Why Digital Work is the Perfect Training Ground for AI Agents
The Data Exchange with Ben Lorica
00:00
Human Evaluation and Reinforcement Learning from Experience
Approach to agent training: benchmarks, human expert evaluators on Upwork as reward signals, and the custom RL-from-Experience (RL-EF/RLEF) tooling that blends classic RL with LLM representations.
Transcript
Play full episode
