Machine Learning Engineering Is the Ultimate RL Task | Jennifer Zhao

Preference Model bets that RL environments teaching AI to do machine learning research—not just general capabilities—are the key lever for safe AI progress.

• $16M seed round: Preference Model, led by CEO Jennifer Zhao, raised $16M led by Andreessen Horowitz and is open-sourcing its RL environment "Karrot" (referred to on air as "carrots").
• Core thesis: Training models via RL environments to perform machine learning engineering/research is the highest-leverage way to shape how AI develops, given short capability timelines.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.