3 Takeaways - Scientists May Soon Design Entirely New Life Forms (#296)

Episode Date: April 7, 2026

We’re entering a world where life itself could become programmable.What if creating new forms of life becomes as simple as writing code? Geneticist Adrian Woolfson explains how close we are — and why the consequences could be extraordinary.

For billions of years, evolution wrote the story of life. Now, for the first time, we may be picking up the pen. Advances in artificial intelligence and synthetic biology are moving us from observing nature to redesigning it, giving us the potential to eliminate disease and even create entirely new forms of life. But if we can start designing life, where do we draw the line and how do we decide what life should become? Hi, everyone. I'm Lynn Toman and this is three takeaways. On three takeaways, I talk with

some of the world's best thinkers, business leaders, writers, politicians, newsmakers, and scientists. Each episode ends with three key takeaways to help us understand the world and maybe even ourselves a little better. Today I'm excited to be with Adrian Wolfson, a geneticist and science writer whose work sits right at the frontier of synthetic biology and the design of new forms of life. He spent years studying how biological organisms can be built from scratch, not just discovered in nature. He's the author of On the Future of Species, which explores how advances in AI and genetics could move us from observing evolution to actually shaping it. Adrian, welcome to three

takeaways. It's great to have you here today. Thank you, Lynn. It's such a pleasure to be on your show. It is my pleasure. Adrian, you say we may soon become the authors of species. That sounds almost like science fiction. What would that actually look like in the real world? What that means, is that we can actually literally take a blank piece of paper, imagine what we want to build in terms of the type of living creature or the types of properties that we might wish to introduce into existing life or to reach back into past life, and then literally write the code for that organism much in the same way

as we author a story or a piece of computer code, turn that into a genome sequence and then build the organism. That is so shocking. For billions of years, evolution had essentially no plan or direction. Are we now moving into a time where evolution itself becomes something we intentionally shape? I think we're moving into a time where life is going to have two co-authors. There'll be evolution by natural selection, what I call natural intelligence or natural engineering. And then there's artificial engineering, which is, is a sort of collaboration between natural and artificial intelligence, are humans largely guided by AI, but humans still having the final say over what happens. Is biology starting to look less like studying nature

and more like writing code? Absolutely. So the history of biology has been observation, documenting species, trying to work out how they connect with one another, how they originated, how they work. and that's really characterized all biology to date. And then at a certain point, Francis Bacon had this clever idea, what have we start to break these systems and crack them open

and then see the way in which they break down and see whether that can tell us something about how they work, and that's the scientific method. And now we're in a new phase of biology, you know, a kind of transition point, if you like, where suddenly we can begin to think about understanding biology through generating it. So build to generate, to test, to understand. And that's where we are today. And that relies on the ability of AI to kind of extract generative rules from existing biology and then our ability to synthesize genomes at scale, get that DNA into cells and boot them up. You introduce the idea of artificial biological intelligence, the ability to construct entirely new species.

how close are we to that reality? We're on the ski scopes. We're not on the double diamond black runs. We're on the ski bump, but we're on the mountain, and we can see the T-bars and the chairlifts and the cable cars that take us to those more difficult slopes. But we're skiing, right? And by that, I mean, Brian He, at the Ark Institute in Stanford,

is used AI to generate the first ever totally artificial species, which happens to be, I'd call it a biological entity, rather than a living species because it's a virus and viruses aren't technically alive, but they do have genomes like living things. When they infect cells, they become alive. So as of right now, as of today, we can design the genomes of viruses using AI. Therefore, we're writing the code of simple biological entities. And then it looks as if we can write the code of bacteria.

And then beyond that, we'll be writing the code of more complex, organisms. We're not there today. We're not writing Doskyovsky. We're writing kids stories as of right now, but we're writing. You know, when kids learn to speak, they don't start speaking in fluent sentences. They babble. Babies babbled. And we're babbling right now. You know, we're babbling the genome sequences of viruses and we're copying the genome sequences of bacteria and of yeast. But we're getting to the point where we're starting to get a real firm grasp of how things operate. And that knowledge will just increase exponentially over the coming years. Adrian, if species are designed rather than based on inherited characteristics,

does the very concept of a species begin to dissolve? Oh, Lynn, I'm so happy you ask me that question, because you're absolutely right. Because the concept of a species is a purely historical one. It means that information is coming down in a kind of linear vertical manner by descent, as Darwin described, and Wallace, using the mechanism of heredity. And the genes and information of different species is partitioned by the fact that they belong to the species and species can only reproduce with other members of the species. And that's the kind of historical way that things have happened. But obviously, when you can just write genomes from scratch, the concept of species becomes irrelevant, actually.

Not for natural organisms, it will always be relevant for them, but for organisms made artificially, the concept of species is totally irrelevant. These organisms that we will design have no historical reference point. They're a historical. They don't need to reference anything that's ever lived before or even anything that currently exists. You can just start from a blank slate and build something that's possible. and then test it in a real world after having simulated it on a computer.

It's so amazing. If we can engineer organisms on demand, does life itself become a kind of manufacturing platform for drugs, materials, even technology? Yes. I mean, let's not lose the wonder of life and the wonder of human existence, because I think that's special, I don't think anything I'm talking about excludes that or I have no wish. to try and destroy belief and all of those things, right? I'm simply saying that, yes, when we turn biology into a predictive engineering material,

we can use it for the benefit of humanity, and we can use biological systems to do things that help humans, like help us grow more food, to store information, to create energy, to create biomaterials. So, for example, you know, right here, right now, the population of the earth is just growing every year, and we're having to destroy rainforest to create the land to farm, to produce enough crops. There was a green revolution, but it's kind of stalling a bit.

The rate at which our food productivity is increasing isn't quite enough. And so if we have to create more land for farming and agriculture, but if we were able to farm, for example, in deserts or in high salinity environments, extreme environments, then that would change everything or increase the yields of crops. And in principle, we can do that. If we understand how the genomes of these cash crops operate, we can actually reprogram them or even totally rewrite them or find new cash crops which have never existed, which could actually be as beneficial as corn, rice, wheat. So there are many ways in which this technology could help us. But obviously, we need to be super cautious and really think carefully about guardrails and safety and ethics and responsibility.

You imagine a future, which you just talked about, where we could grow new crops, but we could also potentially grow houses, where living clothing, and use biological devices in daily life. Which of those feels closer than most people might expect? I think that we're going to start to see biology integrated in every aspect of our life, and that's just going to increase incrementally over time. it's always hard to sort of predict specifically if it's 10 years, 20 years or 30 years. But you just got to remember that even if it's 100 years, that's just the tiniest drop in the ocean of the history of life on Earth and even of humans, right? So it's going to happen increasingly rapidly. And I think within the decade, we're going to start to see biology being present in a lot of

our infrastructure, the materials we use, the energy we consume, the information that we store, that might be, you know, 30, 40 years, maybe less, who knows, but it's going to happen, I promise you that. And what are you most excited about? I'm just excited about getting rid of illness, number one, right? I want all of us to live healthier, longer lives. I don't want to see kids dying of cancer at the age of one or three or five, or frankly, at any age.

I don't want to see people suffering with debilitating diseases or suffering from Alzheimer. I don't want to see families destroyed. For me, the medical side of this is really, really important because if you understand how genome is working, in particular the human genome, you then can reverse engineer human disease. But also, I'd really like to see us stop destroying the planet and using these horrible chemical processes that create pollution. And I want to see microplastics, even in the Arctic, snow or Amazon rivers, species being destroyed, wilderness being destroyed. So I believe that we can use biology itself to preserve biology. So that's a huge part of my vision. Preserve nature, value nature, treasure nature.

It's our heritage. Let's keep it for the future, you know. It's a precious resource. What worries you most? There are lots of things that worry me. You know, unethical use of the technology, use for enhancements, used to basically undermine human nature or what I would call the non-negotiables of human nature like free will, sense of empathy and so on and so forth.

Obviously, the use of technology for bio-warfare or bioterrorism, but also there's a real risk that inadvertently we might cause great damage without actually meaning to by releasing new species whose behaviours are poorly understood and whose effects on ecosystems can only very poorly be modeled. So the key point is we just need to be really, really incremental and cautious about what we do and how we do it and just say, hey, you know, take a step back. Let's just do this cautiously and carefully and responsibly. And to recap, what you're saying is that essentially programming life will be like programming code and we will be able to potentially eliminate disease create new species, create new attributes or characteristics for existing species, including humans.

I personally have absolutely no doubt that that's the case. AIs already demonstrated that it can predict the structures of proteins, any protein that's ever existed or could exist. Honestly, we don't know how it works. What we do know is it doesn't make its predictions by understanding the detailed laws of physics and chemistry, right? It does it through looking at at huge databases and extracting patterns. And that's the way that AI works. It's not like mathematics. In fact, mathematics didn't turn out to be particularly useful for understanding how

biology works. AI does. So there are those who are pessimistic and say, oh, we're way too complicated to be predictable and there's an environment and this. Now, of course there is. But those are kind of details, sometimes important details, but more or less, I think biology is predictable, it is engineerable. we will master those rules, we will master the grammar of life, and we will gain control over the

construction and design of life. And I have absolutely no doubt that that will happen, and actually quite swiftly, because we're already doing it. But we're just in the earliest days, like AI is in its earliest days, and look how incredible chat GPT is. I mean, you can have an incredible conversation with chat GPT, which kind of transcends the kind of conversation you can have almost any human being. So just imagine what AI is going to be like in another decade. Imagine what our ability to design genomes is going to be like in a decade and to build them. So it's happening, you know, and it's happening fast. And this is because the ability of AI to understand the structure of essentially biology and the genome. Already we're seeing this with genetic diseases,

some of them being cured? Monogenic diseases. Those are diseases caused by a single gene like cyclonea or hemophilae. They're pretty easy to fix, actually. They do behave like broken parts in the machine, right? Component-like, and you can put in a new gene or edit them and not using gene editing. That works.

Unfortunately, most human diseases aren't that simple, and they don't behave like broken parts in a human-made machine. And that's where you need a different approach. And I think that it's quite likely that we're going to have to start to think about doing a bit of rewriting. Now, we would only do that, of course, if we were pretty confident that we knew the consequences of doing rewrite. And I don't think we're anywhere near ready to rewrite human genomes in any way that's inherited. So I think in the short term, we're going to get insights into human diseases, which can be treated in different ways which don't involve genome rewriting. but the rewriting the genomes of other species will be extremely useful for us.

Like spider silk, for example, is the same tensile strength of steel. And if we can devise using AI a version of 10 times a tensile strength of steel, that could be a really useful material, for example. Do you have a couple more examples like that? That's fascinating. Pretty much anything you can think of, like clams and mollus, for example, make a kind of natural glue when they stick to rocks. We can take some of these naturally occurring materials and then use AI to kind of re-imagined them

and simulate different versions of them that have never existed and then hallucinate to use an AI term something which has no direct relationship to that, but could be really useful for another purpose. Do you see biology as a treasure chest? Yeah, you know, I think once we're able to navigate this kind of infinite space of all biological possibility, because remember, the species that have existed or still exist represent the most infinitesimal fraction of all biological possibilities. So imagine if evolution had never discovered rice or corn or chickens to make eggs or cows or sheep.

If the tape had been rewound and the tape of life ran again, maybe you wouldn't get any of those things, right? But the potential for them was always there. So there are other things out there in the space of biological possibility that have equal utility or greater utility that could be really, really useful for us. So what we've done is we've kind of unlocked that treasure test. We've pulled it open and now we're exploring it. And there's a lot of interesting and really useful stuff in there for us. It is a Pandora's box and we need to explore that box with great caution, but there's a lot of treasure in there.

So exciting. Adrian, what are the three takeaways you would like to leave the audience with today? The first one, I guess, is we're at this critical transition phase in the history of life on Earth, where all of a sudden, naturally occurring life, which has got to be the way it is as a result of evolution by natural selection, will be joined by living species that are created artificially as a result of a collaboration between artificial and natural intelligence. My second takeaway is that everybody needs to participate in this debate to ensure that the science of generative biology is actually performed in a manner that safe, ethical, responsible, equitable, transparent and ultimately benefit society. The third takeaway of this is that although initially you might say, oh gosh, we're going

to be making artificial life and now scary and terrible, actually, my view is that it can help us to preserve natural life. And actually, if we don't engage with synthetic biology and artificial life, we're just going to destroy the planet, all moan species. We're going to destroy all the ecosystems, get rid of all the rainforests, and we'll be left with nothing. We'll destroy life's heritage. And that belongs not just us, but to our children,

and our children's children, to the future. And we have to be honest and safe and responsible. responsible custodians of nature. Adrian, thank you so much. I really enjoyed your book on the future of species. The future that you describe is both so exciting and breathtaking, as well as incredibly worrying. So thank you.

It's been a great pleasure. Thank you so much for inviting me on. If you're enjoying the podcast, and I really hope you are, please review us on Apple Podcasts or Spotify or wherever you get your podcasts. It really helps get the word out. If you're interested, you can also sign up for the Three Takeaways newsletter at three takeaways.com where you can also listen to previous episodes. You can also follow us on LinkedIn, X, Instagram, and Facebook.

I'm Lynn Toman and this is Three Takeaways. Thanks for listening.