The First Line of Your Job: Default to AI

By Sequoia Capital

Categories: VC, Startup

Summary

OpenDoor made AI adoption stick by making "default to AI" the first line in every job description and performance management evaluation. A non-technical HPM automated his entire role using Cursor and GumLoop, proving that AI-native transformation requires cultural mandate, not just tooling.

Key Takeaways

  1. Make AI adoption a performance metric: Evaluate employees first on whether they "default to AI," embedding it into your performance management system as the primary evaluation criterion.
  2. Run companywide hackathons with vendor support: Bring in Cursor, GumLoop, and OpenAI to coach teams hands-on. This tested whether the muscle existed before mandating the behavior.
  3. Non-technical roles can automate themselves: An HPM with zero coding experience automated his entire job using AI tools, then transitioned to managing software systems—proof that AI adoption isn't gatekept by technical skill.
  4. Be explicit about performance system purpose: Reframe performance management as a team selection mechanism ("professional sports team"), making it clear that AI adoption is non-negotiable for staying on the roster.
  5. Translate hackathon wins into company policy: Use hackathon results to validate AI adoption capability exists, then codify the behavior into job descriptions and org-wide expectations.

Topics

Transcript Excerpt

How do you take what was likely a very nonAI company and make it AI native? To >> calculate the result of what happened, we had a companywide hackathon. We have people we call HPMs. These are people who essentially renovate homes. They're people who manage renovation funds. One of them wrote a piece of software. There's a guy who was like essentially a GC wrote a piece of software that automated away his entire job and he's now a manager of bunch of pieces of software to do that. This guy had ne...