Independent AI lab · Paris
X AssistAI is an independent AI lab studying how autonomous agents coordinate, communicate, and earn trust — and turning each finding into products people actually use. You supervise. AI executes.
Meta FAIR's tri-modal brain encoder won Algonauts 2025 by aligning pretrained model features to neural activity. Notes on why a predictive model of the brain matters for agent design.
Sakana AI's pipeline ideates, codes, runs experiments, and writes the paper — then reviews itself. Notes on what an autonomous research loop says about supervision.
A self-referential framework where the agent monkey-patches its own logic toward a high-level goal. Notes on the design space it opens — and the supervision problem it sharpens.
A workspace of specialized agents — lawyer, accountant, designer, doctor — each tuned to its profession's real workflows.
An AI assistant director for film and video: shot lists, coverage, and cut decisions — a production assistant that knows the director's job.
We’re X AssistAI, an independent AI lab based in Paris. We’ve spent the last few years building agents that do real work — not demos. Everything here follows one thesis: the next decade of software is humans supervising AI agents, and the interfaces, protocols, and trust models all have to be rebuilt around that.
We ship in public, one product at a time, and publish what we learn along the way.