How to run your product team like a research lab | Dan Shipper (Every)

By Lenny's Podcast

Categories: Product, Startup, VC

Summary

Product teams face opposing demands: execute roadmaps while exploring frontier AI models. The solution is separating concerns into dedicated labs teams—even a single person can explore new capabilities full-time and bring breakthrough insights back to production, allowing you to harness early adopters without distracting execution-focused builders.

Key Takeaways

  1. Create a dedicated labs team to separate exploration (divergent, trying many approaches) from execution (convergent, focused roadmap delivery). This eliminates the cognitive conflict of doing both simultaneously and lets you scale frontier adoption without disrupting core product delivery.
  2. You can start with a labs team of one person whose sole job is exploring new frontier models when they launch, running experiments, and reporting findings back to the main product team. This requires minimal resources but captures maximum signal from early adopter instincts.
  3. Identify the natural early adopters already in your org—people who experiment with new AI models on weekends and personal projects. These individuals are living in the future and have critical insight into what your product becomes; harness them systematically rather than letting them distract others.
  4. Never make major product decisions within 30 days of encountering a frontier model. The initial excitement about new capabilities (like Astra or GPT-5) creates decision-making bias; establish a cooling-off period before committing resources to new directions.
  5. Use labs outputs to answer the core question: 'How do I harness early adopters without distracting everybody else?' Structure this by having labs document what new models can do, run proof-of-concepts, and bring validated learnings to the main product roadmap on a scheduled cadence.

Related topics

Transcript Excerpt

Okay. So, I'm going to talk about how to build products on a moving frontier. Um, or as I like to call it, the unreasonable ineffectiveness of business as usual during technology revolutions and what to do about it. So, the world just changed. The world just changed again. Uh, we had Fable 5.1 launched last week. We had Astra 6 launched last week. And these are [ __ ] sick models. Like look at this. This I made this like battle of waterlue 3D like historical re reproduction that's historically accurate with like a single prompt and it just churned for like four hours and then I got that like that's crazy. That is crazy. Someone else on my team made this uh sim agent simulation where he fed it a scientific paper about how agents interact and and how they get simulated together. And then he …

More from Lenny's Podcast