Wednesday, 17 September, 2025
Silicon Valley Ramps Up Investment in RL Environments to Train Smarter AI Agents
Startups and labs in Silicon Valley are heavily investing in reinforcement learning (RL) environments—simulated “workspaces” where AI agents perform multi-step tasks—to make agents more capable and robust. Big AI companies are building these in‐house, while firms like Surge, Mercor, Mechanize, and Prime Intellect are emerging to supply high-quality environments. These platforms mimic real software settings to better evaluate agents.
Read full story at TechCrunch