Synthetic biology

Um, and she really taught me the, the basics of science, and she introduced me to one of her colleagues, Jay Keasling, at UC Berkeley.

Uh, at the time, he was one of the real leaders in synthetic biology.

I got really excited about it.

You know, and there's mentors that you get throughout life, and Judy sat me down one day and she's like, "You know, Matt, you're, you're a fairly smart kid.

I, I feel like you should go to graduate school." And I had no idea what graduate school meant or anything.

Um, so, you know, I got excited about synthetic biology.

I ended up at, uh, Wisconsin in Brian Blaker's lab in the chemical engineering department by getting a microbiology degree, so kind of the fusion of multiple things there.

Um, and just when I was up at Wisconsin, uh, Rudolf Barrengue was also at Danisco working on, uh, CRISPR, literally down the street.

Mm-hmm.

When I joined Colossal, we didn't have a bird program, but I really wanted there to be a bird de-extinction program because a- all of the tools that we are developing for de-extinction are the same tools that we can use to use synthetic biology to modify the genomes of living species and help them avoid becoming extinct.

Mm-hmm.

And so the kit that we're building from multiplex genome engineering to cellular rejuvenation to IPS cell technologies for wild animals to even learning the link between particular letters of the DNA sequence and what those letters actually do to cause an animal to look the way that it does, right? All of that is applicable across the board.

So the stack that we're building for de-extinction, for synthetic biology stack applies to synthetic biology for conservation.

So when I joined Colossal and there wasn't a bird program, they had launched the mammoth program, which is a placental mammal, and the thylacine program, or Tasmanian tiger.

That's a marsupial mammal.

And then based on how the different things of different weights basically spin out to different depths, and so you can bias the likelihood that you'll get a XY carrying sperm or an XX carrying.

And from there, we explore the bizarre behavioral quirks these models are developing.

We also unpack the sheer security nightmare of watermarking synthetic biology.

And finally, we will explore a massive new discovery in AI training.

Which is a breakthrough that really challenges everything we assumed about how models learn.

And it is not just limited to traditional software vulnerabilities anymore.

We also have a rising threat on the synthetic biological front.

AI generated synthetic biology is getting much harder to effectively track.

DeepMind recognized this specific danger and tried to build a solution.

Just visit jordanharbinger.com/start or search for us in your Spotify app to get started.

Today on the show, what happens when the ability to engineer a pandemic stops requiring a secret Soviet bioweapons lab and starts requiring roughly the same technical sophistication as cheating on your freshman biology homework? Rob Reid is back, and today we're looking at the increasingly uncomfortable marriage between AI and synthetic biology, including how researchers can order pieces of deadly viruses through the mail, why robotic cloud labs could be both revolutionary and terrifying, and how a government program somehow arrived at the brilliant idea of hunting down thousands of undiscovered viruses, identifying the most dangerous ones, and then publishing their genetic blueprints online.

The good news, there are surprisingly practical ways to make this much harder, from DNA screening and wastewater surveillance to futuristic UV technology that could potentially kill pathogens right out of the air.

So, you know, light stuff today.

Don't let people get pandemic-causing viruses and just sort of order them up online.

Um, and that's one of the things that we're facing.

And to go back to the passage of time, when they put that genome online, people who understood that, um, synthetic biology, like computing, is an exponential technology, which means it doubles and doubles in efficacy, and after a few doubling cycles, you have things that feel like, felt like science fiction 10 years ago.

Sure.

So y- you've started like four successful companies before Colossal, in, in gaming, in defense, in AI, in mobile.

And so you knew nothing about synthetic biology-

I, yeah.

True

... before you, before you started now a company worth over $10 billion.

Oof.

Uh, so why, what made you make de-extinction your moonshot? And having no background in synthetic biology, was that a hindrance or an advantage?

Oh, I think it's a massive, uh, advantage.

Man, there's things happening that I never could have imagined.

After this, we'll talk a little bit about synthetic biology and editing the DNA of that biology because that's happening too.

And that also can be equally scary.

But the nanotechnology stuff, like...

Like, we actually can cure a lot of genetic diseases.

But I'm very interested in how, you know, us monkey-brained humans, you know, are going to use it.

And then when you combine that with something called synthetic biology, have you heard about this one? That one I don't know.

I know

It's listed in Hong Kong, but has a big Shanghai research base.

And then Beijing, you know, as the capital, of course, it kind of has this strong academic and institutional strengths, but I think it's quite strong as well around, you know, computational biology and AI, SynBio as well.

It's colder up there actually, right? So if you need a lot of power and energy to kind of run platforms, it kind of makes more sense in that sense as well.

you know, like the power generation up there that's, you know, that exists, all of the power stations.

In Hong Kong, yeah, Florida, yeah.

So, you know, that's how I see it breaking down, you know, of course, with kind of nuances and exceptions.

But yeah, Shenzhen hardware, Shanghai, Suzhou, Biopharma, and then Beijing kind of AI and SynBio.

Are the Chinese companies able to recruit globally or even regionally? Like, where are they getting their talent from?