Imagine a world where fusion energy costs less than your monthly coffee habit, biotech startups design new materials the way software engineers write code, and climate tech companies scale as fast as mobile apps. Now imagine all of that sitting in university labs, stuck—brilliant ideas idling on the runway because the institutions around them were built for a different century.
That’s the tension Ramana Nanda dives into. And if you don’t have time to listen to the full episode, this is your guided tour through the ideas that matter most.
The Hidden Architecture of Innovation
Most people look at a breakthrough technology and see… the technology. Ramana sees the scaffolding around it: incentives, capital flows, regulatory environments, and the subtle social contracts that decide which ideas live.
He argues that deep tech’s biggest bottlenecks aren’t scientific—they’re structural. We know how to push electrons, fold proteins, and fuse atoms. What we don’t know is how to build an institutional system that funds, tests, and scales those things with the speed and precision of venture-backed software.
Deep tech today is like a sports car stuck in first gear—not because the engine lacks power, but because the gearbox belongs to a tractor.
A Career Built on Following the Frictions
Ramana’s journey reads like someone rewiring their own mental operating system. Born in India, raised in the UK, trained in chemical engineering, then abruptly pulled toward economics—not because he loved equations, but because he wanted to understand why good ideas fail.
Consulting in London and New York exposed him to a truth that would define his career: capital allocation quietly determines the future. Who gets funded, when, and under what conditions can accelerate or erase entire industries.
At MIT and later Harvard Business School, Ramana sharpened that question into a discipline: entrepreneurial finance—how money shapes innovation.
His move to Imperial College London marks his next chapter: building the Institute for Deep Tech Entrepreneurship, a place laser-focused on the hardest part of innovation today.
Why Software Moves Fast and Deep Tech Moves Like Continental Drift
In the early 2000s, AWS quietly detonated a bomb under the cost structure of startups. Suddenly, founders could run hundreds of experiments for pocket change. Learning cycles collapsed. Risk plummeted. Venture capital adapted.
Deep tech never got its AWS moment.
Imagine two founders:
Founder A: a software entrepreneur who can test ten hypotheses before lunch.
Founder B: a fusion founder whose “experiment” means installing a new neutron detector, waiting three weeks, and burning a million dollars.
These two live in different universes.
Ramana explains why this matters:
Software has short, cheap learning cycles.
Deep tech has long, expensive learning cycles.
Venture capital is optimized for the former.
Societal challenges increasingly depend on the latter.
The mismatch is the story of our time: we expect deep tech to behave like SaaS, then scold it when it doesn’t.
Europe: World-Class Science, Amateur-Level Commercialization
Ramana says the quiet part out loud: Europe produces phenomenal science but often suffocates commercialization.
A few culprits:
Pricing controls: great for consumers, terrible for early-stage frontier tech that needs premium early adopters.
Risk-averse procurement: governments act like referees, not customers.
Cultural discomfort with wealth creation: universities want impact without the optics of financial upside.
Europe has more Nobel-level scientists than early customers who will pay for their inventions.
The US, by contrast, deploys NASA, DARPA, the NIH, and the DoD as active buyers of frontier tech. Capital follows demand. Markets form. Scale happens.
Ramana frames it elegantly:
Innovation requires customers long before it requires markets.
Lab Success vs Market Success: The Gap That Swallows Startups
One of Ramana’s most useful distinctions is between two types of proof:
Proof of Concept: It works in controlled settings.
Proof of Value: Someone will pay real money for it.
Most deep-tech teams treat these as synonymous. They are not. They’re not even close.
A battery that works at 200ml in a lab often collapses when you need 200,000 liters of production. A biotech pathway that looks elegant in a paper turns chaotic in a bioreactor. A new energy source that works on a test rig may fail under grid-level conditions.
Scaling breaks things you didn’t know could break.
The job of early deep-tech support isn’t to make the science “more scientific.” It’s to compress the path to learning:
Which features customers care about
What they’ll pay
What the cost curve must look like
Which beachhead markets make sense
And which don’t
Most startups die not because the science fails—but because they wander into the wrong early market.
Universities: The Underused Superpower in Deep Tech
Universities love talking about impact. But their structures often block the very people trying to create it.
Ramana highlights three leverage points:
1. Equity-light tech transfer
Taking 20% of a startup “because policy” is a tax on ambition. MIT understood this early. Others are catching up—but slowly.
2. Non-dilutive internal funding
Scientists don’t need millions to explore commercial pathways; they need enough runway to learn. Universities can provide that.
3. Time as capital
If an academic can’t leave the lab bench to build a company, the company never forms. Sabbaticals for founders could transform tech-transfer success rates.
Ramana’s point is subtle:
Use institutional design to let the right 1% of academics become founder-builders without derailing their scientific careers.
Government as a Customer, Not a Cheerleader
Ramana makes a razor-sharp observation:
Procurement beats policy.
If governments simply buy early versions of climate, bio, and energy tech, everything else—investment, talent, momentum—follows.
The US does this instinctively. Europe intellectualizes it. Asia industrializes it. The UK… debates it.
Buying power determines which technologies escape the lab.
If He Could Build a New Institution from Scratch
Ramana dreams of a hybrid model:
Philanthropy to shoulder early scientific risk
Patient capital to fund slow, hard learning
Venture discipline for scale-up
A structured path from academic insight to industrial reality
Think of it as an “innovation middle layer”—something between universities and VC that treats deep tech’s unique physics as a feature, not a bug.
If Bell Labs and Y Combinator had a child—and that child had a sovereign wealth fund for a trust fund—that’s roughly the shape.
What Founders Get Wrong About Term Sheets
A quick tactical insight: founders obsess over the valuation of this round. Investors obsess over their ownership at exit.
Ramana argues founders should reverse their perspective:
Map backwards from a realistic exit
Decide what ownership they need to stay motivated
Choose funding partners and pacing that protect that curve
Otherwise, you win the battle (a nice valuation today) and lose the war (a cap table that strangles you later).
What Excites Him Most About the Future
Two themes lit him up:
1. AI as a scientific co-pilot
A future where researchers run thousands of virtual experiments before touching a pipette. If AWS changed software, AI may do the same for science.
2. Fusion and cheap energy
A world where abundant clean energy unshackles manufacturing, materials science, carbon removal, and entire industries we can’t yet name.
Energy abundance isn’t just a technology shift—it’s a civilizational one.
The Takeaway: Deep Tech Needs New Institutions, Not New Slogans
Ramana’s core message lands with quiet force:
We don’t lack scientific ambition. We lack the institutional imagination to support it.
If we build systems designed for deep tech instead of treating it like broken software, more frontier ideas will cross the chasm. Climate solutions scale. Biotech breakthroughs commercialize. Hard problems become solvable.
Deep tech isn’t slow because the scientists lack brilliance.
It’s slow because the world around them hasn’t been redesigned—yet.
👂🎧 Watch, listen, and follow on your favorite platform: https://tr.ee/S2ayrbx_fL
🙏 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast
Chapters:
00:01 – Introduction & Deep Tech Framing
00:49 – Early Life & Career Origins
01:40 – Shift from Engineering to Economics
03:23 – Entrepreneurial Finance Journey
05:02 – Europe vs. U.S. Commercialization Gap
09:11 – Founding Imperial’s Deep Tech Institute
12:54 – AWS and the Cost of Experimentation
16:42 – Deep Tech’s Two Learning Curves
18:27 – Government as Early Customer
20:54 – Proof of Concept vs. Proof of Value
22:07 – Choosing the Right Early Customers
29:18 – Sector Lessons: Energy, Biofuels, Tesla
32:09 – University Incentives & Academic Founders
35:10 – Tech Transfer Models: MIT vs. Imperial
40:11 – Underused Policy Levers
44:26 – AI’s Impact on Science & Labor
47:43 – Future of Scientific Productivity
Transcript:
Brian Bell (00:01:05): Hey everyone, welcome back to the Ignite Podcast. Today, we’re thrilled to have Professor Ramana Nanda on the mic. He’s the Professor of Entrepreneurial Finance at Imperial College London, academic lead of the Institute for Deep Tech Entrepreneurship, and a key thinker at the intersection of science, venture capital, and institutional innovation. Over his academic and advisory career, he studied how deep tech can bridge the gap between lab and market, how capital markets must evolve to support frontier science, and how universities and policy can play catalytic roles in the innovation ecosystem. In our conversation today, we’ll trace his origin story, walk through his research and institutional work, and explore his views on what is next in deep tech and capital. Romano, thanks for coming on. Thanks so much for having me, Brian. So I’d love to kind of get your origin story.
Ramana Nanda (00:01:51): What’s your background? Yeah, so I grew up in India. Until I was 16, I then came here to the UK for high school and university. I worked between London and New York for some years in consulting and then went off to do a PhD at MIT, taught for several years at Harvard. In the last Six years moved back to the UK to be a professor at Imperial.
Brian Bell (00:02:14): Yeah, your accent’s hard to place. Somewhere between an American, British, and Indian accent.
Ramana Nanda (00:02:19): My family jokes that I have a NATO accent.
Brian Bell (00:02:22): Yeah, it’s like a NATO accent. Yeah, that’s a good description. So what initially drew you to economics and finance rather than staying purely scientific or engineering?
Ramana Nanda (00:02:30): Growing up in India, a lot of people in my generation aspired to be either engineers or doctors. And I actually had a similar aspiration. But when I was at high school, I was exposed to economics. I had a very inspiring teacher. And what I loved about economics, or at least the way it was taught to me, was the idea that incentives could shape human action and that you could use incentives to get people to do things that were more positive or indeed negative, but to use the power of incentives to shape human action. And that was something that was very important. New to me, but also very exciting to me and was what kind of drove me into trying to study economics, which I then did in undergrad.
Brian Bell (00:03:16): Were you more fascinated by kind of the behavioral side, like the micro side or the macro?
Ramana Nanda (00:03:21): Yeah, I’ve been always been more attracted to the more the micro side. It feels a little bit more tangible to me. Of course, the macro is very important as well. There’s large flows of money and people and talent across countries and very important policy that gets done at the macro level. But for me personally, the micro elements were more.
Brian Bell (00:03:41): So after Cambridge, you worked in consulting in a few locations before going to MIT for your PhD. How did that, you know, kind of gap year, if you will, of professional experience inform your research questions later?
Ramana Nanda (00:03:52): It was four years that I spent across London and New York. I think I got a real appreciation for the role of, so it was a consulting firm, Oliver Wyman, focused on financial services. And I got a real appreciation for the importance of finance in the global economy. And that was, I think, What then ultimately shaped my research interests towards entrepreneurial finance, because I had this sense that entrepreneurship, as I was studying it, was so fundamentally important for bringing new technologies to market. All the fundamental innovations that we’ve seen probably in the last four to five decades have all got to market through new companies. It’s also so important for, you know, employment growth, productivity growth, but the role of finance in enabling or hindering entrepreneurship was something that I became very interested in both while I was working and then ultimately as I went into graduate school.
Brian Bell (00:04:49): Really fascinating, you know, you’ve kind of been across the, you know, two of the largest economies, both in Europe and the U.S. And yeah, I would agree innovation, economics, finance kind of drives a lot of the prosperity we see. But Europe sometimes gets a bad rap on this, on the kind of the innovation front. Why do you think that is and how have they changed in the last five or 10 years?
Ramana Nanda (00:05:08): Yeah, you know, I think Europe is associated with great science. The UK in particular ranks very high on the list of universities, top 10 universities by some rankings always have four UK universities in the mix. But in terms of commercializing that and turning it into companies, I think Europe has always lagged. I think there’s an element of that that has to do with the capital availability and the large pools of capital that have been present in the US that have enabled the growth of those kinds of companies. But there’s also, I think, a sense that perhaps people don’t take the risk that they would take in the US, in Europe. I think it’s a more complex story, to be honest. I feel like just risk appetite itself is probably not enough to explain it. My view is that there is also Sometimes in Europe, a bit of an aversion towards wealth creation. And if you are going to eat the rich mentality in Europe. Yeah, it’s not celebrated in quite the same way as it is in the US. Right. And obviously no one wakes up in the morning wanting to take risk. They take it. if they can get the reward and so i feel like the the fact that reward isn’t necessarily as celebrated may contribute a little bit to the to the less risk that is taken by entrepreneurs yeah you know america is an interesting experiment right
Brian Bell (00:06:31): if you stop and think about it i mean we’re kind of founded venture capital in a way you know columbus’s journey was sort of like like a venture capital deal right from from the King of Spain and successive waves of capitalism, good and bad, have sort of defined the American experiment. And, you know, it kind of leads to like this interesting cultural, economic, social selectivity and bias in the American society, which is that everybody that came here came here for a better life and for opportunity. Right. So it’s kind of it’s in the ethos. You know, that’s a big risk to move across the Atlantic Ocean or Pacific Ocean and try to improve your life.
Ramana Nanda (00:07:10): Absolutely. I mean, it’s, you know, these are all cliches. But in the UK, I think if you want to compliment someone, you say they’re really clever, you sort of compliment their intellect. I think in the US, if you want to compliment someone, you talk about the fact that they’ve been really successful. And these small differences that show up, I think are emblematic of these bigger differences about what is held up as aspirational in societies. You know, what you were talking about, being founded on venture capital. My former colleague at HBS, Tom Nicholas, has written a great book on the origins of VC and how you can trace back the origins to the wailing voyages that existed back in the time in Boston.
Brian Bell (00:07:53): I’m familiar with the book. I love an intro. He’d be an excellent podcast guest in the future. I’ll be delighted to do that introduction. Yeah. I love having authors and professors on because you guys think so deeply about topics. And podcasts are really kind of almost like little summaries of topics, you know, little conversations on topics. And those conversations are way deeper with authors and professors or experts, you know, generally. Yeah.
Ramana Nanda (00:08:17): Yeah, we often take a, we work at a level of abstraction, which can be interesting in these kinds of conversations. Yeah. Provides a different type of analysis or perspective.
Brian Bell (00:08:28): Yeah, I think in a different, like a different universe, I’m definitely a professor. I kind of miss my calling probably, but I like what I do. So tell us about the founding of the Institute for Deep Tech Entrepreneurship. What is that? What was the vision? How did that all come together?
Ramana Nanda (00:08:39): Absolutely. So as I said, I was at Harvard for many years from 2007 to My wife and I both have family here in the UK and we both did our high school and undergrad here actually in the UK. And so we had been exploring the possibility of coming back and spending time here in the UK. It was through conversations that this interesting opportunity came up at Imperial. Imperial, as you might know, is a very interesting university. It only has four faculties, engineering, natural sciences, medicine, and the business school. Very much a STEM university. It’s a STEMB, your prototypical STEMB university. What’s interesting about that is that I had, as I mentioned, been very interested in the matching of capital and ideas and thinking about the frictions that exist in commercializing ideas that are based on fundamental science that have both technical and market risk. And I’d been interested in, you know, I have a set of ideas about what these challenges were. And I was interested in thinking about how I could actually play a role in fixing it while sitting in the university. And Imperial was very interesting because they gave me the opportunity to pull together this Institute, gave some seed funding to set it up. And it was right actually over COVID, which was particularly Interesting and challenging, but also from a university standpoint, really, I felt a mark that they took it very seriously at times when lots of other places were pulling back funding. They were willing to make the investment in letting me set this up. And the goal of the institute is really quite simple. There’s a view that increasingly a lot of the challenges that we face in the world will have solutions that were based on fundamental science and engineering. A lot of those are going to emerge from universities, but universities around the world, not just here in the UK, in the US as well, really have a challenge in translating the great research that’s coming out into commercially relevant. And that is something that I think is one of these grand challenges in its own right. If we want to solve the other grand challenges that exist in the world, we’re going to have to be able to fix this. And it’s a multifaceted problem. It requires risk capital to engage. It requires universities to engage. It requires government to engage. And part of what we want to do is to study these issues, to help educate people, to convene, to run workshops. And then on our own end at Imperial to help make that process a little bit smoother. And so that’s what we’ve been doing over the last four or five years.
Brian Bell (00:11:20): And what do you kind of see as the kind of major systemic challenges for optimizing this in the current kind of zeitgeist?
Ramana Nanda (00:11:26): Yeah, I think, you know, one of the things that I reflect a lot on is that around 2005, 2006, you were at Amazon, so you know, advent of cloud computing and what that did how it unlocked entrepreneurship in the
Brian Bell (00:11:40): tech industry and angel investing and angel exactly because now you know a couple guys with a couple people with a laptop could you know exactly get a get a working application in somebody’s hands very quickly exactly and so you know it’s hard to
Ramana Nanda (00:11:55): even imagine the the fact that people had to pay you know tens of millions of dollars to buy that server equipment before AWS. Just to get a hello world on a screen. Exactly. What a sea change. Indeed. And so one of the things that, you know, sort of two or three orders of magnitude was the drop in the cost of starting companies. At an intellectual level, at a conceptual level, the way I think about it is that in some sense, what we’re doing as we’re building ventures is running a series of experiments over time. And each of those experiments are generating information that allow us to make go-no-go decisions. And effectively- Almost like the scientific method applied to business. Indeed. And in fact, if one was to think about this in a scientific sense, there’s lots of questions one could ask. What experiment should you run? In what sequence should you generate that information? Who do you need to convince? Sometimes the entrepreneurs should really remember that they need to convince investors. And convincing investors is different from convincing themselves. And the kinds of experiments that they would need to run to convince people who are not themselves might be different from the ones that they run to convince themselves. And so the idea is that the cost of these early experiments just dropped substantially with the advent of cloud computing it allowed all kinds of high option value investments to get started and it as you mentioned unlocked a whole new class of investors who were able to money in meaningful amounts into these early experiments that shifted in many ways the arc of of where entrepreneurship and capital was going towards technologies where you could learn quickly and cheaply about the potential and in some sense deep tech and You know, drug development, nuclear technologies, semiconductors, robotics, all of these are instantiations of ventures where you don’t learn quickly and cheaply about the ultimate potential. And it makes it much harder to finance these kinds of technologies. So I kind of, if you were to think about on the X axis cost and on the Y axis cost, A lot of tech companies achieve that value inflection much more quickly than in a deep tech. You want to think of, you know, in tech, that curve, that S curve takes off much more quickly. it takes much longer for that S curve to take off. And obviously that makes it much harder for investors to eat that lack of inflection in value, which stacks up in terms of time and money. And so a lot of what we think about is how we can compress that curve in terms of making it faster to be able to learn, to get more precise signals of ultimate performance and the degree to which we can take some of that and pull it into non-dilutive pools of capital where what comes out the other end is then closer to those points of inflection that can become more venturable and more possible to finance for those who are in the business of returning capital, which is exactly what you should be doing because you have a fiduciary responsibility to your limited partners to generate returns. And so I see, in some sense, the role that we can play at universities and other organizations where there isn’t that profit motive to be able to internalize some of those costs, the risks, and to deliver to the to the market, more investable technologies that can go off and meet that potential.
Brian Bell (00:15:31): We have a phrase in venture capital, hardware is hard. And basically what you’re describing is government funded non-dilutive grants to try to get a proof of concept in the market that can prove out that there’s a business here, which de-risks it, especially for getting deep tech technologies in the market. You know, there’s been a kind of a rise in deep tech focus venture capital, I’ve noticed probably in the last 10 years here in the U.S. And you spend a lot of time thinking about this from your seat in the U.K. because now you have the U.K., you have the EU, right, after Brexit. That’s two different entities now. And then you have U.S. How do you think about kind of the flow of capital and ideas and science and business between these three entities?
Ramana Nanda (00:16:15): Yeah, as I was mentioning, I think the science is extremely strong in the U.K. and in Europe. the conversion of that science into big businesses has had a less strong track record. I would say all over the world, for all the reasons that we’ve talked about, these are just harder businesses to build. But obviously, in the US, we have had experiences of that happening more. The government plays, as you were mentioning, a very important role as a customer. The military has played a very important role in the US. NASA has played a very important role in space exploration. In the US, the fact that if you pass FDA trials, there will be reimbursement at a high level for on patent drugs has meant that effectively, the US government, US consumers are paying a high level of price for approved drugs. All of these things, I think, have contributed to the very robust life science industry that exists in the U.S. and the kind of the deeper tech that exists. In Europe, there’s been more of a negotiation around cost on the part of government, which I think has made it harder for businesses to truly be able to take off because they’re not getting the the revenue that is commensurate with the risk that they’re taking early on in their lives. I think one thing that you mentioned that I would love to pick up on is that you talked about this proof of concept funding from government. and we like to make this distinction between proof of concept and proof of commercial value because there’s a lot of early stage ideas that show that there is a proof of concept it works in the lab as a early stage demonstration of a technology but it hasn’t really moved to the point where it’s possible to demonstrate that there is a real customer out there who’s willing to pay for that product and that you can make that product at work and you can generate a substantial amount of revenue. And so we like to distinguish between the notion of proof of concept and proof of commercial value. A lot of people end up focusing on the proof of concept, which is necessary, but not sufficient to actually get the venture funding, because if there isn’t proof of commercial value, it’s not going to get backed.
Brian Bell (00:18:39): Yeah, I think in your work, you talk about this with the twin frictions of technology risk and market risk. How do you think about quantifying and structuring capital to manage these two frictions?
Ramana Nanda (00:18:50): Yeah, the fact that in deep tech, you still have a lot of technology risk present. I think makes these businesses much harder to build. And there’s an interaction between the two. So if I’m building a particular technology that it might have two or three potential use cases, the technical benchmarks I need to hit might be different for each of these use cases. And so I have to think carefully about which use case do I want to build towards? Is there a market there? And then what technical experiments do I need to run early in the lab that provide evidence that this might actually work at scale. And I think that’s what makes it complicated because if sometimes I might have a view that this works, but when I scale it up, It doesn’t work at the economics that I thought it was. And we’ve seen this a lot in energy technologies in particular, where we have a view that things are working at a lab scale. But once we actually think about the overall lifetime value of a battery or a biofuel plant or a solar cell, some of the new technologies that are out there are just not able to compete, not because they don’t work, small scale in a very productive way but once you’ve actually taken it to the utility scale and think about the entire lifetime cost and value of these it becomes much harder to justify them from an economic perspective yeah we have
Brian Bell (00:20:13): another phrase in venture which is there’s no bad idea it’s just bad timing and so i think what you’re describing is you’re too early to market right this technology is not ready to scale it’s going to need a little bit more time to incubate and figure out all the the scaling mechanisms and the technology stack to make it viable.
Ramana Nanda (00:20:31): Yes, exactly. And, you know, in some cases like solar, the dollar per watt has fallen substantially. And in part that’s happened because the Chinese government was able to subsidize giga factories of solar panels, which were able to bring the
Brian Bell (00:20:45): the cost down substantially. And why do you think they did that? Do you think it was like kind of a twofold reason? One is, hey, let’s create cheap energy. We’re growing very fast. Our GDP is growing fast. We need cheap energy. But two is, hey, there’s this burgeoning technology, technological field we can dominate. If we pour a lot of resources in, we can become the cost and quality leader and kind of own the market.
Ramana Nanda (00:21:06): Yeah, listen, I don’t know enough about their motivations. I know that they have innovated incredibly in solar, in batteries, in nuclear. So It is true that there is tons of coal-fired power plants going up every week to meet the energy demands, but they’re also innovating in all of these other areas. And I think the combination of pollution and the real costs associated with pollution and probably the desire to have technical leadership and, frankly, the balance sheet to be able to go off and subsidize some of these technologies and get them to become global leaders as companies has certainly
Brian Bell (00:21:45): worked in their favor. So from your vantage point, how has venture capital and LP attitude shifted or not towards deep tech and frontier science?
Ramana Nanda (00:21:54): Yeah, you know, I think there’s a recognition that we need some of these big problems, whether they’re around energy security, healthcare, defense, many of these are problems that can’t be solved with information technology alone or AI alone. We need new innovations in deep tech. And I think there’s been this confluence of the realization that these are big problems. And if one can solve these big problems and get the business model right, they are big, meaningful companies to be built that can generate venture returns. Together with, in the same way that AWS lowered the cost of building tech companies. I think AI is getting to a point where the predictions of being able to think about new material properties, new drug candidates, projecting how the inside of nuclear reactors will work, have allowed that those early experiments to become much more predictive and much more informative. And I think that those two trends together are going to make it such that there will be a greater interest in deep tech and building companies that can kind of solve these big problems. But indeed, as you point out, there are certain structural factors about the venture industry. The fact that it is 10-year funds where oftentimes If you have returns that compound slowly over a 15-year period, you can have a fund that looks very, very nice, but over a four or five-year period, when you’re looking to raise your next fund, it’s not quite showing the returns that you need to show to be able to raise the next fund. And so I think an understanding by LPs that there are these models that can work. Sometimes family offices that are not necessarily tied to these, you know, 10-year fund structures are getting involved in making direct investments. I think all of these are changing a little bit the way people view it. But of course, relative to tech, it is still a small asset class. I think when I looked at the growth in venture from 2010 to 2020, there was a five-fold increase in the overall global dollars that were spent in venture. And almost all of that increase came from consumer internet, enterprise software, financial services, healthcare, IT. Very little actually came from all of these other sectors. My hope is that that will change, but we’ll see.
Brian Bell (00:24:24): Yeah. Well, can you share a few case studies or stories from ventures affiliated with your institute or more broadly that illustrate how deep tech is scaling or failing?
Ramana Nanda (00:24:34): Yeah, you know, I think one of the big challenges, well, I’ll sort of talk about two sets of challenges. One has to do with what I talked about, which is that the economics that you project at small scale don’t necessarily translate at scale. And we saw that this happened a lot in the biofuels industry. There was a lot of excitement around biofuels and a lot of the small biofuels plants worked incredibly well. But it just so turned out that as they got scaled up, the ability to predict how well they would work at scale and how they would work relative to oil was not as good. And so I think the two lessons that I took from that, one, when you’re competing against a commodity, you have to be really, really, really certain that you can be probably several orders of magnitude better in performance at the lab scale. And it’s only then that you’re likely to be able to compete. It’s different. If you’re not able to do that, then you want to think about what kinds of categories can you enter. So if we think about Tesla, they started off with luxury cars precisely because you can get higher margins in that category, and then slowly come down the cost curve. As I mentioned in therapeutics, the fact that you have patent protection allows you to charge these supernormal profits for a period of time, which can allow you to recoup those costs. So thinking about who are the early customers who are willing to pay for version 1.0 of the technology, I think is a very important question that is worth considering. The other thing that I sort of reflect on, particularly around industrial sales, B2B sales, these are long sales cycles. Oftentimes, if your pitch is, I can do this for a little bit cheaper or a little bit faster, pretty risk-averse procurement teams are just not going to be interested in adopting the technology. And so thinking about where you’re selling, why it’s going to be a compelling sale, what the And also how your first set of clients can be good, referenceable clients for your next set of clients is really important to be able to cross that chasm of early adoption, translating into the mainstream.
Brian Bell (00:26:50): So it’s interesting because on the university side, often labs or scientists are not really incentivized to think about commercialization. How do you align these incentives towards a translation to commercialization?
Ramana Nanda (00:27:01): Yeah, this is a big one, I think. And I think about it a lot. You know, not all academics are interested in commercialization. Many have come to a university because they’re interested in research or they’re interested in teaching. And my view is very much that we shouldn’t be trying to force people to do things that they haven’t chosen to do. However, if they’re interested in commercialization, and sometimes people come to this through a roundabout view, they see a colleague who’s been involved in it or a student of theirs has been involved in it and they express an interest in being engaged, then the question is, how can we enable them to reach their potential? And so I feel like a lot of times the question is not, how do you change incentives? Because it’s hard. You know, people are incentivized by the system that exists. But for the set of people who are interested in commercialization, how can we make things as smooth as possible? And there, I think there’s a lot that universities can do, you know, in general. tech transfer offices are set up as cost centers. We’re not thinking of tech transfer as this great engine that might exist to generate impact and to generate return for the world and for the university. It’s typically thought of as a cost center where costs have to be minimized. And then going back to all of these things around incentives, if that’s what you’re trying to prioritize, then that’s what And so I think changing the mindset of university administrators, not just the tech transfer offices, to think about the possibility of generating impact and return and how one can incentivize that through a whole set of things, whether it is allowing academics to take leave of absence for a couple of years to explore these other avenues of impact that they might be able to have, or rethinking the way in which tech transfer offices are incentivized and organized. All of those, I think, are things that probably we need to be thinking about as universities if we are to take seriously the job of taking our science and turning them into impact.
Brian Bell (00:29:13): Yeah. Are there any universities out there you think do a good job of this?
Ramana Nanda (00:29:17): MIT is one that comes to mind. They certainly have very much of an impact-oriented approach to commercialization. I’d like to think that Imperial is another. We’ve definitely made big strides in how we both think about and implement the set of offerings that we have for our colleagues in terms of commercializing So we take a relatively low equity position in any companies that come out of Imperial. And we tend to offer quite a lot of things. Again, the Deep Tech Institute is an example. We will offer at times up to three or 500,000 pounds of non-additionally dilutive funding together with a whole suite of other support services around thinking about customer discovery. around the design of experiments connecting to investors and so on that we bring to bear for the technologies that we think have very high potential, but a very low likelihood of working at this moment in time. And the goal is to try to increase those odds of success before we put them out into the world.
Brian Bell (00:30:22): so we have this phrase in tech which is you know specifically product management because that’s what i did most of my career is you know you start with the customer need and you work your way back to the tech you start with the tech and work your way out that’s a very waterfall process so what degree do you think universities like mit or imperial should keep in mind what the world needs what customers actually need and work their way back to research versus hey let’s just start with scientific discoveries just pure scientific discoveries and then work our way out to the commercialization
Ramana Nanda (00:30:50): Look, I think there’s a place for both. In some sense, that eureka moment cannot exist if you’re optimizing for something, a specific need that you can see today. And sometimes you do get these fundamental breakthroughs that you don’t necessarily know what they’re going to be applied for. Apparently, Edison, when he built the light bulb, someone asked him, you know, what good is this for? And they couldn’t you know, comprehend what use it could be. It turned out that the telephone, you know, they had this invention. They didn’t quite know what use case it could be. And one of the use cases was to, you know, to take music from Carnegie Hall and have a one-way, you know, projection of this music into people’s homes. And so it is true that I think these fundamental breakthroughs can end up being pretty transformative. And so we want to have a space for that. But at the same time, I agree. The sort of DARPA model of innovation that people often talk about, which is there’s a clear use case. We need to build something. Some of that might be engineering. Some of it might be fundamental science. Let’s pull the team together and actually solve that problem. And I think that there is definitely scope for that. And one of the things that we are doing at Imperial, we have these thematic horizontals that are around certain areas around human and artificial intelligence, around space and security, around healthcare and around climate and sustainability. thematic areas are ones that we’re going to be trying to collaborate with at the Deep Tech Institute to think about particular challenges that exist in the world that need to get solved and to think and to bring to bear the consumer or the investor perspective on what are the kinds of things that people are looking for and then try to build back and try to shape what kind of research might. So I would like to think that both of these are areas that are, you know, modes of innovation that are that are important, but it’s important to get the balance right. I think you can’t have exclusively one or the other. And historically, universities have been probably much more focused on the kind of the technology push. We’ve come up with a great invention. Let’s figure out what use it can be.
Brian Bell (00:33:01): Let’s talk a little bit about policy, regulation, public capital. What policy levers do you believe are most underused right now to accelerate deep tech?
Ramana Nanda (00:33:08): I think the role of government as a customer is something that is appreciated, but perhaps not appreciated as much as it could. We’ve seen this time and again where there is a commitment to buy a particular product at a certain price point. That takes out, as I said, because the technical risk and the market risk exist together, if you can have a price point that you can target and a particular type of quality of technology that is also specified, that makes it much easier to build the first iteration of something that is truly transformational. And we saw that this was the case in the semiconductor industry in the US going back, the automobile industry, There were standards that were set by the government in terms of procurement in the World War. World War I, we see that. We often see it in the therapeutics biotech industry. And we saw it again with the solar. industry where they will feed in tariffs and people were willing to pay a higher price for clean electrons relative to dirty electrons. And that led to the growth of the solar industry. So I think, you know, it’s very common in the military to do this. We’ve seen the procurement works in other parts of government. The big challenge, I think, is that oftentimes governments are happy to run small-scale pilot procurement programs, but they don’t end up then providing the big, bold scale-up that happens next. And if there isn’t the promise of some big purchase that comes down the road, then those pilots become less attractive for companies. And I think is one of the reasons why In general, venture, you know, thinks of these as, yeah, good in theory, but in practice, turns out that it’s not, you know, that you go through a lot of hassle through that procurement process and then ends up not translating anything into anything big.
Brian Bell (00:35:02): I’d love to talk about kind of your outlook over the next five to 10 years. What are you excited about? Where do you see changes coming? New classes, vehicles, capital policy, regulatory environments, geopolitical stuff. Anything, any way you want to take it over the next, you know, five to 10 years?
Ramana Nanda (00:35:18): Well, look, I think that if there’s one thing for sure, there’s going to be a period of a lot of turbulence. AI itself is changing the way we work.
Brian Bell (00:35:27): It’s starting to do science and math. Science and math. Engineering, drug discovery, protein discovery, you know. Whole, you know, whole body emulated, you know, digital twins of human bodies. So you know exactly how compounds interact with our complex human physiology.
Ramana Nanda (00:35:44): Exactly. So I think both as the way in which we use it, the way in which it impacts our day to day life as consumers, but also the way in which it impacts innovation itself is fundamentally changing. In that sense, it’s a very exciting time. I mean, obviously, people can be very scared about what happens, what are the implications to our livelihoods, to jobs, and so on and so forth.
Brian Bell (00:36:10): I mean, what will we do when we move off the farm or out of the factory? I have no idea. This is something we’ve always been afraid of every 50 or 100 years.
Ramana Nanda (00:36:20): There’s a circularity to it, because if we don’t do anything, then How do we earn any money to spend on the companies that will need to sell to people?
Brian Bell (00:36:30): Let’s do podcasts and create, you know, TikTok videos and stuff.
Ramana Nanda (00:36:35): There’s a universal income for everybody.
Brian Bell (00:36:37): Maybe that’s maybe that’s the future. We all just, you know, I heard a really good on the Moonshots podcast. They were talking about universal basic income. And they said that it tends not to work because people become depressed and, you know, turn to alcohol and drugs and stuff and become kind of lazy. But they said universal basic services is a better concept, meaning your food, clothing, shelter, your basic human needs, health care are provided. But then you still have to get up, get out and get something in the world if you want. You’re free to go create and build and do whatever you want, but you’re not free to just sit there and watch TV. I mean, you can because, you know, you basically, you know, in a way it’s almost like you’re in a college dormitory or like a assisted living facility or something like that where everything’s provided to you. But, you know, now it’s, you know, what do you want to do with your freedom?
Ramana Nanda (00:37:25): I think humans get utility from differences. I mean, it just seems like a fundamental human desire to feel like you are doing better than the person next door. Comparing yourself to others is a very natural trait. And it’s clear that whenever people get a bump in wealth or income, there’s a moment of which they feel really good about it, but then that becomes taken for granted. It’s always about the comparison to the next person. And I think that aspect is something that I wonder about this universal basic income is, you know, what will motivate people to want to do better than the next person and aspire to be, you know, the standout person in the room, as it were, and how will that be achieved and what will be the skill sets that they will need to do that? What training is required today to be able to be in a position to most utilize that when the time comes? These are all, I think, I guess, existential questions or philosophical questions. I don’t know the answer to it. It seems like AI is progressing at a pace at which so many things will be possible to do, thinking about what are the subjects to be studying today. question that I reflect on. I don’t have a particularly good answer.
Brian Bell (00:38:41): I’ll turn that around and make it more difficult for you. You know, imagine you have children entering college. I do. I do. I mean, yeah. Yeah. What do you advise them to study? I mean, I have six years before she enters college. So yeah, mine’s only two and a half years away. I just tell them to follow their interests, you know, and passions, like whatever is interesting to them. You know, I think that’s probably the most important motivating factor. You know, don’t try to, you know, back to your point about let’s not try to force research scientists to commercialize their discoveries. if they’re not interested in doing that. I kind of think about that a lot. My personal background, you know, I worked, you know, I grew up poor and worked really hard to get to Wall Street and all that stuff, CFA, finance, undergrad. And then I realized like, what, am I really interested in doing this for the rest of my life? You know, or am I just trying to be successful? I think it’s always important to kind of just study whatever you’re interested in. And then I think good things will happen over time. And it’s hard to see, it’s hard to connect the dots looking forward, I think.
Brian Bell (00:39:38): Absolutely.
Ramana Nanda (00:39:38): I mean, that was one of the reasons why I felt I wanted to go back to academia. and move away from consulting, which was pretty much the same sort of motivation, which was I was doing well, but I wasn’t necessarily feeling like I was truly enjoying myself. And so I found my calling in research and academia.
Brian Bell (00:39:55): So if you had to propose one moonshot deep tech area beyond AI to focus on.
Ramana Nanda (00:39:59): Yeah, I’ve been most excited about fusion energy recently, just because of the potential that it has for, you know, near infinite energy. And we clearly know that if we want to be building society that has higher levels of consumption, higher levels of satisfying the various needs that we have as humans, we need more electricity. And we’re seeing that already with AI, the amount of electricity consumption has been going through the roof. Nuclear fusion, as I understand it, has that potential, although there are some other people who feel like it will not be achieved at grid scale by the time our grandchildren turn old. And I think that’s the excitement for me is to try to understand whether there is anything that can be done to accelerate that process, what kind of discoveries are needed, what role can capital play, and how can we be smart about the kind of scientific experiments and the way in which we deploy capital to learn about these possibilities in a way that moves the needle in terms of trying to get grid-scale fusion energy. So yeah, to me, if that works, it will be a massive, massive transformation to our energy.
Brian Bell (00:41:10): Right, yeah. I mean, energy kind of drives all of our underlying economic activity, right? If you have cheap and abundant energy, lots of use cases are unlocked.
Ramana Nanda (00:41:20): Indeed, indeed. And if that can happen at scale, then as you said, the possibilities that can then be enabled by that are also boundless.
Brian Bell (00:41:30): Let’s wrap up with some rapid fire questions. What’s a trend that you believe is dramatically underestimated today by most investors? It could be in science, capital policy, ecosystems.
Ramana Nanda (00:41:40): Yeah, you know, I think one that has a lot of potential is the fact that a lot of family offices invest in new technologies, and on the other side, give philanthropically, but they don’t actually lose the power of these together in a way that is, I think, as synergistic as it could be. So they’re, in effect, proverbially investing in tobacco companies and then giving to lung cancer research, where they could, in theory, invest or give philanthropically to the translation of lung cancer technologies and then invest in the company that came out the door. Now, there are some challenges about that because it has the potential to run afoul of 1C3 rules. You could be seen to be self-dealing, subsidizing a company through your tax dollars. that you invested in for profit. But I think the general view of thinking about how to structure giving philanthropically to support the de-risking of technologies that then can be benefited for the for-profit side of the business is something that probably people can do some more deep thinking on and i would love to see more interesting models
Brian Bell (00:42:48): along that along that what’s a one institutional innovation not yet widely adopted you wish more universities or funders would try i think sabbatical for academics
Ramana Nanda (00:42:59): who are interested in trying to innovate and providing the support that they need to go off and explore and then come back. There are instances where this has been allowed on a sort of an ad hoc basis, but to Make it a policy, I think, would be really interesting to see what kind of innovations it unlocks.
Brian Bell (00:43:17): So if you were designing a deep tech focus investment fund from scratch, what would the structure look like in terms of capital sources, time horizon, governance, alignment with grants, et cetera?
Ramana Nanda (00:43:29): Yeah, this is an interesting one. I mean, obviously, there’s lots of different ideas around things that have worked and have not worked. One idea that has been proposed is to do a mega fund. where if you invest in science that has idiosyncratic risk, then if you invest in a large enough number of companies, one of them is definitely going to pay off. And what that allows you to do is to do a bit of financial engineering. So you can take on some debt, which would be sufficient to be paid back by the one company that worked out. And then the equity stack would take more risk than two or three or others kind of paid off. So I think thinking about how one can use financial engineering to one’s advantage is something that is very, very interesting. The part that’s hard about that is of course, in order to run a mega fund, governance becomes hard. You have to have, you know, how do you make sure that you invest prudently all of that large amount of money? My view is that there’s another institutional innovation that could be possible to do, which is to, as I said, to combine this sort of non-dilutive philanthropic part of the initial de-risking of a fund with a profit-seeking version of the fund. And so in effect, the early stage has a goal to just return its capital and just be self-sustaining forever. And so it funds, a lot of de-risking, that goes and gets funded on occasion by the profit-seeking side of the fund. And it pays back just enough to make sure that the early stage remains viable in the long period. And as a fund structure, that kind of internal accounting, I think, might be possible to generate venture returns for the profit-seeking side of the fund. and to be self-sustaining on the quote-unquote philanthropic side of the fund. And if it can be seeded and funded with these two different pools of capital, then both sides can kind of get there, can see success by their own metrics. I haven’t modeled it all out. But I think that there is something interesting there to think about.
Brian Bell (00:45:33): What’s a mistake that you often see among deep tech founders when negotiating term sheets for their rounds?
Ramana Nanda (00:45:39): Well, I think I’m not sure if it’s deep tech founders in general or just founders. Oftentimes there is a tendency to optimize too much for a given round without thinking about what the value of their ownership at the end will be. And I think thinking about each round as a one step in the overall journey can be very helpful. I’ve had many former students who have talked about the fact that they tried to over optimize on a given round, push the valuation too high in their favor, and then set up a treadmill for themselves that made it very, very hard to actually earn into the valuation, not only required for that round, but to have momentum going into the next round of financing. And so I think thinking about how one builds a company over the longer period and how one optimizes for what one needs to own at the end is something that I often speak to founders about to put in perspective what’s going on in one round relative to the entire journey.
Brian Bell (00:46:41): Yeah, good advice. So if you had unlimited time and resources for one bold initiative, what would it be?
Ramana Nanda (00:46:46): I’d love to be able to just scale up what I’m doing at a much, much bigger level. I think that there is this opportunity that we have to really unleash the potential of science and to rethink The capital models, the talent models that need to be brought to bear to match with these ideas to turn them into great companies. And we’re trying to learn that and implement it at the same time here at Imperial and then to help other people benefit from whatever it is that we’re doing. And I feel very fortunate to be in a position where I’m doing this and really, really enjoying myself. And so if I had more time and money, I’d probably put more of it into this project, which I’m super excited about. Well, thanks so much for coming on. Really enjoyed the conversation. Learned a lot. Thank you. Thank you for having me and look forward to following your podcasts.
