Daisy Lin

LLMs are getting pretty good at talking. Getting them to reliably act on a computer — clicking, typing, and navigating real websites to achieve a goal — is a different beast. At Amazon’s AGI Lab, one of our primary research efforts is to massively scale reinforcement learning (RL) into a practical engine for training computer-use agents (CUAs). Through this work, one lesson has become abundantly clear: useful agents do not emerge from a better model alone.

The day began like any other—a fresh cup of coffee in my company mug, settling into my home office. Ping! A Slack message from my boss. “Can we meet early?” Of course, I replied. My mind ran through the usual list of tasks. A press release? An urgent LinkedIn post responding to new regulations? Such drills had become commonplace in our financial technology world, still reeling from Silicon Valley Bank’s collapse. I joined the Zoom call, unprepared for how swiftly life can pivot.