Sep 28, 2023 · 37m · no-priors
No Priors Ep. 34 | With Ginkgo Bioworks Co-Founder and CEO Jason Kelly
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of No Priors, hosts Elad Gil and Sarah Guo interview Ginkgo Bioworks CEO Jason Kelly to discuss how automated foundries, biological foundation models, and horizontal platform infrastructure are digitizing synthetic biology and revolutionizing biosecurity.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 25.5% of the talking time here. How this is scored →
The hosts as informed peer 6.4 Guest teaching 4.3 Guest disagreement 1.3 The hosts pushing back 1.8
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 16:46 Kelly rejects the biotech industry's platform skepticism
Kelly forcefully rejects the established industry orthodoxy that biological modalities are too dissimilar to support shared horizontal platforms.
Hardest push from the hosts ▶ 30:58 Elad pushes back on employee super-voting governanceElad challenges Kelly's employee-voting structure by pointing out that non-founder CEOs may struggle to make hard, unpopular decisions like layoffs without falling into popularity contests.
Biggest teaching moment ▶ 13:00 Kelly on why AI beats humans in biology faster than human languageKelly re-educates the hosts on AI disruption, arguing that because biological code never co-evolved with human cognition like English did, neural nets will dominate bio far faster than law or text.
The host holds their own ▶ 25:09 Elad cites MIT wet lab realities to counter biosecurity panicElad leverages his personal experience handling viral vectors in MIT research labs to provide grounded context against theoretical AI biosecurity alarmism.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| DNA as Code and the Unpredictability of Biology vs. AI | 7 | 3 | 1 | 1 | Elad demonstrates substantial domain knowledge in biological systems, expanding on Kelly's points by detailing how self-evolving neural networks will mirror messy evolutionary biological mechanisms. | |
| Ginkgo's Platform Model: Abstraction Layers and Foundry Automation | 4 | 5 | 1 | 1 | The hosts ask clarifying questions about Ginkgo's foundry abstractions and client interface, while Kelly explains how Ginkgo automates wet lab work to decouple design from experimental execution. | |
| Biological Foundation Models and AI's Advantage in Biology | 4 | 6 | 1 | 0 | Kelly provides an extensive tutorial on foundation models in biology, arguing that AI will beat humans in biology faster than in English because genetic code is not native to human cognition. | |
| Biotech Industry Structure: Horizontal Platforms vs. Vertical Integration | 7 | 4 | 2 | 3 | Elad details the economics of drug development costs, and Sarah presses Kelly on why the biotech industry has stubbornly remained vertically integrated rather than modularizing into horizontal platforms. | |
| The Data Bottleneck in AI Drug Discovery | 7 | 4 | 1 | 1 | After Kelly details Ginkgo's wastewater biosecurity monitoring infrastructure, Elad demonstrates deep subject mastery by citing historical SARS containment and lab leaks in Beijing and Wuhan. | |
| Evaluating AI Biosecurity Risks and Evolutionary Defenses | 8 | 3 | 1 | 2 | Elad draws directly on his past hands-on research in MIT labs to challenge biosecurity doomerism, arguing that empirical lab practices show how difficult cross-species viral jumping really is. | |
| Ginkgo's Unique Governance Model: Employee Super-Voting Shares | 7 | 4 | 2 | 6 | Kelly outlines Ginkgo's employee super-voting share structure, but Elad mounts a strong pushback, questioning how hired-gun CEOs can execute unpopular necessities like layoffs without founder authority under employee voting. | |
| Energy Efficiency, Alternative Architectures, and Lessons from Evolution | 7 | 5 | 1 | 0 | Sarah and Elad engage in an advanced discussion on energy efficiency and AI architectures, with Elad explaining complex genetic mechanisms like bidirectional RNA transcription and catalytic gene duplication. |
Statements from this episode (12)
Kelly: Neural Network Analysis Will Function Like Systems Biology
“Like you're about to experience, like the analysis of these neural nets is going to look like systems biology, right? It's going to be like, go in and like, try to back, figure out a thing that you didn't design, my friends.”
Gil: Self-Evolving Neural Networks Are Coming Quite Soon
“And so I think people are really underestimating what happens once we have self, self-evolving neural nets, which I think is coming quite soon.”
Kelly: AI will outperform humans in biology faster than in English
“And so I feel like these computer brains are going to kick our ass a lot faster in this domain than they do in English.”
Kelly: Major biotech and pharma companies remain fully vertically integrated
“Where's the fast platforms? Like, where's all the horizontal stuff? Where's the operating systems? And it's like nowhere, like vertical integration, Merck, Pfizer, Bayer, Syngenta, like every one of these companies is like its own tech stack top to bottom, you…”
Kelly: Biological AI's main bottleneck is availability of training data
“The big limitation in bio is the availability of data to train these things, right? And so you have this tough situation where, like, everyone is doing these models of training on the same data, right?”
Kelly: Ginkgo operates a 300,000-square-foot robotic data lab
“Yeah, we have a 300,000 square foot robotic lab that they built in the last 10 years, and so we generate that, and we do it in service of our customer projects, we can do our own data generation, but yeah, that's where it comes from.”
Kelly: Ginkgo runs pathogen wastewater monitoring in US, Qatar, and Ukraine
“One of the things we're doing is, like, with the CDC, we run programs where we collect wastewater from inbound airplanes, and we sequence the DNA, and we look for pathogens, we look, we monitor variants and all this, both for flu and COVID, and I can add other…”
Kelly: Biosecurity should mirror cybersecurity with persistent monitoring and rapid response
“Cybersecurity, I think is a bit the mental model for like what the future of infectious disease response looks like. Persistent monitoring, rapid response, kill it.”
Gil: SARS leaked from laboratories four times in first two years
“Because SARS, they both snuffed out with the original form, but then it leaked four times in the first two years after it was cultured in a lab. It kept leaking from what eventually became the Wuhan Institute of Biology when they moved it from Beijing to Wuhan…”
Kelly: AI biosecurity lone-actor risk is low due to data scarcity
“The idea that, like, we know how to, like, exactly, like, design for that sort of thing is low. You could try something, but it's not like, oh, I know for sure it's just someone's waiting to do it. Remember, you need data. It's hard to accumulate that. Yeah, i…”
Kelly: Ginkgo gives 10x super-voting shares to all active employees
“At Ginkgo, we took this Silicon Valley idea of founder super voting shares and extended it to the entire employee base. So it's not just us, anyone who's at the company, and it goes away if you leave, ah, gets 10 X voting for their B shares versus the A shares…”
Gil: Self-replicating code will evolve hyper-optimized energy efficiency
“And I think the second you have self-replicating systems where you have code writing its own code, And you start going down that evolutionary path, you should have hyper optimization for energetics and for all sorts of other things, because it's just going to …”