Ex-Slack Data Lead Jokes About Founding an RL Company to Fix AI Data Pipelines

josh_wills · x · 2026-08-01

Josh Wills, former Head of Data Science at Slack, jokingly tweeted about wanting to start an RL (Reinforcement Learning) environments company where it's just him teaching AI models how to write data pipelines that aren't complete garbage.

This humorous take highlights a real pain point: current LLMs still struggle with code quality when handling complex data engineering tasks, suggesting a need for specialized RL environments to optimize them.

Original post →

More from AGI Musings

AGI Musings channel →