How AI can unlock life-saving cures

Note: Transcripts are automatically generated from episode audio, and are not fully corrected for spelling, grammar, and formatting.

RANA EL KALIOUBY: If you know me at all, you know that I love tracking my own health data. I am obsessed with my Whoop – and use it religiously to track my sleep, my movement, and now my biological age! And I truly believe this is just the beginning. AI can help us unlock a world where we heal faster and live better, longer lives.

My guest today, Siddhartha Mukherjee, is at the cutting edge of a very specific application of AI in health: drug discovery. His Pulitzer Prize-winning book, The Emperor of all Maladies, redefined our understanding of cancer. Sid has now co-founded Manas AI, with Reid Hoffman, to accelerate drug discovery through AI. In a world where traditional drug development is too inefficient, too expensive, and slow, Manas is aiming for faster, cheaper, life-saving cures.

Sid and I sat down during the Masters of Scale Summit in San Francisco for this conversation.

EL KALIOUBY: Welcome to Pioneers of AI. I’m so excited for our conversation.

SIDDHARTHA MUKHERJEE: My pleasure.

Copy LinkHow an early mentor shaped a career in science and humanity

EL KALIOUBY: So you are a true polymath, you’re a cancer researcher, you’re an author. You now have Manas AI. I am curious — take us back to the early days of your journey. How did you get interested in cancer and medicine?

MUKHERJEE: So when I was 18 years old, I took probably the most consequential and longest trip of my life, which is when I came to university as an undergraduate. Those days were very different days. Remember, no cell phones, no email.

EL KALIOUBY: From India — where in India?

MUKHERJEE: Grew up in Delhi.

EL KALIOUBY: Delhi. Mm-hmm.

MUKHERJEE: And there were four students from India, four undergraduates in total, at Stanford. At that time, I imagine the number has become at least 10 times, if not 20 times that today. And so I landed up at Stanford, and my second year in, became more and more interested in genetics and in particular cancer genetics. And I knocked — and this was what you could do at a place like Stanford; I imagine you could do it in other places too — I knocked on Paul Berg’s door. Paul Berg had won the Nobel Prize for inventing gene cloning, or recombinant DNA. And not only that — this is very important — Paul had then organized what’s now known as the Asilomar conferences, to consider placing a moratorium on recombinant DNA and genetic technology until the ethical, legal, and biosafety issues had been figured out. And that combination of someone who could span these two worlds — a world of what I would call the humanistic sciences, and the world of pure science, the world of how science impacts humanity, or human beings, either directly or at a population level — was very attractive to me.

And so I knocked on Paul’s door and I said, Paul, I’d like to come and work with you. And I was a second-year undergraduate.

Maybe Paul was in a particularly good mood that day. He said, sure, come work with me. And so that began a 40-year relationship, in which we had been friends and so forth. The point of this, of course, is that we would meet often, and every time I was in the Bay Area, he and I would have lunch together. But the important thing was that it was with Paul that I began to learn that science has a human aspect, and everything that we do in human interactions, in human lives can be taken back to science.

And that was a very important formative moment for me.

EL KALIOUBY: Yeah. That’s amazing. And obviously this is an AI podcast and it’s still extremely relevant, this idea, right?

This whole idea. And it’s a recurrent theme in my life.

MUKHERJEE: I take my scientific work, try to distill it into readable forms, try to explain it to myself, to the world, and then I take it back from there into either the AI world or et cetera, and then try to learn something from the history and from what I wrote and from what I thought about while I was doing it.

Copy LinkWhy writing became a way to make medicine more understandable

EL KALIOUBY: Yeah. I am curious about how you started writing. When did that happen? And of course, your Pulitzer Prize — what got you curious about writing, and also documenting? I guess when you were putting the book together, you started documenting.

MUKHERJEE: Yeah. The book, which eventually — I should say that The Emperor of All Maladies is coming up on its 15th anniversary.

It’s coming out in a new edition with four additional chapters and a massive new introduction, because so much has happened in the 15 years. In any case, the book began as a journal, and then I soon realized that it was much bigger than a journal, that there was a lot more to say and a lot more to think about. It became bigger and bigger and bigger. And at one point in time I thought to myself, even as an oncologist, as a cancer doctor, as a cancer scientist, I didn’t know so much about why or how I — we, researchers, scientists, doctors, patients — had arrived here.

And that’s how this whole thing began.

Copy LinkWhat cancer research has taught us about genes and the tumor environment

EL KALIOUBY: Yeah. I’m Egyptian so I was struck that in your book, the earliest references to cancer were actually in ancient Egyptian texts. We’ve been grappling with cancer for a long time. What have we learned from the evolution of this disease over the years?

MUKHERJEE: Well, we’ve learned a lot. It would be hard to distill it in a few sentences, but the big picture is that, first of all, cancer is a genetic disease.

EL KALIOUBY: Mm-hmm. It is.

MUKHERJEE: It occurs because of changes in genes that either activate or inactivate cellular growth. And in doing so, the original cancer cell takes in nutrients, performs metabolism, alters the way it uses energy — but it’s a cell where the genes have mutated. So that’s the first thing. The second thing we’ve learned is that the mutations can come from many sources. They can come from random chance, which is when cells divide, they could make a mistake in copying, and so the daughter cell, the cell that’s born, could have a mutation or a change in the genes. They can come from inherited changes.

EL KALIOUBY: Yep.

MUKHERJEE: They can come from your parents’ DNA, they can come from viruses — in particular some viruses that can either cause mutations or introduce new genes into cells. And number four, they can come from carcinogens, from chemical or environmental insults that alter the genetic makeup of your cells. That’s the second thing we’ve learned. And maybe the third thing I’ll say that we’ve learned that’s sort of very globally important is that we now know that cancer doesn’t occur in isolation, but rather it utilizes its environment — its microenvironment, its tissue environment — to essentially sustain itself.

It borrows nutrients and signals, and that’s why only certain cancers grow in only certain kinds of places. Of course there are many, many more things we’ve learned.

EL KALIOUBY: Yeah. Yeah.

MUKHERJEE: But those would be sort of very high-level, three very big lessons in cancer. And I think what a lot of your writing has done too is that it’s opened this kind of science and these findings to a much wider audience, humanized it and made it accessible.

EL KALIOUBY: Well, absolutely. I hope it has.

MUKHERJEE: The book is written for everyone. Cancer biologists read it. Young doctors read it. Patients read it. A very interesting group of people who read it are people who might be completely outside the field but have an idea about it, and might come and write me a message saying, has anyone thought about trying this, that, or the other?

And that’s why the field is constantly mobile, and that mobility is very important to me because it allows me to explore new spaces for discovering new drugs, making new medicines, and changing the way we treat cancer.

Copy LinkHow to think clearly about rising cancer rates in younger people

EL KALIOUBY: Very cool. So before we dig into AI and Manas, I wanna ask you about the state of cancer diagnosis that’s making the news.

We’re seeing a lot of headlines about the rise of cancer in young people. What do you make of that? Is it more cancer for real, or is it just better diagnosis? Both?

MUKHERJEE: I recently wrote a long piece for the New Yorker about this.

They asked me specifically to tackle this kind of question.

EL KALIOUBY: The problem with diagnosis is that, of course, if you invent a new diagnostic test, the minute you invent the new diagnostic test, the number of diagnoses is gonna go up, right?

The most striking example is years ago in South Korea, they started giving out ultrasounds to primary care doctors and they would use that ultrasound to diagnose thyroid cancer. And so the incidence of thyroid cancer skyrocketed, the number of surgeries for thyroid cancer skyrocketed.

Guess what difference there was to the deaths from thyroid cancer? Zero.

So that’s an example where increased diagnosis does not mean increased mortality — it means increased diagnosis. That’s what it means. So we are yet to find out — I would say that there are certain cancers where we know for sure there has been an increase in mortality. A modest increase in mortality. Yeah. That’s a real statistic.

MUKHERJEE: For virtually all cancers, there’s been a decrease in the United States of mortality. I can give you real numbers.

So in 2000, the mortality from cancer — all cancers, all types — in the United States was about 200 per a hundred thousand people. In the year 2024, 2025, it’s going to be on the order of 140 per a hundred thousand people. So there has been an absolute reduction in cancer mortality. Now, that’s true for most cancers, but it’s not true for some.

So we’ll have to see whether these new diagnoses of cancer in younger people really translate into increasing mortality. I’m afraid it might, and in which case we’ll have to really find out why.

EL KALIOUBY: Our understanding of cancer has evolved, allowing us to fight the disease in innovative ways. But how can leveraging AI help? For Sid, diving into this question was a leap of faith. We’ll find out why, after a short break.

[AD BREAK]

Copy LinkWhy AI could reinvent the slow and costly process of drug discovery

EL KALIOUBY: Alright, let’s talk about Manas. So you and Reid recently founded Manas AI together. And you’ve said Manas AI is your leap of faith.

MUKHERJEE: Manas AI is a leap of faith because when I started thinking about AI, I thought that every aspect of drug discovery was inefficient, was not working, and could be made AI native and much more efficient. And by that I mean let’s just walk through them. So you start with a target usually, and by target I mean it’s usually a dysfunctional protein — could be a dysfunctional molecule in a cell.

So that’s your target. From there you validate that target using experimental methods in the laboratory. And then you try to use any mechanism — you try to jam it with a small molecule like a lock jamming a key, or you try to prevent that target from being active using a variety of strategies.

Manas AI is an attempt to take all of these steps all the way to human clinical trials and make them AI native. Right now our focus is mainly on making molecular generators. So we have partnerships that do these other aspects of drug development.

But right now the centerpiece of Manas AI is to build, we hope, one of the world’s most powerful molecule generators. A true generative foundation model for making new molecules that will become drugs.

EL KALIOUBY: And with the business model, you partner with other pharma and biotech companies and license the molecule generator to them?

MUKHERJEE: Yeah, we license them in partnership. So what we don’t do — just to be very clear — we don’t create enterprise software, for instance.

EL KALIOUBY: Yeah. The LLM. Okay.

MUKHERJEE: We use LLMs a little bit, but really we make foundation models and then use those foundation models to make new molecules and then get those molecules out through partnerships with various companies, until you reach an inflection point medically.

And that’s the purpose of Manas AI.

Copy LinkWhy generative chemistry needs more than a standard LLM

EL KALIOUBY: I imagine that biotech and pharma have been using AI for years. So what’s different now?

MUKHERJEE: The kind of AI that biotech and pharma have been using has been, I would say, very first-generation AI. And in fact, it’s not even clear that they have been using AI for true molecular generation.

EL KALIOUBY: But there’s a lot of machine learning baked into the discovery process already, right?

MUKHERJEE: There is some machine learning baked into some parts of the discovery process. Yeah. So there is certainly machine learning baked into target discovery. What target do you wanna send your drug after? And you can imagine that’s amenable to machine learning — you can imagine that’s something that LLMs would be very good at. You could search through vast data space to find targets.

And that has become relatively AI native. You can imagine a protein LLM. So those are now being incorporated into drug discovery, but they weren’t before.

Where there’s not been a lot of AI is in the actual molecular generation — like generative chemistry, where you’re coming up with new molecules.

EL KALIOUBY: Yeah. LLMs won’t work here.

MUKHERJEE: LLMs won’t work here. You can borrow some important principles from LLMs and from what’s been learned before.

But these are — I’m gonna use a fancy word that I’m happy to explain — these are true neuro-symbolic systems.

A neuro-symbolic system is a system that incorporates, on one hand, neural networks and AI, which of course is important because it’s the learning aspect of it, but it’s also symbolic in the sense that it is constrained by the rules of physics and chemistry.

Just to give you an example, you could say that a sentence generator is largely dictated by tokenizing language and then putting those tokens together in a meaningful way.

And it has a little bit of symbolic gestures, like grammar, but it’s largely governed by that.

When you enter the world of drug discovery, the symbolic rules take on a much larger role because physics is physics.

Chemistry is chemistry. You can’t violate those rules. Those rules are inviolable.

EL KALIOUBY: Think of it this way: an LLM for images can make up a horse with five legs, or a camel with wings. It can bend reality because in this case, it’s just generating a picture.

But to make medicine … you need a model based on the immutable laws of physics and chemistry. You actually don’t want it to make up something that cannot exist in the natural world. So, off-the-shelf LLMs won’t work here.

And so you have to build a system which on one hand learns, but also is cognizant of the rules, has to keep note of the rules. And not only that, it has to pay attention to those rules. Some rules are more important than other rules. To build a foundation model is more complicated — you have to hold both.

MUKHERJEE: What is the data you feed into these models if it’s gonna be generative? So there’s a lot of data out there. It’s mostly structural data — since the 1960s, scientists have been depositing structures, crystal structures of proteins, into a database called the PDB.

The Protein Data Bank. It’s a vast database. It’s a huge public service. And so we use some of that. We use laws of physics and chemistry. We also use programs like AlphaFold. And finally, we use molecular dynamics — in other words, how does the protein jiggle in space in real time, and if that jiggling makes a difference to a drug or not.

EL KALIOUBY: Given that much of the data Manas AI is using comes from public data sets – that any drug discovery company has access to – what gives Manas AI an edge? More on that in a minute. Stay with us.

[AD BREAK]

Copy LinkWhat gives Manas an edge in a crowded AI drug discovery field

EL KALIOUBY: With my investor hat on, I invest in early-stage AI companies. I am wondering what is your competitive moat? If all of your data sources are these public data sets, what makes Manas different?

MUKHERJEE: Well, I think there are two or three big differentiators. Number one is, we are not an enterprise software company, so we actually want to make medicines.

EL KALIOUBY: You wanna make end-to-end.

MUKHERJEE: Yeah. So that’s a big strategic difference. The second is that, even though there’s a lot of data out there, we are building what I would think is one of the world’s most powerful foundation models. It all depends on the model.

EL KALIOUBY: And also these are very disparate data sources — how you synthesize them, the multimodal approach for that kind of data. Exactly.

MUKHERJEE: So the way we harmonize all these data sets is very different. It’s proprietary. We have a very proprietary approach to making generative chemistry. So we do both — we do search and fit, which is non-generative. You can search molecular space and try to fit that.

EL KALIOUBY: And I guess AI helps make that faster, more efficient.

MUKHERJEE: It can make it faster and more efficient, but it’s fundamentally not generative, because you’re searching and fitting. But one of the jewels in our crown is that we have a generative foundation model, in which we basically look at a molecular pocket and try to actually make a chemical that fits. And that’s completely new chemical matter. The third big difference is that we are multimodal in a way most companies are not.

We’ve already started making drugs that are moving towards clinic right now. Even as we speak, they’re going through pre-IND, preclinical. They’ve been partnered, they’re sort of moving forward. Very exciting space. We also have a sister company that we helped launch called Athea, which works on siRNA — RNA as a drug.

EL KALIOUBY: Yeah. Which is the same kind of technology behind the COVID vaccines, I think.

MUKHERJEE: In a sense. But it’s a different kind of RNA in this case.

And we work across molecular disciplines. In fact, we have a sister company or partner company which does cell and gene therapy. Again, that makes us very different because we are multimodal. And being multimodal, we can combine therapies — we can take an antibody and add to it a small molecule and make a combined therapy that’s a combination of both. The last point is that we have a global infrastructure. I think it’s important to note that the United States has basically priced itself out of many clinical trials. Just to give you one example: in 2010, the United States in-licensed about $5 billion of drugs from China. In 2025, that number is gonna be about $60 billion. So you can imagine how much work is going on outside of the US.

We have a global infrastructure in India, spreading across now into Australia and other places. And that’s another big differentiator for us.

Copy LinkHow AI native drug development could make medicines cheaper faster and better

EL KALIOUBY: So let’s say Manas succeeds — and it will — what will drug development look like in the next decade?

And if the whole thing becomes AI native, what will that mean for drug development?

MUKHERJEE: The ultimate aim is cheaper, faster, better. Pharmaceutical companies claim — correctly in some cases, and incorrectly in some other cases — that the reason novel medicines cost hundreds of thousands of dollars is because 90% of them fail in early trials.

If you could change that number from 90% to even, let’s say, 60%, you would dramatically change the economics of medicines.

And so that’s the cheaper. The better is: if you start with garbage in, you’ll end with garbage out.

MUKHERJEE: In a lot of cases there’s this problem — they’re not better because they’re not using the enormous power of machine learning that we could have. And faster because, obviously, a machine-learning-enabled, AI-native system could take hours to survey the literature, take an initial stab at the draft of all this big stack. Yeah, exactly.

EL KALIOUBY: Fascinating.

MUKHERJEE: And again, I’m really focusing on things I don’t think are sort of pipe dreams in some kind of distant, vacuous future.

I’m talking about things that are very proximally available, and there are companies that are already working on this right now.

EL KALIOUBY: Yeah. Very cool.

Where is the human in the loop in all of this?

MUKHERJEE: We have human in the loop at every step. Just to give you one example, we recently encountered a medicine which was moving into quite advanced trials.

Our computational chemists took one look at it and they said, that’ll never make a medicine because it’s gonna be so insoluble. It was a very aromatic planar structure. Now you don’t know — I know, but

EL KALIOUBY: Yeah.

MUKHERJEE: these flat, aromatic planar structures, which look like Frisbees, are very insoluble. And they tend to make very bad drugs.

And so the trick is now to find out how to tweak that system. And that knowledge needs to be funneled back so that if you do a search and fit the next time round, don’t make a structure that looks like that.

Copy LinkHow AI and sensors could shift medicine toward prediction and prevention

EL KALIOUBY: One of the things I’m most excited about at the intersection of AI and health is this idea that we are at the cusp of a health span revolution.

It’s personalized medicine, but it’s also really focused on wellness. My thesis there is that there’s the trifecta of sensors, which are becoming ubiquitous, data, which is also becoming very accessible and available, and then both generative and predictive AI. So zooming out of the cancer world, the disease world — where do you see medicine going, and what role can AI play in that?

MUKHERJEE: Well, obviously for any medical discipline — and this has been proved time and time again — prevention is better than the cure.

But you need a lot of things in order to prevent; it’s not so easy to prevent.

So there are ways to get around all of this. One of them is sensors.

A sensor could be an AI-native biomarker.

MUKHERJEE: And by biomarker, I mean it marks something. It marks the future presence of a disease.

EL KALIOUBY: I’ll give you an example. Take an EKG or ECG as an example. All of us get ECGs and then a doctor looks at them and says, oh, you have or don’t have a problem with your heart rhythm. Right.

Mm-hmm. There’s much more information we now know in an ECG — there are small things that no one has ever even looked at. And if you put a massive data set together — let’s say, 5 million ECGs — which exists, right?

MUKHERJEE: And it is digital. And then you ask the question: which of these people is likely to develop congestive heart failure five years from now? Now all of a sudden you’ve used that as an AI problem — a Bayesian problem. Now I know a population that’s going to develop a disease in the future, and I’m gonna now make an intervention, whatever your intervention might be, and try to see if I can prevent the disease from happening in the future or not. So this integration that you’re talking about is an enormously important thing.

EL KALIOUBY: My last question — I always ask all my guests on the show. What do you think it means to be human in the age of AI?

MUKHERJEE: I obviously am very immersed in the technology. So I’ll have an answer which I hope doesn’t sound technocratic but sounds humanistic. I also am a humanist. I write books. I care about compassion. I’m a physician.

EL KALIOUBY: I think your partner is an artist. Yeah, exactly.

MUKHERJEE: I think it’s important that we improve human lives first, and that the goal should be to use the machines that we are making to make lives better. And I think they will, as long as we keep that in mind.

And as long as we don’t make machines for the sake of making machines, and as long as the goals are clear — the goals should be to increase compassion, dignity, empathy, the great human qualities that we have, creativity — and use these as tools rather than ends in and of themselves. I think that’s sort of what it means to be human in the age of AI.

EL KALIOUBY: I love it. Sid, thank you so much for joining us on the show. What Sid is doing with Manas AI is a perfect example of how AI can be harnessed for the greater good of humanity. Sid has spent decades advancing cancer research — and now, AI is helping supercharge that effort. We’ve often spent time on this show talking about the potential harms of AI, but conversations like this underscore how powerful AI can be when it is used to solve meaningful problems in the world.

But what do you think? What are other ways AI is supercharging benefits to humanity? Reach out to us at ‪(601) 633-2424‬. That’s ‪(601) 633-2424‬.

I spoke with Sid during the Masters of Scale Summit in San Francisco. You can find more thought-provoking videos from Summit at the Masters of Scale YouTube channel.