OpenRouter Asks What AI Learns After the Human Leaves the Loop

The model can keep training after the human leaves the room. The harder question is what remains available for it to learn.

OpenRouter hosts Alex Atallah, co-founder and CEO of OpenRouter, and Garrett Lord, co-founder and CEO of Handshake, for Can AI Learn Without Humans? on December 3 in New York. The conversation reaches into reinforcement learning, self-improving systems, human expertise and the future of work.

Why this room matters

The internet gave models a record of what people had already written. The next generation of systems is being asked to learn from actions, environments and feedback that may be generated partly by other machines. That changes where expertise enters the system and what happens when the easiest signal to optimize is not the outcome people intended.

OpenRouter sits close to the routing layer between models and applications. Handshake sits close to the human talent and expertise those systems may increasingly consume. Putting those perspectives together makes the question less philosophical. It becomes a discussion about the infrastructure, incentives and labor behind the next learning loop.

The event is free. The exact New York location is provided through the official registration process.

Event details

• Date: December 3, 2026
• Time: 5:30-8:30 p.m. ET
• Location: New York City; exact address after registration
• Host: OpenRouter
• Speakers: Alex Atallah, OpenRouter; Garrett Lord, Handshake
• Access: Free registration
• Register: Official registration page