Episode summary
Faster Science, Better Drugs
The a16z Show
· · Hosts: Erik Torenberg · Guests: Patrick Hsu, Jorge Conde
Summary
The conversation centers on Arc Institute’s “virtual cells” moonshot: using foundation models to simulate human biology well enough to guide experiments and eventually accelerate drug discovery. The guest argues that science is slow because it is physically messy, incentive-fragmented, and multidisciplinary in a way universities and companies often do not handle well. Arc is trying to counter that with a single roof for neuroscience, immunology, machine learning, chemical biology, and genomics, plus flagship projects that can turn models into practical lab tools.
A major theme is that biology is behind language and image AI because the domain is harder to observe, harder to label, and far less native to most researchers. The guest compares the desired “AlphaFold moment” for cells to a model that can predict perturbations from one cell state to another, even if it cannot fully explain every mechanism. The discussion emphasizes that current progress is still early—somewhere between GPT-1 and GPT-2—but may improve through better data, benchmarks, and lab-in-the-loop experimentation.
The second half turns to industry economics. The guest and the host argue that biotech needs better target selection, better molecules, lower capital intensity, and faster development cycles, but clinical trials remain the hardest bottleneck. They point to GLP-1s as proof that large, societally important drugs can create enormous value, and say future breakthroughs will likely come from combining better models with better chemistry, gene editing, and more scalable experimentation.
Key takeaways
- The guest says Arc’s core bet is that virtual cells can make biology as programmable as software by predicting perturbations that move a cell from one state to another.
- The guest argues biology is harder for AI than language or images because scientists cannot directly “speak” biology, and many key signals are still unmeasured.
- The guest places current virtual-cell progress somewhere between GPT-1 and GPT-2, with future gains depending on public data, private data, benchmarks, and lab-in-the-loop iteration.
- The host says the industry’s three biggest bottlenecks are capital intensity, slow clinical timelines, and weak effect sizes from poorly chosen targets or molecules.
- The guest cites GLP-1s as evidence that tackling large patient populations can create enormous value and reset ambition across biotech.
- The guest says the most useful AI applications today are strongest in protein design, pathology, and workflow automation, while multimodal biology remains more hype than heft.
Notable quotes
“our moonshot is really to make virtual cells at ARC and simulate human biology with foundation models”
“I think the most important thing is science happens in the real world”
“Why are we so worried about modeling entire bodies over time when we can't do it for an individual cell”
“I think most people would agree where somewhere between GPT 1 and two, right”
“the thing that is very clear in complex diseases, right, where you don't have a single cause of that disease is there's some complex set of changes”
How the episode opens
I want to make science faster. Our moonshot is really to make virtual cells at Ark and simulate human biology with foundation models. Why are we so worried about modeling entire bodies over time when we can't do it for an individual cell?
We can figure out how to model the fundamental unit of biology, the cell, then from that we should be able to build. My goal is to really try to figure out ways that we can improve the human experience in our lifetime. There are a few things that if we get them right in our lifetime will fundamentally change the world.
Patrick, welcome to the podcast. Thanks for joining. Thanks for having me on.
I've been trying to have you on for years, but finally I could get your time. Here I am. I'm excited to do it.
It's going to be great. For some of the audience who aren't familiar with you and your work at at Arc and Beyond, how do you describe what's your moonshot? What what is what you're trying to do?
I want to make science faster, right? You know, we can frame this in high level philosophical goals like accelerating scientific progress. Maybe that's not so tangible for people.
I think the most important thing is science happens in the real world. If it's not AI research, which moves as quickly as you can iterate on GPUs, right? You have to actually move things around.
Atoms clear liquids from tube to tube to actually make life-changing medicines. And these are things that take place in real time. you have to actually grow cells, tissues, and animals.
And I think the promise of what we're doing today with machine learning in biology is that we could actually accelerate and massively uh massively paralyze this.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai for biology
- virtual cells
- drug discovery
- clinical trial bottlenecks
- capital intensity in biotech
Companies and tickers mentioned
- GOOGL
- GOOG
- Arc Institute
- Alphabet
Watch the episode
Watch “Faster Science, Better Drugs” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from The a16z Show
-
How AI Could Bring Halo and Other Classic Games Back to Life -
Mike Volpi on Why AI Infrastructure Will Outlast the Hype Cycle -
Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse -
Inside Salesforce’s Acquisition Strategy for the AI Era -
How AI Is Turning Figma’s Finance Team Into Builders -
The New AI Agents That Shop, Plan, and Gossip
The a16z Show, this episode and its artwork belong to Andreessen Horowitz. Octagon is not affiliated with the show; this page is an independent summary.
