I believe AI's most beneficial application will be science. It's also one of AI's hardest tests: breakthroughs demand deep domain knowledge, multimodal data, and genuinely novel insight.

To understand where the field is heading, I went to the people building it, interviewing founders, CEOs, and researchers from Ginkgo Bioworks, MIT, Lawrence Berkeley National Lab, Boltz, and beyond. Here's what I learned, organized by area:

📋 Table of Contents

🤖 Autonomous Labs

Autonomous Labs – Primer

My mapping of the ecosystem including key players and technologies:

Autonomous Science Opportunity Map

Peek at the state of the art of lab automation, 2026:

https://youtu.be/DDivALKEqf8?si=KnbATz12wZEpSSkS

Universal Lab Automation – Rick Wierenga of PyLabRobot

Writeup:

https://www.discoveryengines.co/p/universal-lab-automation

Interview Only:

https://www.youtube.com/watch?v=tCxF37cCcWM

Biotech Lab Automation – Luis Villa of Bay Area Lab Automators

Writeup:

https://www.discoveryengines.co/p/inside-biotech-lab-automation