How Autonomous Labs Will Transform Scientific Research: Ginkgo Bioworks’ Jason Kelly

Watch on YouTube ↗ | March 24, 2026 at 12:01 | 58:28 | Sequoia Capital

Speakers

Jason Kelly — Co-host, The Deal Podcast

Summary

Jason Kelly, founder and CEO of Ginkgo Bioworks, explains how AI reasoning models combined with robotic autonomous labs can run real experimental science, citing a Ginkgo-OpenAI project that beat a cell-free protein synthesis benchmark by 40%. He argues the bottleneck in science is manual labor, so automated and cloud labs should collapse discovery costs, improve lab utilization, and disrupt biopharma, with strategic implications as the US loses drug-discovery ground to China. Kelly also lays out Ginkgo's commercialization path (automation racks, a Department of Energy deal, a $39 cloud-lab service) and near-term biotech applications he favors: GLP-1 and consumer-health products plus longitudinal molecular diagnostics.

Ginkgo's OpenAI project: a reasoning model with a robotic lab beat the cell-free protein synthesis state of the art by 40% after six rounds.
Kelly argues most science spending is manual overhead, so robotic and cloud labs could lift utilization from under 20% toward 70% and deliver roughly 10x more data per dollar.
Ginkgo sells reconfigurable automation racks (97 to the DOE Genesis mission), runs ~50 robots scaling to 100 in Boston, and launched a cloud lab service with $39 experiments.
He expects AI to disrupt biopharma R&D and warns that China's share of novel drug deals rose from under 5% to over 40%, with faster trials and more research per dollar.
Near-term biotech applications he is excited about: GLP-1s as consumer products, broader consumer/healthspan products, and weekly molecular health monitoring with diagnostics.
He doubts humanoids will do lab work and criticizes life science tools for incrementalism; he sees AI superintelligence arriving category by category with breakthroughs ahead.

Ideas

Jason Kelly Co-host, The Deal Podcast 0:22

AI and autonomous labs will disrupt biopharma

Unlike prior tech revolutions (internet, social media) that were just back-office IT for biotech, AI plus autonomous labs will change the fundamentals of how experimental science is done, and big science industries like biopharma will be disrupted - he really believes this. Drug development keeps getting more expensive because discovery runs on manual labor (Baumol's cost disease): less than 5% of research budgets go to reagents while most goes to overhead, people, and lab space, so every biopharma R&D head and research university must engage with this shift and cannot ignore AI.

Jason Kelly Co-host, The Deal Podcast 3:25

Ginkgo commercializes autonomous labs for AI science

Ginkgo has built and is commercializing the automated, AI-driven lab infrastructure for experimental science. Its Boston lab runs about 50 robots scaling toward 100, with in-house scientists dogfooding the system; it sells reconfigurable automation racks on a per-box service fee plus software subscription (97 sold to the Department of Energy for the Genesis mission); and it launched a cloud lab service with experiments as cheap as $39. The OpenAI collaboration showed a reasoning model with a robotic lab beating the state-of-the-art cell-free protein synthesis benchmark by 40% after six rounds of ~30,000 experiments by running 24/7 and sharing data across 100 parallel hypotheses, validating Ginkgo's bet that removing humans from the bench collapses discovery costs, lifts lab utilization from under 20% toward 70%, and captures a large new market.

Jason Kelly Co-host, The Deal Podcast 38:13

China is winning drug discovery research

China is rapidly winning drug discovery: its share of the industry's novel drug deals went from under 5% three years ago to over 40% last quarter; it has as many scientists as the US who are paid less, so it generates more research per dollar, and it can run phase 1 trials in about 6 months versus roughly 2.5 years in the US. He does not see why China would not win in research, making this a structural competitive shift the US must respond to and a signal of rising Chinese biotech strength.

Jason Kelly Co-host, The Deal Podcast 51:39

GLP-1s are trillion-dollar consumer products

GLP-1 drugs are worth close to a trillion dollars and are effectively a consumer product rather than a disease treatment - the best thing since the iPhone - because users stop thinking about food and get their willpower back; he personally takes them. Their value comes from serving people who are not sick, which shows how much larger biotech's market becomes when it addresses consumer wellness (weight, sleep, muscle, feeling better) rather than only disease.

Jason Kelly Co-host, The Deal Podcast 52:31

Consumer biotech applications have massive upside

The applications of biotech in the consumer space are bananas: products that add healthy years of life or improve weight, sleep, muscle, and daily function are worth effectively unlimited amounts - he frames the value of five extra healthy years as near-infinite - yet there is no good approval pathway because the FDA and the industry are oriented around treating disease. The shift toward health when you do not have disease (the MAHA-style framing he praises) is good for the industry and should open a very large market, and he expects it to happen.

Jason Kelly Co-host, The Deal Podcast 54:12

Longitudinal molecular monitoring is a major opportunity

The real signal about the body is molecular, not wearable heart-rate data: people should get a weekly blood sample and a broad molecular readout tracked longitudinally so they can test different interventions and see how each affects them personally. This monitoring trend - normalized by figures like Bryan Johnson, who spends ~$2 million a year on diagnostics, and enabled by services like Function and Quest - is still very early, but he sees it as the other big near-term application for biotech alongside consumer products like GLP-1s.

Up Next

Why Now is the Best Time to Buy Public Software Companies

ILTB Podcast · Mar 24

Iran War: Kuwait, Bahrain Hit by Iran Strikes As Optimism Fades | The Opening Trade 3/24/2026

Bloomberg Markets · Mar 24

She's building a company to solve America's caregiving crisis

Why Iran Is Trump’s Greatest Taco and the Fed’s Impossible Split: Bits + Bips

Unchained (Chopping Block) · Mar 24

Businessweek Convenes: Black Leaders Face the DEI Backlash

Bloomberg Markets · Mar 24

This Sequoia Capital video, published March 24, 2026, features Jason Kelly discussing XBI, DNA, Chinese biotech, GLP-1 drugs, Consumer biotech / healthspan, Molecular diagnostics / longitudinal health monitoring. 6 trade ideas extracted by AI with direction and confidence scoring.

Speakers: Jason Kelly · Tickers: XBI, DNA, Chinese biotech, GLP-1 drugs, Consumer biotech / healthspan, Molecular diagnostics / longitudinal health monitoring