Subvolts take AI automates lab work, slashing scientific discovery costs and accelerating breakthroughs. Mar 24, 2026 - Developers
- Learn how AI agents can automate complex experimental workflows and accelerate research cycles.
- Business
- Understand how AI-driven automation in R&D can drastically reduce discovery costs and unlock new market opportunities.
This interview features a founder discussing a concrete AI application that significantly improved scientific research outcomes.
How Autonomous Labs Will Transform Scientific Research: Ginkgo Bioworks’ Jason Kelly
Audio streams from Sequoia Capital. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Sequoia Capital's Training Data podcast features Jason Kelly of Ginkgo Bioworks discussing how AI is realizing his vision of programmable cells. A collaboration with OpenAI used a reasoning model and robotic lab to advance biochemistry research by 40%. Kelly argues AI's scaling power, not just intelligence, will reduce discovery costs and accelerate breakthroughs in health and biopharma.
Imagine a robot scientist that can run thousands of experiments at once, day and night. This episode explains how AI is making that happen in biology labs, speeding up discoveries that could lead to new medicines and better health.
This episode offers insights into how AI is moving beyond theoretical models to practical, automated scientific discovery. It's for researchers, biotech professionals, and anyone interested in the future of scientific innovation. The discussion on AI-driven experimentation and accelerated breakthroughs directly relates to ongoing efforts by frontier labs like OpenAI and Google Gemini to integrate AI into complex problem-solving domains, pushing the boundaries of what's possible in fields like drug discovery and materials science.
- HostSequoia Capital
- GuestJason KellyFounder, Ginkgo Bioworks
- Guest: Jason Kelly, Ginkgo Bioworks
- Title: How Autonomous Labs Will Transform Scientific Research: Ginkgo Bioworks’ Jason Kelly
- Show: Training Data (an official AI show)
- Length: 58 minutes
- Ginkgo Founded: 2008
- Biochemistry Improvement: 40%
Jason Kelly founded Ginkgo Bioworks in 2008 with a simple but radical idea: DNA is code, and cells are programmable. Sixteen years later, AI is finally making that vision real in ways that could reshape science itself. Jason describes a landmark collaboration with OpenAI in which a reasoning model with access to a robotic lab beat the state of the art in biochemistry by 40% - not by being smarter than scientists, but by running experiments 24 hours a day and sharing data across a hundred parallel hypotheses simultaneously. He argues that the biggest inefficiency in science isn't intelligence, it's manual labor. Once AI helps scale research, the cost of discovery collapses and breakthroughs follow, with profound implications for biopharma, national competitiveness, and human health.
Hosted by Sonya Huang and Pat Grady, Sequoia Capital
- reasoning model
- AI models that can process information, draw conclusions, and make decisions.
- AI that can think and make choices.
- biochemistry
- The study of the chemical processes within and relating to living organisms.
- The chemistry of living things.
- hypotheses
- A proposed explanation made on the basis of limited evidence as a starting point for further investigation.
- Educated guesses to test.
1 h 08David Silver and Lasse Espeholt of Ineffable Intelligence: A Superlearner for the Age of Experience
WithDavid Silver, Lasse Espeholt
1 h 041 h 0552 minMaking Cities Awesome: Peregrine’s Nick Noone & Ben Rudolph
WithNick Noone, Ben Rudolph
1 h 03
37 minInside the Model Spec - Episode 15
WithJason Wolfe
Questions people ask
- What is Ginkgo Bioworks' core idea?
- Ginkgo Bioworks was founded on the idea that DNA is code and cells are programmable, a vision now being realized with AI.
- How did AI improve biochemistry research?
- A collaboration with OpenAI used an AI reasoning model and a robotic lab to beat the state of the art by 40% by running experiments 24/7 and testing many hypotheses simultaneously.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Sequoia Capital.