Transcription
Vinod, welcome to Sparks. I've been waiting for this conversation for a long time. Uh, you may not recall, but I have been to this office a few times. I've met you in different contexts, and it was good to see you at the AKA event as well.
I want to start with, uh, I'm in the Valley for the last few days. I, I come here often once or twice a year, but the pace seems very different now. There's so much action, and a lot of it seems to be concentrated in San Francisco. Is this, you've been in the Valley for the last 40 odd years. Does this feel different?
>> It definitely feels different.
>> There's a lot of activity, a lot of genuinely good startups, but also a lot of money.
>> Yeah.
>> And that combination has resulted in velocities that are very, very high.
>> Yeah.
In one of your talks, you have mentioned there, and I think it's, it's, I guess fairly obvious also, there are, there are only three, four very deep model companies now in, in the US at least, and then maybe there are an equal number in China. Do you think most of the innovation is concentrated inside these companies, and there's a delta between what the companies outside these companies are able to do?
>> No, I think there's a lot of innovation outside these companies. A lot going on, uh, at least in this area. Like, every day there's a new idea.
>> That's worth considering.
>> If not funding. Uh, result of which, you know, the way best way to understand Silicon Valley today is, of all the unicorns, 91% are in the Bay Area. The rest of the world combined accounts for 9%. Which is pretty stunning. The second highest.
>> This is all unicorns, not AI unicorns.
>> I, I think it refers to AI unicorns, but, uh, the second highest is New York with 2%. So it's even the number two is largely irrelevant.
>> Yeah.
>> And so the action is here, almost all the innovation is here. And see, uh, if you look at any AI company today, the engine of all AI companies, obviously, is the inferencing, which is powered by one of these these companies.
What is the nature of innovation outside? But I want to just, maybe I should want to touch upon your investment in OpenAI in 2019. It was, you have been known for contrarian bets, but this was way out there contrarian. If you recall your thinking and the discussion, what convinced you? It was not even clear where OpenAI was going to be a for-profit company, nonprofit yet. You wrote a pretty large check.
So, I wrote a check that was twice the size of the largest initial check I'd ever written.
>> Cuz I don't like writing huge checks. We like very small checks.
>> So, I'd never written in 40 years a check over $25 million initially. Later on, we'll add additional money.
>> And we wrote a $50 million check. And when we wrote the check, we also sent an apology letter to our LPs saying, "Sorry, we're making this investment. It doesn't make any logical sense.
>> It's a nonprofit. No product plan, no business plan. It's just a tech technology effort, but the idea was very simple."
>> So today, that letter is sort of a funny thing.
>> Yeah. Um, and I should probably tweet it. It was so funny nowadays in retrospect. Um, but what was very clear to me is when you get AI,
>> Mhm.
>> It'll be very consequential. So, it wasn't when we would get AI.
>> Mhm.
>> It was when we get it, whenever that is, whether it happens in two years or 10 years, didn't matter. Uh, it would be very consequential. So the bet was very much on the very large upside if you were successful, and that's our kind of bet.
>> Mhm.
>> I love making those bets.
And then each bet has to be, I guess, two parts. One is, you were convinced that AI, whenever it does happen, will be very consequential. But this particular company, because when you're dealing with so much uncertainty, almost no precedence, and almost no clarity about timelines, picking a team versus the other also has to, there's probably some method behind.
>> Absolutely. You judge the team. I thought this was a very good team.
>> We're very confident in the team.
>> Um, but there weren't also a lot of companies. Yeah.
>> There was Google.
>> Mhm.
>> Yeah.
>> Uh, OpenAI and BU, maybe in as a Chinese company. So there wasn't a lot of AI efforts. It's not like today.
And there's this whole thing about opportunity meeting a prepared mind. I know you've been talking about AI for a while, including, I think, 2012, you've been talking about AI doctors and so on. So, is it almost a case of you were looking for something like this, and you happened to eventually, you know, I assume you must have got.
>> I was absolutely looking for it. You know, there's some.
>> Important areas of technology that I tend to like to identify and say.
>> We should, if we find a great startup, invest in it.
>> Mhm.
>> By the way, same year, too. It was 2018 that we committed to OpenAI. It closed in early 2019.
>> Uh, but same year, we invested in Fusion.
>> Mhm.
>> Nobody was expecting Fusion in the next 30 years.
>> Yet we placed that bet, and I think it looks pretty good today. So, you know, I like unusual bets. We also placed a bet that year in, uh, public transit.
>> Which is also not considered a venture area.
>> So, um, I think around then, we also invested in hypersonic aircraft.
>> That was also unusual for a startup to be doing hypersonic aircraft. So, we, we're not afraid to place large bets. We tend to be cautious about how we guide them so they don't run out of money, because some of these areas become very difficult. But I'm always trying to say, what will be hugely disruptive 10 years from now that people aren't expecting? If everybody's expecting it, then it's not interesting. But it has to be at the sweet spot of people are not expecting it. But it should at least, I know, as on on average, should become real in the next 10 years, because if, let's say, OpenAI bet, if you were to place the same bet in 2009, the story will not look that great because of finite lifetime.
>> No, look, this is the hard part of venture investing is guessing when you will have breakthroughs, when will.
>> Large changes happen. So I'm always looking, like everything I look at is between 2030 and 2035, and say, what can be consequential?
And this, this philosophy as a venture, uh, um, capitalist, or I guess you would like to call it as a venture advisor, has been a work in progress. You know, you've kind of honed it over a period of time by seeing a lot of different patterns. Like, what is the backstory behind doing this way?
>> So, I like to differentiate from people who are investors.
>> You know, they have to run spreadsheets.
>> Yeah.
>> We never run a spreadsheet on rates of return.
>> Mhm.
>> So, rate of return calculations are just mostly misleading in our business.
>> For what we do.
>> Mhm.
>> Most other investment firms use it.
>> It's misleading because of the artificial mark-to-market that you.
>> Well, it's not artificial mark-to-market. Mark-to-market in private markets is always artificial.
>> That's not the real issue. What the real issue is, you can't predict these things.
>> You know, if I said to you, OpenAI will have no revenue in January of '23, but January of '26, it'll have $20 billion, you'd say I'm crazy.
>> So sometimes it's better to realize you can't forecast instead of trying to make a forecast and then using it as a basis for calculation.
>> Mhm.
>> You just say it's not possible to forecast. Or.
>> See, today is obviously very impressive. It can do a lot of things, and one can also argue that we are not fully exploiting what's possible in AI even today. But a lot of the AI investment, there are now, I think close to a trillion dollar of capex, which is being spent this year, and perhaps the next few years, every year. This also assuming some, this the scaling to continue for the next few years, is that you almost take it for granted now, we have hit an inflection point, the scaling will continue because it can also stop, the asymptote may play out, and this may have implications for the whole journey.
>> And if you look at the five years journey.
>> Could you have 10 times the amount of inferencing? Absolutely.
>> There's demand for that.
>> 10 times more intelligent? Like, what's the metric? 10 times more inferences.
>> Mhm.
>> Than we do today.
>> Mhm.
>> Could you see a 100 times? Possibly.
>> Mhm.
>> U depends on the price decline.
>> And you're not commenting on quantity of inference, not quality of inference.
>> No.
>> Yeah.
>> Uh, just the number.
>> Yeah.
>> Yeah.
>> Got it.
>> Sort of measure it in tokens produced. Now, the intelligence of each token will also go up.
>> Mhm.
>> Which is the quality aspect. So I suspect we'll see continuing increasing increases in the number of tokens, almost exponential.
>> Yeah.
>> We'll keep seeing, um, large but more modest increase in the intelligence per token.
>> Mhm.
>> Which is human intelligence delivered.
>> Yeah. And we'll see pretty rapidly declining costs per token.
>> Mhm.
>> And if the cost declines, the demand will be there. I'm pretty confident about that.
But not, is any of this contingent on newer breakthroughs? Or.
>> I don't think so. I think we.
>> Look, there's a set of breakthroughs we can forecast pretty easily because we know how to get there. If we scale X or Y, you know, we know scaling laws apply.
>> Uh, we know certain cost curves of silicon and other things. You put all that together, I don't think any massive breakthroughs are needed. I consider things like memory in AI systems, continuous learning in AI systems as sort of routine things.
>> Again, a question of when it'll come along, not if.
>> Yeah. And the bigger question is, when they come along, how good are they?
>> How good is the continual learning?
>> How good is memory?
>> How, how well can it be utilized?
>> Uh, got it.
>> So, I, I think those are predictable. I also expect we'll see unusual of.
>> Phenomena in AI.
>> Progress that we aren't planning on.
>> Mhm.
>> Unusual good, unusual bad, we can't say. I think mostly unusual good. There will be some bad things for sure.
I want to bring the conversation to health. I think around 2016, you wrote this paper, I think it's called something like "20% Doctor," which was about the role of AI in medicine and ubiquitous, you know, world-class doctor available to everybody on their mobile phone. In fact, incidentally, we know 2016 was when I was starting my second company, and it was a health company. We wanted to solve for holistic lifestyle across everything from fitness, food, and marrying to primary care. Most of those things didn't work out. Fitness worked out very well. So today, it's a very large fitness company. But the dream of that AI doctor was, at least that we couldn't pursue. But fast forward 10 years from now, do you think we are much closer to now that vision of where everyone has.
>> There's no question. Even five years from now,
>> You won't need a human doctor other than in interventional medicine. That means like surgery and heart surgery or things like that, or burn victims or broken bones.
>> Uh, I think five years from now, the expertise involved in a doctor.
>> Is almost completely already in existence. I no longer consult doctors. I fractured my wrist last week.
>> Um, got an X-ray.
>> Sent it to ChatGPT and told me what to do.
>> Now, I have had to have a real surgeon. I called the surgeon.
>> You know, like any surgeon, he said, "Can you come in for a consult?"
>> I said, "I don't think I need it. Here's what ChatGPT told me." I sent him, uh, the exact comments from ChatGPT. He said, "This is exactly what I'd tell you." So you don't need to consult. We'll go directly into my first time I met him was in the surgical theater.
>> Uh-huh.
>> And it worked out. And since then, I consulted on every question, like, can I do this or can I do that, or when will the cast come off, or.
>> When can I put pressure? Or.
>> It's a much better.
>> Uh, I, I would say it's at least as good a doctor.
>> And a much more accessible doctor. I can ask it a question at midnight and get an answer within.
>> Five minutes.
>> Yeah.
>> As opposed to waiting multiple weeks for an appointment. That's possible for every person in India today.
>> Mh.
>> And I, I, one of my pet projects is to get that going and deliverable in India as a nonprofit effort within the next two years.
And why does it need to be a new effort? GPT already is available in India. They have, I think, pretty wide distribution for one of the largest user base for GPT, as you say. I also, you know, consult GPT, my first protocol for any health issue, and I'm pretty happy with the outcome we get. Why does it require a separate effort at all?
>> Um, I think this is misunderstood. The safety requirements.
>> For AI are much, much higher for some areas like health.
>> Than in others.
>> Yeah. So today, if you consult ChatGPT, um, the general triage error rate will be 20 to 30%.
>> That means you get the wrong answer 20 to 30% of the time.
>> Now, humans are pretty bad at triage too. So.
>> You have to be careful.
>> Uh, but when you add a health-specific system.
>> And my son has a company that does that.
>> But they use GPT-5.
>> Mhm.
>> But layer on things on top.
>> The triage error rate goes to zero.
>> Mhm.
>> Right.
>> Much better than human beings.
>> And so I think in all these areas, if you're talking about national defense, or healthcare, or financial trading.
>> All of them will are essential.
>> They're necessary but not sufficient. You need additional things.
>> Yeah. Or various levels of insurance.
>> So this is not about making AI better with further scaling. It's about building enough guardrails and checks and balances and human.
>> Kinds of every domain is a little bit different. You need different things in different guardrails.
>> Mhm.
>> But yes, you know, we know AI systems hallucinate.
>> Yeah.
>> And it's not been easy to reduce hallucination.
>> Mhm.
>> If you're doing customer support and you hallucinate somebody's bank balance.
>> Uh, that's a problem.
>> So, there are special systems that'll ensure that the systems don't hallucinate and still use these systems, but with the extra guardrails and safety. So, whether it's financial or healthcare or other critical areas.
>> And these, these additional systems are more deterministic, rules-based systems. They are not additional inference to.
>> They're not rules-based systems. I think rules-based systems, uh, are too hard to do and manage.
>> They're just better, different techniques with different tradeoffs.
>> Right.
>> Got it.
>> Got it. So maker-checker, and probably multiple checks, but they're all inference-based systems that are layered together in some way.
>> So that's why there's opportunity in application areas too.
>> Right.
>> Right. And this is, I'm just thinking about now, India healthcare, since you mentioned, we have a massive public healthcare system. There are, I think, if I'm not mistaken, hundreds of thousands of doctors work there. So, one possibility is to co-opt them in some ways to be the human layer.
>> Yes. You know, whether you need a human layer or not.
>> Mhm.
>> Um, and for what.
>> Is an important question.
>> And I think each human being will make, every citizen in India will make their own decision.
>> Mhm.
>> Um, obviously, if, if an AI is going to tell me I got cancer, I might prefer a human being to tell me that.
>> Than, uh, get a text message.
>> Um, so, but different people may prefer different things.
>> Um, others may say, "Hey, tell me as soon as possible." Text is fine.
>> I don't want to wait one week for a doctor's appointment to learn that.
>> Mhm.
>> Um, so different people will prefer different things, and I think human preference.
>> Will become a big part of how we use AI systems.
>> Yeah.
What does education look like in this? You know, it's a, both, I maybe I'm asking from a two-part question for a grown-up person, you know, someone who is in his 30s, 40s, 50s. Probably, I assume people need to learn a lot of different things and perhaps unlearn. And then a similar question for someone who is coming of age now, how should they think about education?
>> I spend most of my time today.
>> At age 71, learning.
>> I'm learning new subjects.
>> If I'm interested in fusion, I learn about fusion. If I'm interested in biology, I learn about biology, cancer, drugs, whatever I need to learn, I can learn now. So, the opportunity for 30 and 40 year olds, and 50 and 60 and 70 year olds is huge as far as learning, and it's almost all free.
>> Mhm.
>> So, that's the big, big advantage.
And when you say learning now, it's mostly going to GPT and just finding out the right resources from there and then following them.
>> Yep.
>> That kind of thing.
>> Yeah.
>> I, I often have long dialogues with GPT on a subject. Now, even a typical, how do you design a much better drone motor?
>> Mhm.
>> I've not designed a motor in 50 years, but I can still design one with GPT and say, I understand the basic scientific principles, and I know how to apply them, and I'll do these exercises.
>> M.
>> I say, how do you, how would I design a cancer drug for this cancer?
>> All those are available and possible.
And this is something you said, this is something one one can do, or one should do?
>> Well, that's each person's choice.
>> To, to, to survive and thrive in this, you know, new world.
>> Survive and thrive, you have to do it.
>> Right. If you don't want to, um, keep up, that's fine. But I'll give you the best example. You know, most people in the software business know GitLab as a company. He was one of our founders. We invested when there were four people, became a big company, public company, and then the founder got cancer.
>> Mhm.
>> And if you look at his history, he basically decided very quickly, he wouldn't talk to doctors.
>> He would essentially design his own cancer treatment.
>> Yeah.
>> And he got deep in, he learned everything from scratch as a software person.
>> Mhm.
>> And had, did his own cancer vaccines, cancer drugs, custom drugs for each stage of his cancer.
>> Yeah.
>> He wasn't given that long to live. He's alive and well and healthy now.
>> And doing even more drugs for what he might need as cancer drugs in the future. So if a software engineer can do that.
>> And you don't need a medical degree.
>> Uh, that's the world we can live in.
>> We can choose not to, or choose to live there.
>> I find it very, very inspiring.
>> Yeah. This is an incredible example. Does he talk about this publicly?
>> Oh, he talks about it. Um, in fact, he's given a couple of talks.
>> Uh, called, uh, it's called "Going Founder Mode on Cancer."
>> Uh-huh.
>> It's a beautiful talk. I'll ask you for an introduction. Try to invite him on this, because this is the ultimate example of what someone powered by AI and the right mindset.
>> With no knowledge can do.
>> Right.
>> Just general intelligence.
>> Right.
>> And, you know, you don't have to have a degree to have general intelligence.
But what does it mean for someone who is going to college now? Why go to college?
>> I think the only reason to go to college is to learn how to learn.
>> To learn curiosity.
>> Why, why do I work 80 hours a week? Cuz I have curiosity about so many things.
>> Yeah.
>> You know, um, I asked ChatGPT, what do you think of all my searches over the last couple of years?
>> It freaks out because it says you've been all over the map.
>> Yeah.
>> You're learning about cancer, you're learning about fusion, you're learning about designing motors, you're learning about,
>> You know, various AI algorithms, just everything.
>> Yeah.
>> Uh, even gardening. I learned so much from ChatGPT about gardening, my own garden.
>> Yeah.
>> Um, so you go because you learn how to learn.
>> Mhm.
>> You go because you learn curiosity.
>> Yeah.
>> And by being active and online, and this may be the single most important thing.
>> Yeah.
>> You get agency.
>> Mhm.
>> You feel like you can do things. I think that's the difference between founders and non-founders. Founders have agency.
>> Yeah.
>> Instead of looking at a problem and saying, "They should do this," which everyone says.
>> Founders say, "Oh, there's a problem. I'll do it."
>> Yeah.
>> I'll solve it.
>> So that's founder mode.
>> Yeah.
>> And say, there's no, it doesn't matter if I know about it or not.
>> Right.
>> I'm just going to go start learning and attacking the problem.
>> Yeah.
>> So agency becomes very important to learn for.
>> Both a five-year-old and a twenty-five-year-old.
So perhaps it has a huge implication for how schools and colleges cultivate this curiosity and agency, because a lot of it is about teaching you something that's in textbooks and being able to repeat it back in exams and so on, and that's how the whole system works, and people become very good at it. But that's not exactly fostering curiosity and agency.
>> No. And, and I think universities need to change. I was, I was at IIT Delhi in February. Myself and Sam gave a talk, and then I was talking to the director.
>> And he was saying, "We're going to expand the university. This is what we're going to do with classrooms."
>> And my answer was completely irrelevant. Do you think any of the students in our audience today will know more than the AI in their area of specialty? The answer is obviously not.
>> Right.
>> Right. So what should you do with IIT Delhi? I said, "Open more dome rooms to have more students."
>> Yeah.
>> Have more meeting rooms, not classrooms.
>> Have places where people can gather and discuss things and debate things.
>> And do the learning at home and come back and discuss topics.
>> Yeah.
>> So there is a role for universities, but it's not the old role. It's not lecturing out of textbooks. Those can all be thrown away. They're all obsolete anyway, because the world keeps changing, especially in technology areas like engineering.
>> Every day. You can say, "How do I design a new magnet without railroads?"
>> Right?
>> You're not going to find it in a textbook.
>> You're going to find it on ChatGPT.
>> And then pursue trails and build a little model and then experiment. We have people doing things like that.
>> Yeah.
>> Uh, you know, the country doesn't have rare earths. Can we not depend on China? Do our own.
>> The answer is probably a small group of people could solve that problem.
>> That's called having agency, if they choose to go do it.
>> Yeah.
And I want to just underline that. I've very often noticed that sometimes I stop myself from asking the question. But if I do ask a question and engage with any of the AI models, it leads to a totally new trail, and you didn't realize it's possible. So then having that agency and somewhat courage to act on it, just, you know, like this, this is this founder you mentioned, you know, just trying to solve, figure out his own cancer situation with AI. Most it will not occur to most people to even attempt. And a lot of this promise of AI, we know, uh, today, everything is, in some ways, digital medium. They're all mostly text-based. Yes, can process some image, etc., but the majority of, uh, both input and output is is text with some translation. Uh, how, where are you on the whole, this physical AI aspect, you know, AI.
>> Well, physical AI is coming. In the next couple of years, we'll have the ChatGPT moment of physical AI.
>> Um, I've tweeted a bunch about it.
>> Uh, so I don't think there'll be an issue with robotics in five years from now.
Can you envision what does a ChatGPT moment can look like, or what are the possibilities where we say this becomes relevant to all of us?
>> You, you know, it's, I think when you can put a robot in your home, it can cook your meals.
>> Yeah.
>> That's a pretty stunning point.
>> Mhm.
>> Obviously, there's lots of industrial application.
>> But a robot in your home.
>> And these look like humanoid robots, not specialized arms or what?
>> The most likely humanoid. I think there'll be multiple specialty form factors, but humanoids will be the largest form factor by a lot.
>> Mhm.
>> And because they'll be the largest form factor, they'll have the most scaled manufacturing.
>> And they'll have the lowest cost.
>> Yeah.
>> So you'll need auto manufacturing like facilities to make large scale, very low cost.
>> Mhm.
>> But, you know, cars don't cost that much more than the cost of steel.
>> Right.
>> Right.
>> If you look at it, because the value add is so automated and so scaled.
>> That it's very low overhead. So.
>> And if you play out this exponential growth and impact of AI, then the world will be unrecognizable. The.
>> I think 15 years from now, the world will be unrecognizable. A five-year-old kid today.
>> Will grow up in a very different world by the time they're 20 or 25.
But it has also implications for our social, political, economic structures also. They can't just hold up the current, because in some ways today, if you know Anthropic has gone from nothing to a trillion-dollar valuation four years, and it seems, you know, Google is now $5 trillion, probably most likely going to a $10 trillion company in the next few years. So it seems like the benefits today are getting highly concentrated, while there's some innovation outside, but the most of the benefits seems to be flowing in a very, very small.
>> Area. I disagree with that. I think the question to ask is, is 95% of the population better off?
>> Mhm.
>> Not what's happening with the 5% who are making all the money. And that is happening. There's concentration of wealth.
>> Yeah.
>> And technology has always resulted in a concentration of wealth.
>> And they'll be more nonlinear.
>> Yeah.
>> But if you stop worrying about the few percent.
>> Yeah.
>> And say, will 95% of the people be better off in 2040?
>> Yeah.
>> Absolutely. No question. They will be better off because they are making more income in their jobs, or they will require some redistribution where proactive.
>> No, because have access to services.
>> Yeah.
>> If they need education, it's free. If they need.
>> Health care, doctor, it's free. If they need legal services, it's free. If they need more entertainment, it's near free. Hopefully, even food and others will get cheaper because of robotics.
>> So, I don't need to make $100,000. And even if I make $50,000, I have a dramatically better lifestyle, possibly if.
>> My bet is we'll have a very deflationary world before 2040.
>> And prices will decline very dramatically for most goods.
>> A few things that aren't done by robotics or have other physical constraints.
>> Will be more expensive.
>> Uh, but we'll have, for today's measure of GDP, which is defined as a basket of goods and services that we measure today.
>> Yeah.
>> That basket will become much cheaper.
>> Yeah.
And I think some of the forums, you know, you have argued for that it will at some point require some government ownership for.
>> I think there's many different systems possible. U, each country will adapt its own.
>> Yeah.
>> You know, likely the US is very different than India is very different than Germany.
>> Yeah. Some people will resist AI, some countries will, and it depends on the politics in the country. So it's very important for a country like India.
>> To show the benefits of AI to people first.
>> Yeah.
>> Free doctors, free teachers, free agronomists. So every small farmer can.
>> Uh, worry about it, and the price of input.
>> Yeah.
>> Goes down.
>> Yeah.
>> And so they see the benefit for.
>> Mhm.
>> Uh, they see the asymmetry. So there will be more asymmetry.
>> Yeah.
And for India, I guess, there's an additional challenge of figuring out that this whole IT service industry and BPO industry, which is a huge source of, uh, foreign income, uh, for the country.
>> Business will be gone. There may be new opportunities in deploying AI to the planet.
>> Mhm.
>> Uh, because India has an advantage in learning how to deploy AI. Yeah.
>> Very few of the companies are doing it today.
>> But, uh, if they do, they'll be in good shape. If they don't, they'll be in very bad shape.
What are the examples of that? You know, what can India develop at scale which can employ millions of people and participate heavily in deployment of AI?
>> Yeah, it's, it's a complex area. Um, I, I, you know, I, I've written a piece about it earlier this year for AI specific to India.
>> Yeah.
>> So I'd refer you to that. It's about 20 pages long.
>> I'll go through it.
>> So it's a fairly detailed look at.
>> What we'll do. But, you know, the basic things people need.
>> Yeah. Health, education, um, entertainment, food, those should become much, much cheaper. Housing is the one area that's troubling because that needs to scale and still mostly depend on materials costs.
>> Yeah.
>> Not so much labor.
>> Right.
>> Yeah. I guess if you're physical AI and the really cheap humanoid, humanoid, so probably the.
>> Yeah. But, but cement will still cost like cement costs and.
>> Hopefully, we can do something about those, scaling those things.
>> Yes, cement also, I guess, is mostly energy cost in some level. So if energy becomes cheaper, then I guess you can, you'll have more options.
>> And where in this whole, uh, the, the, in the quest for solving for energy, for, uh, an AI space has become a factor now? And you have, I think you've been investing in space as well with the rocket layer way back, I think 2012, '13, so on. Have you, are you convinced that space has a role to play as for to in in these, uh, data centers to support massive compute infrastructure?
>> Look, I would say data centers in space don't make sense today.
>> But there's certain cost parameters like transport cost to space for a kilogram of weight.
>> Things like that.
>> Mhm.
>> Um, that could make it more cost-effective.
>> You know, are factors of 10 possible in cost reduction? Yes. And if that happens, then the equation might change. But.
>> With today's increase in, uh, today's costs and even normal declines in costs, uh, data centers in space haven't seemed to make sense.
>> But I don't rule out larger increases in decreases in cost. And if that happens, then assumptions will change around data centers in space.
If you apply your framework of what will be feasible by 2035, which side of bet today you will take, the space-based data centers a big part of our life, or unlikely?
>> I would say unlikely, only because the cost on Earth will decline very rapidly.
>> Got it.
>> Once you have fusion, you don't have a power generation problem, for example, by 2035.
>> So if power is dirt cheap on the planet, then why go up in space? Why get the inflexibility of fixed infrastructure that you can't change quite as easily as walking into a data center.
>> And changing a GPU that failed.
>> Right?
>> You know, suddenly GPUs will have to be pretty different.
>> Mhm.
>> Cuz the radiation is not healthy for semiconductors. Power generation in space is hard. Uh, cooling in space is hard. Now, there's possible solutions to each of these problems.
>> But if we, and if you're lucky enough to get improvement in all those dimensions much faster than cost decline on Earth,
>> Uh, then space centers may make sense. Today I'd bet that, uh, data centers on Earth make more sense.
>> Got it. And I think we.
>> And they will decline rapidly in cost, and power will not be an issue.
>> Yeah.
And hopefully, I guess, we'll start to get some data back as well in the next few years with all these plans of various companies to have at least some experiments going on. So maybe two, three years, we'll have some more evidence start to.
>> Uh, pile up as well.
And what is the process of looking at it? And maybe we can discuss in the context of fusion, because, you know, fusion, even today, most people don't believe is going to be realistic, the net gain from fusion over the next five years. I think, this is an area I tried to study the last couple of years. In fact, I was looking for a possible investment in India. We have a few fusion startups. And maybe I'm guilty of, you know, when falling looking for consensus and not outlier belief. But pretty much anyone you will talk to, even today, they think it's not happening.
>> And, and that's the key. I don't talk to experts.
>> You know, experts are experts in a previous version of the world, not the one you're trying to create.
>> Yeah.
>> Like that's the key to remember.
>> If I get three teams to look at fusion.
>> And all three agree, then it's probably not an unusual bet. It's consensus.
>> Yeah.
>> If teams disagree, then it becomes more interesting. So it's, look, it's hard to explain this process.
>> I think we do a pretty good job of judging where technology is going.
>> Yeah.
>> With probably 60, 70% accuracy.
>> Mhm.
>> And then we make a lot of mistakes, but we're not afraid of those mistakes. So we've made our fair share.
>> Mhm.
>> And not never afraid to be mistaken.
>> Mhm.
>> But, you know, when you're mistaken and you make the wrong bet, you lose one times your money.
>> Yeah.
>> But when you make the right bet, you make a 100 times your money. Then.
>> Then it's worth bet worth placing, right?
>> And that's our paradigm. You know, if OpenAI is successful, we should make more than a 100 times, which we will.
>> But that's always been my approach. Most people are afraid to fail.
>> Mhm.
>> I like to say my willingness to fail is what gives me the ability to succeed. I'm not embarrassed when I fail.
>> Yeah.
>> Um, and so I don't worry about it. If you write every investment off the day you make it, then you only have upside.
>> Same is true of Fusion.
>> There's a trillion-dollar company in Fusion to be had.
>> Most people, you say, don't agree it's feasible. I'm very sure. I think at this point, there's greater than 80% probability in five years, nobody will be debating fusion works.
>> And fusion works economically. My bet is in two years, people will say fusion works, and five years, people will say fusion works economically, because you have to pass through both gates.
>> And once it's economic, it's the source of energy for the whole planet.
>> Mhm.
>> By the way, there's other sources. Super hot geothermal is another one I'm very excited about.
>> I think it can be cheap, cheaper than natural gas and oil as a source of energy.
>> Both fusion and super hot geothermal. That means geothermal above 400 degrees centigrade.
Let's talk about both. But I want to double-click on fusion. Just, I'm trying to get to, you know, your process and way of thinking. Let's take fusion. There are two ways one can look at, you know, one is macro. We obviously know fusion works. So we can look at the sun and know it's a, it's a, it works. Uh, but there are so many competing platforms, and to really get into the nitty-gritty of fusion, you know, different technology and so on, you need to be fairly technical in nature. Do you at Host Venture employ consult, or maybe, you know, goes against your philosophy of consulting experts? Is it like a macro bet, or is it like looking at technology has now reached these many gates and readiness, therefore it might be realistic in five years?
>> No, I think we look at fairly detailed look.
>> Yeah.
>> Uh, obviously, we employ experts from time to time.
>> But most of our gut gut call is ourselves.
>> We are pretty technical. We can make these calls.
>> So, more expert diligence is confirmation rather than decision.
>> Mhm.
>> Uh, the question you have to ask in something like fusion is, why now?
>> Yeah.
>> In 2018, what has changed that would say it's not possible before, but now it might be possible? And that was high-temperature superconductors.
>> Mhm.
>> And that dramatically changes both the experiments you can run and the rate at which you can run them.
>> Yeah.
>> And also the economics. Uh, you know, if you can use high-temperature superconductors, which was the big risk to build a, in 2018, to build a 20 Tesla magnet.
>> Mhm.
>> Um, you know, if the magnet is four times more powerful than what was possible before, your reactor is going to be 250th the size.
>> Right?
>> It suddenly becomes much more feasible to experiment, try things, build things. It doesn't take 30 years to build. It takes three years to build.
>> And so everything becomes faster.
>> And that was the bet we placed.
>> And that's the kind of thing we look for. You know, every, every time there's something on the horizon that others don't believe.
>> Mhm.
>> Um, that we have to believe, and then find the right team. The right, finding the right team is very key.
>> Mhm.
>> When I met Bob Mumgard, he was a senior fellow at the MIT Plasma Fusion Lab.
>> I was just visiting because I like techy, geeky places to visit.
>> And he showed me this reactor where, before I stepped on it, the temperature was.
>> Couple hundred million degrees, like 30 seconds before.
>> Mhm.
>> And now you can step on it.
>> Uh, pretty cool. Uh, and so I like geeky things. We got talking, and the more we talked, the better I liked him and the what he was thinking, and said, "Let's build a plan."
>> So that's how that came to be.
>> Got it. Got it. So a lot of, I guess, in this case, I probably bet on the person as well.
>> And bet on the person for sure. Bet on the technology direction.
>> What is feasible, what's not? That's a risk assessment. Mostly I will do my own risk assessments, and other people in our firm are very good at it too. Right?
>> So together as a group, we can do that.
>> Yeah.
One of the things when I, you know, when I try to study fusion, one thing keeps coming up that what about fission? We have not really exploited full power. We have so much uranium, thorium, and so on. Small modular reactors. And do you also look at this that, that we need, the world needs a lot of nuclear energy? It will be great to have a lot of our energy, especially now with the, the power energy that AI needs to come from nuclear process, one way or other. Do you think about fission as well, or or fusion is something that once it's solved, is obviously we have infinite.
>> Once fusion is solved.
>> There's no need for fission.
>> And fission has a different problem. At least in the West, nobody wants a fission plant in their neighborhood.
>> Right.
>> So fission technology is not hard. Yeah.
>> It's possible to do fission.
>> Uh, technologies, and even like the Thorium reactor India has been working on, which is a very good idea.
>> Um, but my view is very simple. If, after you decide to build a plant and have the technology, you're going to have 10 years of lawsuits.
>> Nothing's ever going to pencil out.
>> Uh, so you have to spend a lot of money before you know somebody will let you build in that neighborhood. And if you're building 5,000 plants for the United States, it just doesn't become feasible.
>> Right?
>> So you have to look at the world holistically and say.
>> The politics will not allow fission to have 5,000 reactors in the US.
>> Or 5,000 in India. India will need more than 5,000.
And is there a risk that same politics will also try to at least get some hurdles for these massive data centers which are going to be energy guzzlers and which are absolutely essential for this?
>> Yes and no. But there's solutions. A number of the data center providers are saying, for everybody in the neighborhood of.
>> A data center will pay free electricity for residential purpose.
>> It's a small tax.
>> On their energy consumption.
>> Mhm.
>> You know, if they're doing a gigawatt, they give 50 megawatts of free electricity to neighborhoods. People are fine.
>> So there are solutions. You can also develop your own reactors, whether they're super hot geothermal or fusion, or nuclear. You can do your own, so you're not tapping into the grid. So there are solutions.
>> But it is a political problem.
>> Mostly because people haven't worried about the politics.
And given the current conventional source of energy, the available next four, five years of AI acceleration, is energy likely to be a bottleneck, or is it where we're okay?
>> I think energy is likely to be a bottleneck the next four or five years.
>> Why? Because four or five years ago, we weren't thinking about the problem.
>> And nobody thought it was a big problem. Nobody believed AI would be a large consumer of electricity. Nobody that.
>> And so nobody prepared for it. I think there's a lot of preparation for it now, and so I think five years from now, it'll become less of a problem. It takes five years to build a power plant, maybe longer.
>> And so in the early 2030s, enough power will come online, enough capacity, enough turbines, enough alternative sources of energy. I think.
>> It's a short-term problem. It's a five-year problem.
>> Got it. Correct. I want to bring the conversation back to, uh, AI and what's possible with AI, and that's the agency I guess you're talking about.
>> This is the agency idea.
>> Right?
>> Uh, Vinod, I think we're nearly out of time. On the, just closing note, I want to really zoom out and the big picture question about how all of this pans out in the longer term. You know, one scenario is with this whole, let's say, the entire promise of AI is fulfilled. There is, you know, abundance. Uh, where are you on this whole so-called notion of singularity that regards the will? And I think you've been to one of the Peter's, you know, even.
>> I take a practical view.
>> If I look into the 2040s or 2050s.
>> It is possible, in fact, very likely.
>> Mhm.
>> If we have the right government policy.
>> Yes.
>> That the need to work.
>> To survive for a living will go away.
>> Right.
>> We people will still work, but work on things they want to work on.
>> Yeah.
>> Not things they have to work on because it's the only job they have.
>> You know, who wants to work, uh, in a field doing farm work.
>> In 40, 50 degree heat?
>> Who wants to work on an assembly line for 8 hours a day for 40 years?
>> Yeah.
>> That to me is slavery. It's servitude to survival because you need a job. Those are not jobs with human dignity.
>> I think those jobs can disappear.
>> And people can work on what they want to work on, as they want to work, or get good at singing, artists.
>> You, you name it.
>> Sports, competing.
>> So many things to do.
>> So freedom is the ultimate promise of AI, and just people free to live life, and this, that becomes a new normal. Hopefully, whatever next for humanity, space or beyond, I guess will play out. Let me say, and I say this in a piece I wrote two years ago. I wrote a piece, "Will AI Lead to Dystopia or Utopia?"
>> That was on AI in the Western world. I wrote a piece earlier this year for India.
>> I believe AI will ultimately free humanity.
>> To be human beings.
>> Not servitudes to survival. Survival has been.
>> The goal of every species till now.
That's an incredible message and something, you know, something to really look forward to. Vinod, I want to thank you for taking the time for this podcast, as well as incredible career and impact in the world through all your investment, and also inspiring so many VCs as well as entrepreneurs. And I think you've had a very different way of doing things, but, you know, this track record speaks for itself. So I think, you know, talking to you is very inspiring. I'm pretty sure our audience will be inspired, and hopefully, it will foster more agency and more people to really embrace, you know, the promise.
>> I hope so. I was inspired myself learning about a Hungarian immigrant coming to Silicon Valley to start Intel.
>> That inspired me and got me on the entrepreneurial track. So, I hope more people get inspired.
>> Thank you, Vinod.
>> Thank you.
At Sparks, we aim to bring to you stories of exponential impact. We share in-depth analysis of what goes behind success stories. If you find our conversations interesting, you can join us by subscribing to our YouTube channel. You can also listen to Sparks on Spotify, Apple Podcast, or any other audio platform of your choice. If you have any suggestions on who we should invite or what topics we need to cover, just let us know in the comments. We are always listening, looking for ways to improve, and keep getting better as we go along.