

Quentin, founder of WeaveMind and co-founder of Prism Evals, built the behavior elicitation benchmark for AI safety evaluation. His frustration with how AI systems are currently built led him to develop a framework where everything operates at the same level of abstraction — giving developers the structure and verifiability of code while preserving the flexibility needed to work with language models. The goal is to make what AI agents are actually doing visible and controllable, not just at the output level but at the level of the system itself.

Speaker: Quentin Feuillade-Montixi (WeaveMind)

This session was part of Intelligence at the Frontier, held March 14–15, 2026, at Frontier Tower in San Francisco.

Learn more at https://www.fundingthecommons.io