For the past few years, humans have been doing “prompt engineering” to coax the best performance out of different LLMs. In this work, we exp…

For the past few years, humans have been doing “prompt engineering” to coax the best performance out of different LLMs. In this work, we explored what happens if we train an AI to do that job instead. By training a Conductor model with RL, we found that it naturally learns to write highly effective, custom instructions for a whole pool of other models. It essentially learns to ‘manage’ them in natural language. What surprised me most was how it dynamically adapts. For simple factual questions, it just queries one model. But for hard coding problems, it autonomously spins up a whole pipeline of planners, coders, and verifiers. Really excited to see where this paradigm of “AI managing AI” goes next, especially as we start moving from single-agent chain-of-thought to multi-agent “chain-of-command”. Link to our #ICLR2026 paper: https://arxiv.org/abs/2512.04388 Along with our TRINITY paper which we announced earlier, this work also powers our new multi-agent system: Sakana Fugu (https://sakana.ai/fugu-beta) 🐡 Introducing our new work: “Learning to Orchestrate Agents in Natural Language with the Conductor” accepted at #ICLR2026 https://arxiv.org/abs/2512.04388 What if we trained an AI not to solve problems directly, but to act as a manager that delegates tasks to a diverse team of othe…

Source: David Ha (X) | 2026-04-27

Related