#76 - Jeff Huber: Innovation, Google, GRAIL, Triatomic

Foreign.

Welcome to the Insightful Investor Podcast, a weekly series that seeks to share industry, investment and market insights. We define insights as concepts that are counterintuitive, widely misunderstood or underappreciated. In other words, unique ideas that you probably won't hear elsewhere. I'm Alex Shahidi, the host of the podcast and Co CIO of Evoke Advisors and a leading investment advisory firm. Learn more about our show@insightfulinvestor.org Today's guest is Jeff Huber. Jeff is the founder of Triatomic Capital, which is an early stage venture capital firm focused on applied AI and is dedicated to helping entrepreneurs build century defining companies, which we're definitely going to talk about. Previously he founded Grail, whose mission is to detect cancer early when it can be cured. And Jeff is also a former Senior Vice President at Google where he helped build and scale several multi billion dollar businesses including Google Ads, Gmail, Calendar, Google Docs, Google Maps, and Google X Life Sciences. Things that we've all heard about and appreciate that you were involved in creating those. Thank you for joining us, Jeff.

Thanks Alex. I'm really excited to be here on the Insightful Investor Podcast.

All right, so let's start with your background. You grew up on a small family farm and started your first business at age 14. Would you tell us about those early experiences and how they shaped your approach to innovation and entrepreneurship?

So I'm the youngest of five growing up at a farm in the very northwest corner of Illinois, and my siblings and I always joke that growing up on a farm is a character building experience. And specifically we were primarily a dairy farm, but very democratic in the other things that we did. So we had beef cattle and hogs and then crops to feed everything. But one of the implications in particular of being a dairy farm is cows get milked twice a day every day, whether you want to or not.

They don't take vacations.

They do not take vacations. So in fact, my father, who became the patriarch of the family and his family and then ours obviously from when he was 14 years old when his father passed away, was never away from the farm for more than a day until he retired in his mid-70s. So the sense of accountability, responsibility is very high. And I knew it was a character building experience. I was up every morning at 5:30am before going to school to help my father milk the cows, before jumping on the school bus to go to school, frequently with cow poop in my pants. But the school that I went to, my graduating 8th grade class, was 12 kids and roughly half of them were the same situation that were showing up with cow poop on their pants. And while I recognize it was a very special place to grow up, character building was a great experience. I also recognized from a pretty early age that I was really not cut out to be a farmer.

You saw what the path was.

I saw the path, and there was a moment in particular that was quite crystallizing. I was 12 years old, and in the springtime, my job was to let the cows in and out of the pasture. And to do that, you had to cross the cow yard, which in the springtime, between cows doing what cows do and the spring rains, it ends up being a bit of a soup that you're walking through to let them in and out. And I was doing that one day in my rubber boots, and I got about halfway across the cow yard, and I got to the point where, literally I was stuck. So I was in the middle of the cow yard with cows looking at me and had a moment of clarity as I went through the first stages of deciding whether I was going to step out of my boots and then go barefoot the rest of the way or what exactly I was going to do. But in that moment of clarity, I saw my future, and it was filled with a lot of cow manure. So that was really a motivating factor to figure out that there had to be a better way, there had to be a different future. I mentioned I was the youngest of five. My parents had established a policy in the family that they were very supportive of education. My mother earlier had been a grade school teacher, but their policy was that they would pay for the first year of college. And then beyond that, you were on your own. You had to figure it out. And from my older siblings, I had enough advance awareness of that, advance warning of that, that when I was 14 years old, inspired by computers and the beginning of the computer revolution, this was the very early 1980s, decided to start a business that I ran out of my bedroom at the farmhouse. And specifically, it was selling computer equipment via mail order. So in a lot of respects, it was E commerce before the E existed.

So you had this fascination with computers. How did that develop? And ultimately, you left the small town in Illinois and ended up in Silicon Valley. Tell us about that.

So the initial inspiration around computers came because I had the good fortune or luck of having a couple of mentors that were very material in my thinking and development. Specifically, I had an older brother who was 14 years older than me, and then also an uncle who had grown up on the Farm, left the farm, and were working in professional jobs in Chicago, which was about three hours away. And specifically they were both accountants, which doesn't sound like a very cutting edge discipline necessarily. But actually at the time it was because accounting was an early adopter of personal computers and technologies like spreadsheets. So they were at the leading edge of going through this revolution that was changing completely how their discipline was done. And they were very excited about computers, were big advocates of it, and quickly infected me. And I was able to convince my mother, or parents more broadly, who had just gotten an inheritance from her uncle, that computers were the future. So she took her very small inheritance, a few thousand dollars, and basically spent all of it buying a computer for her precocious son, who had been able to convince her that that was the future. And that ended up being an Apple II computer that I got again in the very early 1980s and managed to teach myself programming and computer operations. But then the issue I quickly ran into was you needed content for it, you needed to create and save things. And the medium then was floppy disks, which as it turned out were about $50 a box, which is probably about 3 or $400 a box in today dollars. So on my $5 a week salary that I was getting on the farm, it took a long time to be able to do that. So at 14 years old, I managed to find a wholesale distributor that would sell them for $20 a box. And while I was able to convince them to sell to me, which they probably shouldn't have officially, then the initial ones, I then saw an opportunity of being able to build a business selling those. So the E commerce business started with floppy disks and then moved into software and ultimately software development and many other things over its evolution.

Were you one of those kids that would open up the computer and try to figure out how it works?

Absolutely. So that was my reputation before getting the computer where just about every electronic thing I was taking apart and trying to figure out how it worked. And I couldn't resist the temptation when the new computer came. So there was a community of hackers that were essentially souping up the computers. So at the time, the Apple II is actually an Apple II plus computer. All of the input to it was uppercase and it was 40 characters wide across because that was a limit of the resolution of the screen that we're using, typically a television screen. Even so, there were modifications to make it do uppercase and lowercase to make it do 80 columns instead of 40 columns to add memory to the system. The complication is that all of those things involve a soldering iron. So the first thing I did after getting the new computer and orienting myself was take a soldering iron to it and do some surgery to soup it up. Much to my parents horror, but it worked out okay.

So when you look back, what values or lessons from that upbringing do you feel have stayed with you throughout your career?

I mentioned character building, experience. I think one of the key lessons is really that sense of accountability and responsibility. If something happens on the farm, nobody else is going to fix it. You have to figure out what to do. So my father, yes, was a farmer, but he was also a plumber and mechanic and hydrologist and whatever it took to keep things going. I think also there's a key lesson in resilience. There are many, many things outside of your control on the farm, and whatever comes your way, you need to be able to deal with it. And as a foreshadowing for later in my life, one of the things that I've really taken to heart is that you can't choose what happens, but you can choose how you respond. And I think that sense of resilience, of how do you rise to challenge and what are you going to do about things really stands out. I think a final one is just humility of on the farm, there's no job that is too big or too small and no job that's too dirty. You do whatever it takes to get the job done.

I've always felt an admiral goal is to leave the world a better place. How has that philosophy influenced your decisions and shaped the direction of your career?

So I've had a very strong sense of mission and purpose from the very beginning of my career. I think the beginning of that was so after the time on the farm, I applied to colleges and compared to kids today, it was a very simple algorithm that I applied. I went to the best school that I could afford to go to, which turned out to be the University of Illinois, which had a great engineering program, but also in state tuition. So I think I was paying $1,500 a year or something like that to go there. And the application process was submit your test scores, and if your test scores were high enough, you were admitted. So I feel sympathy for kids that are going through the crazy college process today. But one of the things that I was incredibly lucky when I was at University of Illinois because I was living in the future and I didn't completely recognize it yet. University of Illinois was an early adopter around Internet technologies. So when I was a student there in the late 80s. We already had Internet access and we had email and we had online discussion groups using that news access. We had instant messaging between terminals in the system. So when I graduated in 1989, I was released to the real world and quickly discovered that the rest of the world wasn't there yet and I didn't have those tools. So my first mission on graduating was I wanted to get those things back. I personally wanted to have. But then I wanted all of my friends and family to, to have access to the Internet because I could see how it would change how people were informed and entertained. So that sense of mission from that first job and then that was reflected in my first, what I would say, real experience in Silicon Valley, which was working for a company called At Home Network, which built out all of the broadband networking infrastructure. And then that continued at Google. There was another step or two along the way. But ultimately my landing at Google was being inspired by Google's mission to organize the world's information and make it universally accessible and useful. And when I went to Google in 2003, it was still a very small company. It wasn't the Google of today. I was employee number a thousand roughly when I started. So I wasn't there in the garage days. But it was still a very small organization and an incredible sense of mission and purpose behind what people were doing. People were going there because of the mission, not because Google was Google. And then later, and we'll probably talk more about it, but my experience with Grail is incredibly purpose driven, mission driven. And Grail's mission is to detect cancer early when it can be cured. And I think Grail, and we'll probably Dr. Moore is the clearest path that I can imagine to ultimately save many millions of lives. And a takeaway reflecting across all of those that crystallize it for me. When I was at Google, my boss at the time was Larry Page. And Larry had a very strong sense of anything you're going to do is hard. So if you're going to do it, you might as well make sure that when you achieve it, it matters. And that's really resonated for me as well. And in fact, many times it's easier to when you have a big mission, a noble goal, something that's going to have impact, it's easier to get others rallied to the cause and supporting it because they identify and they buy into the mission.

And I'm sure it's also easier to actually accomplish the mission because the, the end goal is in Sight and you can see the positive impact it would have.

Yes. Or another way I phrased it is when you have the right mission, when you have the right goal, the universe conspires for you to help make it happen.

That's interesting. So you joined Google before it was public and you played a key role in building products that we all use today, like Google Maps, Google Ads and all the Google Apps. What would you say was the vision of the company in those early days? And how did its unique culture enable such extraordinary growth and innovation?

So the mission for Google was, I mentioned it already was set pretty early. Organize the world's information and make it universally accessible and useful. The lore of the company is that the mission was created in a hot tub in the original garage where Google was founded and had its first half a dozen employees. And I mentioned that that was really the mission that drew me to Google. I had been at ebay previously, which was a great company, but in a lot of respects ebay when I was there was much more of a marketing driven place as opposed to Google really believed fundamentally in innovation through technology. So the kind of things that would have taken weeks or months to accomplish at ebay were a five minute discussion at Google because it was the right obvious thing to do to drive innovation through technology. Interestingly, Google also as a complement to that, had a portfolio mindset from the beginning that was articulated as 70, 2010. So 70% of the resource and focus should be around the core business, which at that point was the search business. But already Google was starting to invest in new categories like Google Apps. So what became Gmail, Calendar Docs, that set of things. There were nascent efforts of exploring and building prototypes of those even in those earliest days. And then the 10 were more crazy out there, game changing things that were low probability, but you wanted to be investing in already. So again, even in those early days, there were efforts around what became Google Maps and Street View, which was the very audacious program to basically go out and photograph the world, which is reflected in the Google Street View project, where you can see your family home where you grew up as a kid, or you can see where you went to school, or you can check out a restaurant and see what it looks like before you go and visit it and make sure it looks interesting or it's in a nice neighborhood. But what was really going on behind the scenes was that was a massive computer vision project that was both improving the data quality and accuracy of Google Maps because not only could we tell where a business was but we could literally be at the level of seeing the sign on the door of what are the hours. And that data also then ultimately fed into Google initiatives like self driving cars, because we knew where every street sign and street light and every line and curb on the road were. So those were seeds that were planted even in those early days when I started in 2003, 2004, it's interesting.

So you had that 70, 20, 10 and some of the 10 eventually fell into the 70.

Exactly.

And then you have a new 10 that replaces that. That's how you keep innovating.

So it's dynamic. And then Google Apps, when they were up and launched and that became a multi billion dollar business, that became part of the 70. And then you're continually exploring and looking for the next 20s and 10s and then even within an area it's fractal. So as we were ramping up Google Apps and all the Google workspace, we had our core products, but then we were continually looking for the next things coming along.

There's another aspect of Google that I think is interesting and you held significant leadership roles there, but you have to earn the respect of your teams through technical strength and vision. Would you talk about that as well?

So I think one of the things that Google got really right about culture was reinforcing that innovation through technology dimension. In fact, when I started a reflection of the culture, I had been at at Home Network where I was a senior vice president of engineering and operations, had a very large team. I then went to ebay and was a VP of architecture and systems, had a very large team. But then when I contemplated going to Google, my boss at the time, the then CTO said, jeff, I recognize this is going to sound strange. We are making a bet on you for the future. We really want you to be here. But the best way for you to do that is to come in as an individual contributor. And we recognize that that's a leap of faith. But you just have to trust me and Google culture. And then we talked about Google culture and why. So I went from a very fancy title and a big team to coming into Google as an individual contributor. And the encouragement of my then CTO boss was come in, engage deeply in the technology. People will recognize your technology depth, but your natural leadership will show true and people will naturally want to start following you and and essentially they will nominate you as the leader. Whereas if I declare it from day one when you come in, then people will be skeptical and it'll actually be much harder for you to build the technology foundation of how things work at Google. And that ended up working out. Exactly. I joined within six months working with the team and really in addition to working on some technical pieces, I put the focus on what are the goals, what are the things that we're trying to accomplish, what's the story behind what we're doing and the impact it's going to have. And that was the thing that then resonated within the team, where later I had a hundred people on my team within six months. Interesting. Another quirk of Google culture. Google was ferociously distributed and flat. So while I had actually it was 86 people on my team, they were all direct reports, there was no management layers between. So it really meant at Google that we had to find people who were very smart but ultimately were very resourceful and really took initiative for getting things done. It was a very light management hand or it was a very goal driven management style at Google.

What did you learn from co founding Google's life sciences efforts in Google X and how did that experience prepare you for your next chapter?

So that inflection at Google came when I had been at Google for 10 years. I had been at Google for a decade and it seemed like that milestone at that point. I had helped build Google's ad systems, which is the economic engine behind Google. My team built Google apps now in Google Workspace, the consumer products outside of Search, Gmail, Calendar docs, et cetera. And then I had been responsible for Google Maps when it was a very large Enterprise, a $5 billion P&L 5,000 people. But when I hit that 10 year anniversary, it felt like that was a good opportunity to pause and reflect a little bit, look back on, okay, what did I accomplish and what were the things that contributed? Because if I were contemplating starting a second decade at Google, I wanted to make sure that both I was doing things for Google that were going to be high impact, but also that I was excited about and energized about and I wasn't just turning the crank. And when I reflected on that and looked back, it was really the early days of ads, the early days of apps, the early days of Maps where I felt like I had the biggest impact, where it was building the team, it was getting the strategy set, it was setting the ambitious, audacious goals that we were going to go after. So one dimension of it was that I wanted to be in the early days again of building at Google. The other one was I self assessed that I was very energized by learning of throwing myself in the deep end and having to figure it out. And that process of learning I wanted to find again and I had done it previously. So before going into ads, I had no background, I had technical background but no business background. So it was figuring it out quickly. And same thing with Maps. And with that then insight, I went to my boss at the time, Larry, and said I would like to make this pivot move from driving the battleship here where I was running the Maps team and move into Google X which is a collection of crazy new things like self driving cars and balloon based Internet access for the world and drive a new area which was life sciences. And my insight there or the motivation for that was I saw out in the industry, out in the world this just tidal wave of data that was being created around genome sequencing. And it felt like with all of that data, if Google, if I at Google could apply some of the technologies and approaches that I'd spent the last decade developing there, where my teams, for example built the first machine learning systems at Google, the first AI systems at Google and then the first deep learning systems. If we could apply those machine learning, deep learning, big data approaches to this new category of data. It was both a Google scale problem, but also an area where Google could have a real impact on driving insights and the next breakthroughs in science and research. And then the other was back to the personal motivation is I had always been fascinated by biology. I had no official background, but I saw a massive learning opportunity there. So as I made that pivot into it, I joked with friends that I was starting a night and weekend PhD in biology. And it was kind of true now with the Internet and now even More so with AI and LLMs, anybody can learn anything. But that was at this point now, 12 years ago, pre commercial AI and LLMs. So I started it's probably there in my bookshelf. It's called the Big Red book on biology, which is the biology 101201 book that undergraduates do. And I started on page one of the big Red book and started going through teaching myself biology. But it's possible now.

That's amazing. Would you describe your transition from Google to Grail and how all that came about?

So I was a few months into actually rewind a little bit. I was literally a week into the process of making that decision to focus on life sciences at Google X. And again one of the driving opportunities that I saw there was I'm a big believer that innovation happens at intersections. So there are people who spent a lifetime in biology and life Sciences. There are people who spent a lifetime in computer science. There's incredibly capable people who have driven fundamental innovations there. But very often you see that the biggest opportunity for what's next is at intersections of disciplines. So it felt like driving that discipline of computer science meets life science, good things at the middle was a driving factor. About a week into that process, after I had officially made the decision, I got a call out of the blue. The universe is conspiring in good ways. Got a call from Illumina, who was the leader in genome sequencing. And they were thinking similar things where they saw this tidal wave of data and they wanted somebody to help them figure out their big data strategy. So I was invited by them to join the board of directors and focus on that with them. So it was the first step and I was about six months into that process where then the universe conspires in this case, not in as positive a way. Where my wife Laura was diagnosed with very late stage cancer. And that was a shock. She was at the time 46 years old, super healthy, super fit, did everything right, ate right, exercised way healthier than I was, had no family history of cancer whatsoever and started having some vague symptoms. She was more tired than usual, that was very atypical for her. Started having some joint pain and went into her doctor for annual physical and doctor look at her and saw a healthy 46 year old woman and said, welcome to premenopause. Because that was the most likely description of what was going on. Things continued on. She started having some other symptoms, GI symptoms, which again were very atypical for her. Went back to the doctor, doctor ordered a series of tests, the test came back all normal, nothing going on. Another couple of months go past, she starts having some more, greater symptoms. They went back and said, oh, maybe it's mild Crohn's disease or irritable bowel syndrome. We'll do a combination of endoscopy and colonoscopy and just see if anything's going on. And then from that they identified what was bad news but potentially good news, which was they identified a small colon tumor, a 2 centimeter tumor. So bad news, you have cancer. Good news is you may have caught this early and this might be the positive case. When we did the next level of workup, unfortunately it wasn't that at all. When you do a CT exam and it just lit up where she had extensive metastasis to her liver throughout her lymph system, it was very advanced stage four cancer. And through that process in parallel, I was on the board of directors at illumina. And we're in a community where we have access to great physicians. So we began the treatment process for laura initially with optimism, because we have access to the best doctors, we have access to the best diagnostics. And she and I had a very positive attitude of, this is beatable, this is winnable. Unfortunately, what we found was that it wasn't. When cancer is diagnosed at late stage, you, are continually behind and you are continually chasing. And part of the issue is that cancer is a disease of mutation. And over time, the mutational complexity of cancer increases exponentially. So that's not a single thing that you're fighting. Cancer is evolving under the darwinian pressure of the immune system and its own evolutionary growth. So even though initial treatments were positive, ultimately, it kept coming back and coming back in parallel. When I was on the illumina board, it's a very technical place as well. Shared many properties with google. The R and D teams were sharing their latest projects with the board, and one of those projects was the seed crystal that ultimately became grail. They were sharing data that showed when you took illumina's sequencing technology and really turned it to its most extreme level, turned it to 11, that they could demonstrate finding traces of cancer, fragmentary DNA that's shed by cancer at its earliest stages in the blood. And this is a category that's now called liquid biopsy. And by extending illumina's technology to a further extent, There was potential to be able to detect cancer early. And it instantly resonated to me that if that technology had been available three or four or five years earlier, it could have profoundly changed the outlook for laura. But more broadly, it could profoundly change the outlook for many, many millions of people in the future. So as laura was going through treatment, you, could see this technology coming along. Though we had access to the best doctors, best diagnostics, best treatments, Unfortunately, Laura's battle was a losing battle. She passed away in the fall of 2015. She was diagnosed in 2014. And in parallel, this technology had been coming along at illumina, and the illumina board knew I was very passionate about it and asked me to ultimately lead the cause and taking that technology out of illumina and commercializing it where we could get the capital and the technical resources we needed and just the focus to drive that technology forward and to create what ultimately became grail. So Laura passed away in november, and grail was launched in January of 2016, basically two months later. And back to my earlier observation of, you can't choose what happens, but you can choose how you respond. That was an incredibly difficult period in time for me, for my family. But I saw this technology, this capability that had so much potential, and that's when the foundation was being built. So I dove headfirst into that with a great sense of mission and purpose.

Thank you for sharing that story. I know it's difficult to go through that. What would you say were the biggest challenges you faced in building Grail? Especially as someone coming from outside the life sciences, having to overcome this steep learning curve.

There were many challenges. There was a challenge of expectation. Grail was launched with very high profile and very high expectations. So in some respects, we were shot out of a Cannon with a 40 person science team that came over from illumina on day one with a over $100 million Series A, which was unheard of at the time, with the additional expectation of when we mapped out the clinical study program that was needed, it was going to be over a billion dollar program. So we needed to instantly get on the road to fundraise and raise the billion dollars that we needed to do it. Though there was great initial data, it was still an unproven category that it could be done at scale. And while there was a prototype that had been created at Illumina, there was also a strong sense and a little bit of dissonance between my view and Illumina's. Illumina viewed it as, here's the proof of concept, just commercialize that. But it had a pretty profound limitation, which was it could detect cancer, the existence of cancer, but it couldn't tell you anything more about it. And I felt very strongly from the beginning that whatever we were doing needed to be actionable. So part of our clinical study structure that we created was not just using the initial prototype that had been created at Illumina, which was a great product, but looking at additional dimensions of what were the signals that we could detect in blood at the earliest stages and using additional dimensions that could inform more. And one of those is a technique that was very cutting edge at the time called methylation, which is not just looking at the DNA, but it's essentially the chemical decoration on DNA that differentiates cell behavior in your body. So you have roughly 35 trillion cells in your body. If you don't have cancer, all of your cells have the same DNA, but obviously your cells behave differently. Your brain cells behave differently than your gut cells behave differently than your heart, than your lungs, than your skin, than everything else. And methylation is essentially the software layer that runs on top of the DNA hardware that differentiates that behavior. So by using this additional dimension of sequencing where we were looking at methylation, not only could we detect aberrant cells, mutated strands of DNA, but we could also interpret the methylation to be able to predict where they came from. So it's not just saying. The initial version of Grail would have been, you take the Grail test and it would say you have cancer or you don't have cancer, but if you did have cancer, good luck, now you need to find it. What we were able to do by pressing to this next level and really focusing on the right product for patients, for clinical use, we were able to have a very highly accurate prediction of where it was. So not just you have cancer, but instead you have lung cancer or ovarian cancer or pancreatic cancer, so that it could be actionable. So that was another challenge. The final one was, and I under appreciated at the beginning of it really was a multidisciplinary effort where we needed the best science in the world, we needed the best clinicians in the world, we needed the best engineers and machine learning, deep learning AI engineers to be able to pull this off. And the culture differences between those groups was greater than I initially recognized. So I thought at first naively, that we have the noble mission, detect cancer early when it can be cured. We've got our set of goals, everyone should be able to align on those. But what I ultimately found is while everyone had the right intentions and positive intentions, there were a lot of cultural discontinuities or mismatches between them and the nature of how people did work. So we ended up investing a lot of time in articulating our Grail values that were the unifying elements ultimately of Grail culture, so that we could bring those disciplines together in the right way to achieve what we needed to.

I assume there were also many skeptics and potentially competition as well. When you started, how did you navigate that landscape?

I would say an absolute of startups and innovation is overcoming skepticism, because if everyone agreed it was a good idea and everyone agreed it was possible, somebody would have done it already. So you always have skepticism. And in fact, if there isn't skepticism, your idea probably isn't unique and novel and innovative enough. So part of it is you just have to recognize that that part of the landscape that you need to overcome, and the way you overcome it or the way that we thought about it at Grail, is just really putting incredible focus on integrity and rigor around what we do. I mean, if you think back to the day that was in the window where you had companies like Theranos that were out there that were making grandiose claims that had no substance behind them. And in fact, Theranos had just hit the wall right in the window where we were launching Grail. So one of the first questions when I would do interviews for Grail actually was one of the last questions from journalists, and they felt like they had to ask the question, and they would ask it almost sheepishly, is why is this different from Theranos? And the easy answer was, it was different on every dimension because we believed in. The only thing that was common was their starting point of we take a blood sample. Beyond that, we were putting incredible scientific and technical rigor behind what we were doing. We had a board of experts that were the leading cancer doctors in the world. Our board of directors included the former head of the National Cancer Institute. On every possible dimension, we were different and differentiated. We felt that fundamentally, but also just given the skepticism from the outside world, we had to really demonstrate it at a whole nother level.

Obviously, when you build a business like that that becomes successful, you have to scale it to grow. Are there lessons from scaling Grail that you now apply when advising or investing in new startups?

Yes. So as mentioned earlier, Grail was shot out of a cannon. There were incredible expectations around what we were doing. And one of the biggest challenges was that we needed to execute the largest clinical study program ever done. Ultimately, it's been over 300,000 subjects in the clinical study program. And our goal initially was clinical studies are done all the time. Our goal was to take best of breed and put together a system and architecture that was using best of breed of contract research organizations and existing tools. And what we found was those systems and tools had been created in a prior era that lacked the imagination of what was possible or what was coming. So for example, the state of the art tool for tracking clinical study data had literally a hard limit in it of tracking 10,000 subjects, because they couldn't imagine anyone would ever do a clinical study with more than 10,000 subjects. So in a lot of respects, it's like the famous Bill Gates quote from the mid-1980s when they were designing DOS for the original IBM PCs. Why would anyone ever want more than 640k of memory? So is that equivalent in life sciences, biotech? And we just found that repeated over and over again as we looked at the labs information management systems and the limitations of those. So ultimately, what we had to do to scale was really focus on first principles and in most dimensions actually build it ourself. We had the luxury of access to Google caliber engineers to be able to do that. So we ended up building our own systems ultimately for how we executed and ran the clinical studies, how we built the laboratory information management system, how we aggregated and analyzed our data. So I guess the net is really encouraging first principles. Thinking of what are the things that are going to differentiate you, what are the things that you have to do because it's not possible given that you want to achieve, and then what are the places where you can leverage other capabilities or state of the art so you're not doing everything.

So eventually you transitioned to Triatomic capital. Would you talk about the mission and origin story of Triatomic capital?

The transition there from Grail to Triatomic was driven by. So I went from being the founding CEO of Grail to moving to being the Vice Chairman. And part of that was that as I reflected a couple years into that journey, I had gone directly from my wife's passing to immersing in Grail and I had been working incredibly hard on that. I was getting on. It dawned on me as I was getting on a plane for I think the fifth or sixth time that month that my children had already lost their mother and in some respects they had lost their father as well. And that wasn't fair and appropriate for them at all. So that was a tipping point for me of we've got the foundation built and there are people out there who have experience in doing this at the next level that could probably do as good of a job as me on running and executing the day to day. And I could step back a little bit, focus on my family and then still continue to be very actively involved. So I made this step to becoming Vice Chairman. Grail continued along, we raised the billion dollars that we needed to raise, we ran the clinical studies that were the foundational product development studies, drove an entirely AI based, machine learning, deep learning based product development process which I think was was the first in life sciences to create the Grail product. And we were heading towards IPO with a successful launch of the product when then a suitor came knocking. Illumina came back and expressed interest in acquiring Grail. And boards of directors always make unanimous decisions. And I was the last holdout of Unanimity, but ultimately agreed and supported the acquisition because it could and should be an accelerant for Grail to get to scale and ultimately be accessible around the world. So when that deal closed and depending on how you countered, it was somewhere between an 8 and $10 billion deal. So grail, while it's technically scientifically clinically successful and impactful, was also a very good investment for our early investors when that deal closed. That then freed me up for what am I going to do next? And as I thought and reflected on that, both in the end of my time at Google and then as I was chairman at Grail, I had started ramping up more investing, primarily angel investing, seed stage investing. And I found that process really exciting and rewarding and reinforced the thing that I talked about earlier of continual learning and curiosity driven approach. And it was just so much fun working with young entrepreneurs who were incredibly excited about what they were doing and wanted to change the world. So Triatomic, the genesis was I ended up teaming up with two friends, colleagues were longtime investors. One at Coutu Management where it spent a decade there, another who'd been at Deerfield Management spent a decade there, were really builders of those firms. And we decided to join forces in Triatomic. And our mission with Triatomic is that we want to work with great entrepreneurs to build sensory defining businesses and technologies.

Very high bar.

Yes. But back to the point earlier, if you're going to do something, anything you're going to do is hard. So you might as well make sure that it's going to have an impact and make a difference.

And I'm sure it's very gratifying too in some ways you're on the other side of it. Where initially you were involved in building these companies and these technologies and now you're assisting others who have similar long term objectives and you can give them advice on the experiences that you've had. Your learnings how to build these businesses, how to scale them and you can watch them grow as well.

Correct. It's a different way of achieving scale and there is a lot of benefit. Given my operator background and having built systems of scale, built multiple products used by a billion people at Google and then having built Grail versus where we had to run the largest clinical studies ever done. So that operator mindset and perspective of having been in the front lines and been there, done that is a benefit for our founders and something that resonates with them, but it's in a more scalable way. That said, I do have with the operator in me sometimes is a little bit impatient of I want to jump in and be very closely involved. But I try to strike the right balance of guiding, mentoring, advising rather than hands directly on the wheel myself.

Well, I know you focus on five megatrends at Triptomic. Would you describe what they are and why you've chosen these areas?

Maybe one step back first on that, because it's really the foundation that all of those are built on. So our core strategy with triatomic capital is focused on what we call applied AI, which is a very opinionated view of where AI will ultimately have greatest impact and value. And it's a very data first, data centric approach. And that thesis is informed by, if you look back on the history of AI over the last, certainly 10, probably more than 15 years, increasingly the output of AI systems is entirely defined by the data that you feed it. So we look for teams, companies that are generating interesting, important, valuable, proprietary data and then using AI to unlock and accelerate the value of that data. And then as a complement to that or another dimension of that, we also look at companies that are the fundamental enablers of the generation of data or management of that data at scale. So those are the two dimensions of applied AI. We then concentrate that in our century defining themes. And the concept behind the century defining themes is these are our bets on the areas that will have biggest impact over the next decades. But ultimately the absolute test is imagine yourself at year 2100, looking back and if you said try to distill what defined the 21st century, what are the areas that mattered the most? So those five are first we call engineered biology, which is now ability to engineer biology at a cellular level. There were some big breakthroughs on that starting at the beginning of the century, the massive breakthrough of the Human Genome project, where now we can read DNA, we can read the code of life. Barely a decade later, we've got breakthroughs like crispr, where now we can edit and write the code of life. We think the intersection of those capabilities is incredibly profound. And you can already see it with the ability now to cure previously incurable diseases. We've already made huge progress on things like cystic fibrosis and sickle cell disease. And there was a recent case at UPenn of a baby being cured of a rare genetic disorder with a custom therapeutic that was created for them, delivered via crispr. So we think that the implications of that are very significant. And at the limit, we now also control our own evolution as a species, which is both exciting and scary at the same time. But if there were one thing that stood out, we think there is potential that this century will be the one that's defined by our ability to now engineer biology at a Cellular level. The other themes are important and very complementary. So the next is, we call it new materials, which is deterministic material design, computational chemistry, engineered nanomaterials. How do we build the stuff the future is made of? Third one is next gen compute, which is what are the compute architecture and system breakthroughs that'll drive the next inflections in AI. So I started at Google 2003. My teams built the first machine learning systems at Google in 2003, 2004. The architecture of the systems being built today aren't really different than they were then. The one addition is about seven or eight years later, the introduction of the GPU to be able to handle some of the math for calculations in AI. But it's not a fundamentally different model. We think that there's opportunities. When you look at the software architecture evolution of AI, increasingly it's inspired by biology and the architecture of the brain. So you'll hear about neural nets or transformers as instantiations of neural nets. Your brain doesn't have a CPU and a GPU and memory that's thrashing back and forth when you need to do things. It's actually a very inefficient model. Your brain has distributed memory and computer. And there are architectures now that instantiate in hardware the direction that software architecture has gone inspired by biology and the architecture of the brain. We're very optimistic about those fourth theme called new energy, which is when you think in decade and century terms, energy is ultimately the limiting factor for anything that you do. And we're already seeing that with AI. So what are the next breakthroughs in energy? Whether it's next gen fission, fusion is actually becoming increasingly real. And we'll see fusion on the grid in our lifetimes. But as a complement to that, what are storage technologies, battery technologies that can help make that transition happen? The final one is probably the most broad. We call it new economy, which is the digitization and automation of everything. So everything from robotic automation to software platforms and agentic AI systems that are changing the nature of work. So those are collectively our century defining themes. Those are the themes that we're building the firm around. Our first fund is investing in those themes. Because we have five themes, it doesn't mean that we do 20% in each. It really is. We're opportunistically looking at the best opportunities within those. And back to this concept of intersections, also looking at intersections between those, because we think in those intersections that's disruptive technology from outside that has potential to drive products and businesses that have disproportionate returns.

So you have founders coming to you looking for capital and or counsel. What would you say are the most common mistakes you see them make when pitching to triatomic?

So we see a lot. We went back and did account in Q1, we do a quarterly update for our LPs and we went back with our Q1 investor letter and did account. We talked to in Q1 400 companies and through that process made two investments but one of them had actually started earlier. So we basically made one and a half investments out of the 400. So the selectivity is incredibly high. There's one level of just is an investment the right fit for us and what we do so we screen on is it really an instantiation or application of applied AI with our definition? Is it one of the themes that we're focused on? So there's a set of things that no fault of the entrepreneur go away with that. The other one where I would say where they're not errors but it's just the maturity is as an entrepreneur you're continually balancing near term execution with the longer term. So what so many times you'll get entrepreneurs that just focus on the technology. And this technology is amazing without the what are the implications, what's the business, what's the financial model, what's the distribution model, how do you get there from here? Or then you have others that are great at the marketing but don't have the fundamental breakthrough of technology. So it really is, I think that balance and finding the right balance of the differentiating technology, the thing that's going to make a difference and be defensible and protectable within the discipline of what's the path to get there and get to impact?

What's your outlook for the pace of discovery and innovation in biotech over the next decade and what would you say is the base case and what's the upside scenario?

Biotech as an industry has had a tendency of going through swings and I would say at the moment the pendulum swing is out of favor. But I think that that presents a massive opportunity because if you look at what's going on behind the scenes, all of these trends of being able to generate increasing quantity and quality and fidelity of data, the ability to generate data is dramatically higher than it has been and it's getting better every day. And now we have increasingly capable AI systems to integrate that data and integrate everything that we know. So that intersection we think is going to drive incredible breakthroughs and progress. But now you've got at the moment, a macro market where the pendulum has swung against biotech. So the bar is incredibly high for entrepreneurs to build in the area. That said, if you look back in the history of other industries, for example, when I went to Google in 2003, 2004, that was shortly after the Internet bubble meltdown. The Internet was out of favor then. You could have bought Amazon for $4 a share. Google was toiling in obscurity and it was an incredible time to build because you could get the talent that you needed, concentrate the talent that you needed and build without the overhang of intensive competition or crazy valuations or things like that. So I think now is an incredible time to be building in biotech and life sciences. And I think you'll see generational companies that come out of this period like generational companies. Amazon, Google, Facebook came out of the Internet meltdown of the early 2000s.

You mentioned those five century defining themes, but would you say are there any emerging markets or technologies you believe are currently underappreciated by the broader venture community?

Life sciences, biotech? I think with the pendulum swing is currently underappreciated. I think another that is significantly underappreciated is around next gen computer. There is a tendency of people wanting to anoint winners and assume that they will always be the winner. An example of that now in AI is Nvidia. And Nvidia is an incredible company. They've executed well, but in a lot of respects they were lucky to be in the right place at the right time where they had created a product, the gpu. And a lot of people have forgotten that G stands for graphics. It was a graphics accelerator for video games and scientific computing. And they got lucky and were in the right place at the right time where that capability was able to be applied to machine learning and machine learning training, where the vector math for graphics was able to translate into the matrix and tensor math needed for AI. And in a lot of respects it is good and useful and was a significant step forward. But it's using a hammer for every job where now there are tools that are much better suited. So an example of that in our next gen compute category is we made an investment in a company called D Matrix. And in our view D Matrix is really the right architecture for where AI goes next. And specifically D Matrix was the king of the era of training where everyone was building these large models. The GPUs were the right hammer for the job of being able to do that better than otherwise could have happened. But GPU architecture isn't Particularly efficient where the market is now exploding, which is using those models, which is an area called inference or AI serving. And we think that the inference market is going to be orders of magnitude larger than the training market. And D Matrix was a company that was built from the beginning with an architecture that's optimized for inferencing. I already alluded to it earlier of their model, their hardware model is distributed memory and compute. So a hardware architecture that's much more like your brain works and it's all computer engineering, so there's nothing for free. But by having the right architecture that matches the software architecture more efficiently, depending on the benchmark, they can be anywhere from 5 to 30 times the performance and price performance, and even more importantly increasingly the power performance for AI inference serving applications. So if you think about your own personal behavior of how Frequently you use ChatGPT or Gemini or Anthropic, and then imagine what that's going to look like a year or two or three from now, where not only are you doing it on your phone and on your computer, but you've got a thing whispering in your ear all the time, your useful AI assistant. We think that the market is going to be just massive and there's going to be a big drive for much more price and power efficient solutions for doing that.

I wanted to close with a few questions on leadership. You've obviously led transformative teams at Google, Grail and now Tritomic. What would you say are core leadership principles that have guided you across such different environments?

I think the core ones are clarity of mission, clear objectives, and then leadership by example. On clarity of mission, each of the companies from At Home Network to Google to Grail, and now with Triatomic, those are companies or organizations that have had very clear missions. And I mentioned Grail. One of the first tasks out of the gate was to raise a billion dollars. And that was incredibly challenging and was really unprecedented for the time. But the best, most productive discussions that we had with prospective investors were we would go in. The first slide that we would bring up was the mission. And for most of the successful meetings, we basically never got past that first slide. It was just a discussion about the mission, the importance, how we were going to do it, the culture of achieving that. And then yes, people wanted to do their diligence and they would dig into the background and the model and plans and all of that, but really people either got the mission and aligned with it or they didn't. But thankfully for us, many, many, many did. And it was really joining the cause to do that. And when I look back at Google, it was joining the cause of Google to do it. And now with Triatomic we are very mission focused in what we want to achieve. So I think when you have the right mission, it really helps define the culture and helps build the team of people who are passionate about that.

Especially if you're good at articulating that mission and explaining why it's the mission, what drives that goal, and then gaining a following. Then basically everything else is just details.

One of the things I used to joke at Grail was it always bothered me that the mission had one too many words. I wanted it to be a seven word mission and it was an eight word mission. Detect cancer early when it can be cured. And that was a little bit of a joke, but it is just reinforcing the more crisp and clear you can be around it. That said, if we hadn't been acquired and if we had stayed involved, there was actually an articulation of the mission that accomplished that, which was take out the word cancer. And Grail could and should have been the early detection company. So detect early when it can be cured across all diseases and conditions. So that was the next horizon to go beyond the bright spot that we're aiming towards in the near term. Second is around just clarity of goals. One of the things I really took away from Google, and again this was something that Larry Page was very passionate about, is having the right goals. So at Google, and then I've applied at every other place I've been, is having a very clear goal setting process, specifically we call them okrs, objectives and key results. But what's the target that you're shooting for a year from now and then what are milestones along the way where you can see that you're making progress towards it and then also having a mindset where those are stretch goals where you're really pushing yourself and the organization to achieve those. And I think the power of that multiplied across the organization. And it was a fractal model. So there were company level goals, yearly goals and quarterly goals that then would reflect down to the organization level and ultimately down to the personal level where they're nested and everyone can see how those connect together was a very important one.

And I'm sure it builds momentum as well because you have these milestones and you reach it and you feel like you've made a positive step in the direction towards the mission. Rather than waiting until you reach the finish line, you're making progress and that builds that following and drives incentives and continues to motivate the team.

Yes. And additionally, from a management perspective, it's also creating a more frequent feedback loop. So you can see every quarter, are we on track or not? Or if we didn't achieve this goal, what happened? What's the post mortem on? Why didn't we achieve that or make the progress that we wanted to make? So I think that combination of the right goals and then that built in sense of feedback loop so that you're continually improving is a really important one. And then I think the final one is just leadership by example. I've always been a huge advocate on my teams. I would never ask anyone to do a job that I wouldn't do myself. And I think reinforcing that and demonstrating with the team that you're there with them is really important. And they also see the standards that you execute those to. So if you're doing it right, it raises the bar for them around their personal targets.

You've described multiple examples of this. You've never been afraid to jump in the deep end, even in fields you knew very little about. What advice do you have for others facing steep learning curves or are considering career pivots?

So I share this frequently with my kids that we now live in an incredible time and I mentioned it earlier where it was a little bit harder when I made the jump into life sciences biotech after a career in tech. But we now live in a time where anything is learnable. The only limitation is the curiosity and willpower to do it. So just let curiosity drive you. And yes, now the Internet exists. Everything is discoverable. Now we have incredible AI tools that can be your coach along the way to be able to learn and develop new skills. And I think a key one of those is AI itself. I think you're going to see differentiation between people who are on the outside and not engaged and people who are on the inside and using the tools and really pushing what's possible. And Jensen Huang recently had a quote that really resonated for me. It was asked, is AI going to take away people's jobs? And his take was no. Somebody using AI has the biggest potential to take away your job unless you're there and leaning in and continuing to learn. So my encouragement for everyone, for my children, my son just graduated college last year and my encouragement to him was, congratulations on graduation. You finished the first chapter, now you've got the rest of the novel ahead and you need to be continually learning as you go. Your education is a milestone step along the way it's now lifetime learning that you're engaging in.

I think of it as there are certain things that humans are really good at, and there's certain things that computers are really good at. And if you can figure out the things a computer is better than you at and use a computer to help you do those things and then focus your time and energy and all the things a computer can really replace, that's a very powerful combination.

We're on the verge of every person has potentially superhuman skills, but it's incumbent on you to figure out and learn how to adopt those.

So the last question I'll ask you, Jeff, is what advice would you give to founders who want to build not just a successful company, but one that truly changed the world?

I think a huge element of it, and there's some themes that we've already touched on, is really being clear on your mission of what are you trying to accomplish and then what's the culture that you're going to build to get there. And if you look at the biggest differentiators of companies that are successful, I think it ultimately comes down to the culture, which is a combination of values and then execution of those values. So yes, you need to have the right technology, you need to have the right business model to get there, but ultimately it's the people and the culture that you're building. And that's something that we really look deeply at. So our framework when we look at startups is team technology, timing and then a bar of impact. And the team really is about the culture and values and focus on building that.

Jeff, this has been great. Highly insightful. Thank you for sharing your story, sharing all the insights that you've learned along the way, and for making a big difference in this world. So truly appreciate it.

Thank you very much, Alex. It's been a pleasure and an honor.

Thanks for listening. We hope you enjoyed this episode. Please visit our website@insightfulinvestor.org to access past shows and learn more about our podcast. If you have questions, feel free to email us@info insightfulinvestor.org and if you enjoyed the discussion, please subscribe to this podcast to ensure you don't miss future episodes. And don't forget to forward today's conversation to others you think would enjoy listening. This podcast is provided for informational purposes on only and should not be relied upon as legal, business, investment or tax advice. All opinions expressed by podcast participants are solely their own opinions and do not necessarily reflect the opinions of evoke advisors, their affiliates or companies featured. Due to industry regulations. Participants on this podcast are instructed not to make specific trade recommendations nor reference past or potential profits, and listeners are remaining reminded that securities trading, commodity trading, and alternative investments are complex and carry a risk of substantial losses. As such, they are not suitable for all investors. Listeners should be aware that guests featured on the Insightful Investor may have current or past associations with Evoke Advisors or the host, including as an investment manager of a private fund opportunity by Evoke or accessed through an affiliated Evoke Fund, or as a client. Participation as a guest on the podcast should not be perceived as an endorsement or testimonial with respect to Evoke advisors, the podcast host, or their services. Similarly, the inclusion of a guest on the podcast does not imply that Evoke Advisors or the host endorses the guest or any company with which they may be affiliated or employed. Evoke has neither paid nor received compensation from guests for their participation.