The Rise Of AI Coworkers

The Rise Of AI Coworkers

More

Summary

In this episode of Y Combinator’s Light Cone podcast, Lauren Tan and Roman Ugarte of the Grockbot team explain how they took an internal prototype to launch in just seven weeks. Grockbot lets users create AI teammates with their own names, identities, skills, tools, and even their own computers, and it is now built into the default DM experience on X.

The guests describe three design principles: work organized around bots rather than chats or files, bots that use the same tools humans do rather than relying only on MCPs, and bots that are always on and proactive. They share how teams use the product, from performance marketing bots that manage Google and Facebook ads with a budget and collaborate with a design team’s bot, to a full feature built end to end in a single Slack thread of about 200 messages, with a Cursor cloud agent producing screenshots and videos along the way.

The hosts and guests also discuss what it takes to work this way, including codebase guardrails, static analysis, and encoded review practices that free up engineers’ time, and why they think this style of feature development could grow dramatically in the coming year.


📺 Source: Y Combinator · Published November 14, 2026
🏷️ Format: Interview

1 Item

Channels

1 Item

People