EVERY SPOKEN WORD
25 min read · 4,563 words- 0:00 – 1:20
Introduction
- NKNikhil Kamath
[upbeat music] I'm not going to ask you anything generic or anything I've heard you answer already. Uh, I'm going to be a little selfish and ask you questions that, uh, I try to answer for myself a lot of the time. Maybe like, you know, uh, there's no better person to learn from. [upbeat music] So we do this once in a month, Bill, and, uh, we primarily cater to the young entrepreneur in India. Uh, the idea is to get really good people on the show and help a young twenty-year-old boy or girl in India who's looking to become an entrepreneur, give them, uh, advice, guidance, uh, help them learn from what you might have learned already.
- 1:20 – 2:20
Nikhil’s past Interactions with Bill Gates
- NKNikhil Kamath
I've met Bill a bunch of times. I think the very first time I met you was, uh, when I caught you at WEF and bugged you for an hour with random questions a few years ago. I joined Bill's pledge last year.
- SPSpeaker
Yeah. Thank you.
- NKNikhil Kamath
Uh, the one thing I want to clarify is a lot of Indians, when they hear about an Indian sign on to the pledge... I think there's four of us, right? We're all from Bangalore, and all of us are friends. Uh, people somehow assume that the money is going out of India, the money which is going to make society better. Uh, it's actually not the case. And, uh, I, I would also like to say, I think The Gates Foundation, who I've worked with in the past, uh, we've done a school project together, we're doing something around, uh, malnutrition together. I think the impact and the effort that they're putting into India in improving society, uh, is, is, uh, is great, and I hope you guys continue this association with India for a long time.
- 2:20 – 6:10
Bill’s Experience with India over the years
- NKNikhil Kamath
Do you want to talk a little bit about your relationship with India and how it's evolved over the last many years?
- SPSpeaker
Well, I mean, I've had a fantastic relationship with India, starting with the, the Microsoft experience, where, you know, we hired some very smart IT graduates, brought them to Seattle. Later, they go back, uh, create a development center for us, uh, that's now in four locations, twenty-five thousand people. And, of course, a lot of the amazing people I work with and have so much fun, uh, in the Microsoft success with are, uh, part of the team we have hired from India. You know, top of that list is Satya-
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
... uh, who now, uh, is doing a, a great job as CEO. So in my digital first career, uh, the connection to India was, uh, fun and made a huge difference, um, in what the company was able to achieve. It was during that time that I was kind of learning, oh, wow, India is such a study in contrasts, you know. You know, first class in so many ways, but still a lot of poverty, uh, and challenges. And, uh, you know, I was beginning to think, "Okay, how do I give this money back?" Uh, together with, uh, Melinda, and so she came over as we were starting to do foundation work. Um, I start the foundation, I fund it in the year two thousand, so I'm forty-five years old and still, um, full time. I stayed full time for another, uh, eight years. Uh, but India becomes, you know, an important place for us because we realized that vaccines, uh, were this missing, uh, thing, and it was a crime that these, uh, diarrhea rotavirus vaccines were getting to the rich kids. So, you know, I get to know Serum, I learn about making cheap, uh, vaccines. Uh, you know, the Indian government at the time had not adopted new vaccines for a long time. You know, so, uh, how do we, as foreigners, uh, form partnerships and not come across, uh, in the wrong way? So it's a heck of a learning curve. Uh, another big program we had here at the start, called Avahan, uh, was about making sure that the HIV epidemic didn't explode, you know, by making sure that, um, the sex workers were insisting on condom use and so that the, the numbers would stay small. And that was also quite successful and very much a partnership with the government. Um, you know, other than the US, this is the country we spend the most money in. Uh, we have a fantastic team of people here.
- NKNikhil Kamath
Do you want to put a number to that in terms of the scale of how much the foundation does in India?
- SPSpeaker
Yeah, we do. Uh, if you take direct and indirect, uh, it'd be close to a billion a year. You know, we help buy vaccines through, uh, Gavi. We, uh, fund a lot of different things. Our foundation is about nine billion a year, and our health is our biggest, agriculture is our second biggest. Education, uh, although super important, uh, is not, uh, super high percentage because we're so specialized on the health stuff, uh, and we've had so much, uh, progress on that. We want to, uh, make sure we do that super well.
- NKNikhil Kamath
I'm not going to ask you anything generic or anything I've heard you answer already. Uh, I'm going to be a little selfish and ask you questions that, uh, I try to answer for myself a lot of the time. Maybe like, you know, uh, there's no better person to learn from.... when it
- 6:10 – 10:20
Does philanthropy actually work?
- NKNikhil Kamath
comes to, uh, society at large, a lot of really wealthy people today, uh, are virtue signaling in the manner that being wealthy is a bad thing. Somehow it, it's, it's starting to seem like the class and race divides of yesterday has become the new, uh, wealth divide has become that thing in society today. Uh, which is very socialist in the very thinking of it, inherently socialist, when people think of it that way. And when I go on my own path of philanthropy, I often ask myself, like even at the pledge, I find myself asking this question: Does philanthropy in that sense truly make sense? And me, as a thirty-seven-year-old who's starting on this path, how do I think about this?
- SPSpeaker
Well, there are some things that only philanthropy can do because they require taking risks.
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
And, you know, the basic needs of the people, the safety net, that's not really the role of philanthropy, because you need to count on that year in and year out, and the government, you know, through the political process, needs to decide that. And of course, the market-based system is where most of this wealth comes from. And anything the market can do, you know, like how many restaurants like this should we have? You know, [chuckles] the market does that so well.
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
But we know there are diseases and challenges that either the government's not innovative enough or the market opportunity's not enough, that the two big actors, you know, private sector, you know, which is like, you know, seventy percent of the economy, and, you know, the government, which is like twenty-eight percent, this third sector, which is only a few percent, can step in to do. When we had, uh, democratic capitalism, there was always this concern that the poorest would kind of overtax the rich and that the incentive system would be broken down. In fact, you know, that really hasn't happened. You, you have, in Northern Europe, very high tax rates, but e- even there, the incentive system works. So I think it's important to separate out, is my country's tax system as progressive as it should be? Uh, so when you look at billionaires, you know, do they kind of have more than they should? And then, if you look at whatever the billionaires have o- left over after they've paid their taxes, are they engaging in thoughtful, uh, philanthropy? Are they kind of spoiling their children, uh, who are kind of arrogant and view themselves in a class of their own, uh, and being kind of extravagant in what they do? And that's very hard, you know, to draw the line. What is, uh, that responsibility? Even with the giving pledge, this idea of give back half, you know, some people say, "Well, half, if you have, you know, four billion dollars, that means for your own consumption and your, your family, you're keeping two billion." Well, it's not a upper bound. [chuckles] We'll have, you know, people who, who give away ninety-nine percent. But it... And it's great to debate these things because, um, you know, we're not trying to take over government functions. We couldn't possibly afford, afford to do that. Uh, but a, a lot of the best government programs, if you look into them, some philanthropist funded the pilot, you know, brought together talented people. And often you're using the s- same skills that you-- caused you to be able to be sel- successful, and now you're using it on behalf of, of equality. You know, how do you design a school or take AI technology into a, a farmer's a advisory system? Uh, and that, that experience, s- that sense of urgency, [chuckles] uh, can be very helpful in philanthropic innovation.
- 10:20 – 11:50
Capitalism Vs Socialism
- NKNikhil Kamath
I'm, I'm with you when you say that the anomalies of capitalism, the billionaires on the very end, should be taxed higher. But if you had to redesign society, do you think capitalism over socialism any day?
- SPSpeaker
Well, capitalism gives you the freedom to start a new business and try out a new product. And the idea of, okay, who do people choose to buy from? That's a form of freedom because it's not based on your background or your class or something like that. And so the discovery power of the capitalistic market to, "Oh, this guy knows how to do this cheaper. This guy h- knows how to do this better. Oh, you know, people are interested in that kind of a movie." Um, you know, top-down systems, even though with lots of digital data, you could run a top-down syster- system less ineptly than, say, the Soviets or the North Koreans did. But still, the discovery process and the price comparison that takes place... So it's, it is, it is the fundamental way that things should be done, but government has to look for monopolies. It has to create the safety net. And you can argue about the relative role of the government in education and, and healthcare, but they need to make sure that, uh, everybody gets those services.
- NKNikhil Kamath
Right.
- 11:50 – 12:56
How does Bill stay healthy?
- NKNikhil Kamath
You're looking fitter than I remember seeing you the last time.
- SPSpeaker
Yeah, I know. I've been playing more tennis-
- NKNikhil Kamath
Uh
- SPSpeaker
... a little more pickleball. Uh-
- NKNikhil Kamath
Is pickleball really picking up that much? It is in India.
- SPSpeaker
In, in North America, it's unbelievable-
- NKNikhil Kamath
Right
- SPSpeaker
... particularly California, Florida.
- NKNikhil Kamath
Right.
- SPSpeaker
I mean, it's actually more people play pickleball now than play tennis.
- NKNikhil Kamath
Right.
- SPSpeaker
Uh, so it's actually passed it by.
- NKNikhil Kamath
... Uh, the last time we met, I kept bugging you about what supplements do you take? Now that you've gotten even healthier looking, is there any hack that you would like to recommend to young entrepreneurs?
- SPSpeaker
I take vitamins, you know, because it's, there's no downside. So getting blood tests every once in a while and seeing if there's anomalies, uh, that's good. And now there's a thing called the Grail test. Um, it's not covered by all insurance in the US, so probably not by much in India. But if you can afford it, uh, that is almost magical at helping to see cancers way before other techniques would catch those.
- 12:56 – 16:34
Energy Transition : What are the opportunities?
- NKNikhil Kamath
Uh, so we have a foundation, Bill, which does work around climate. It's called Rain Matter. Uh, just for the sake of not having this debate publicly, let's between us assume climate change is real. Uh, I've heard you speak about, uh, the Breakthrough Energy fund. Energy transition appears to be a big opportunity for young entrepreneurs to work in over the next decade. Are there any sub-facets of that? If I were a twenty-two-year-old boy or girl wanting to build a career in energy transition, uh, where do you think there are most tailwinds? What factor should I focus on?
- SPSpeaker
Well, because the sources of emissions are so broad, you know, we need deep innovation in many, many areas. So for electricity generation, you know, we've got to make solar cells better. We have to make off wind cheaper. Um, there's probably something dramatic that can happen in nuclear fission. Uh, hopefully something dramatic that can happen, happen in nuclear fusion. Uh, in the long run, that may be, uh, the cheapest source of electricity, and now we have some companies, uh, trying to get there. You know, over in the agricultural space, um, rice, uh, is a huge source of emissions, and there's new ideas about genetics and direct seeding. With cows, you can vaccinate them, change the genetics. Uh, in the industrial side, you can try to get cheap green hydrogen, or you can try and change the process. For example, with steel, instead of using a form of coal, uh, you could use electricity for the reduction. And so I'd say there's probably twenty areas that if you want to help in climate innovation, uh, these things are, are begging. Uh, and I think it will be startups, and then, of course, they'll have tie-ups with big companies, but the open-mindedness, most of these breakthroughs won't come straight out of the big companies.
- NKNikhil Kamath
Is there a starting point, though? Let's assume I have no capital at hand, and I'm twenty-two, starting off in Bangalore. Uh, how do I, how do I go about attempting to be a part of the energy transition solution?
- SPSpeaker
The ideal is actually to find a philanthropist. Um, Breakthrough Energy has a fellows program where we look for people with really, uh, good early-stage ideas, and we do that as a grant, uh, with a little bit of, of options, but you don't need a business plan. Then, if you get past that stage, either on your own or with a, a fellows type grant, then there are venture capital groups that really specialize in finding climate technologies, and they're willing to fund very risky stuff. Um, and I know all of them, including Breakthrough Energy, are like: "Gosh, are we missing some in India?" And so, you know, there's a network. You know, if you have a good idea, you know, if you pass it to me, if Nandan passes it to me, there's a bunch of US venture capitals and hopefully some who would also come in domestically. Uh, so for risk capital, uh, this is a very good time if you have something that looks attractive, could have a significant impact.
- 16:34 – 19:40
AGI: Impact on Jobs, Money and Society
- NKNikhil Kamath
Artificial general intelligence. Uh, you know more about this than most people. How do you think society changes? If you could paint a dystopian and a utopian picture, let's say there is AGI. What happens to the society post-AGI from, uh, jobs? What is money? What is society? What is family standpoint?
- SPSpeaker
Well, the amazing thing about this technology is, um, we know it can help, uh, in key areas. We know it can create, like educational tutors. A lot of great pilot work, uh, several projects here in India. Sal Khan in the US is the one that, um, may have started the first and is, is, uh, showing great results on that. So the potential is incredible. And if, if we knew that all it did was make a lot of jobs, say, three times more productive, so doctors are three times more productive. So, you know, the demand for more people to see doctors and get good advice, that's probably good. For software, um, you know, we'll still need those software engineers. We'll-- We may take some of the extra productivity for quality, some for quantity, but we're not gonna, um, you know, start, uh, not, not needing software engineers. The fact that over time, and we don't know the, the rate of improvement, um, and there doesn't seem to be an upper bound, if it gets so that it's a complete-... replacement, that it can do all the medical advice, or it can do all the coding. That's a world of excess where, you know, it's, it's got to be wonderful in the sense that, you know, handicapped people, old people, you're, you know, taking care of them. You're shortening the, the work week. But you can get to a point where the very organizing principles of society, uh, and the whole philosophy about, you know, what you value, um, that world of excess, which I personally think won't come in the next twenty years, but I can't guarantee that. You know, I was personally not expecting this reading and writing breakthrough to take place just because, you know, GPT-3 got scaled up to GPT-4, but that really showed something profound in knowledge representation, uh, had taken place. And so most of us are working on the, "Hey, let's go build that tutor and that health advisor and farmer advisor."
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
Um, but while we do that, we should be aware that, okay, so much extra productivity will, uh, surprise people and require us to rethink, uh, about a lot of different things.
- 19:40 – 22:06
How can Indians leverage AI?
- NKNikhil Kamath
So when Sam Altman was in India, he fielded a bunch of questions around ChatGPT and, uh, generative AI, per se. When these large language models take as much computational power as they require, an Indian entrepreneur who has lesser access to capital compared to a Western one, how does an Indian entrepreneur compete with the West in this domain without the same kind of access?
- SPSpeaker
You know, that's a very good question. So if you're an entrepreneur who's trying to build foundational models, models from the very ground up, like Google does, or Microsoft with OpenAI does, and they have four or five others, you have to raise, you know, two, three billion dollars. Uh, now, maybe some government will decide because you're domestic, they'll help you do that or provide some of that. But [clears throat] there should either be, like, zero, one, or two cases where an Indian company goes after that very base level, uh, and hopefully focuses on some specific Indian applications of that. Uh, most of the opportunity is at the level above, you know, where you're saying, "Okay, let's take this and help architects," or, "Let's take this and help this group of farmers." Or, you know, the Indian legal system is very different. Believe me, what legal things are done for the US won't apply here. But, you know, just imagine if you could make everybody in the legal system four times more productive. You know, that changes justice, because right now the backlog, uh, is kind of nightmarish. Um, and, and here, you know, comes a potential solution. So I'd say there's a thousand opportunities where you, you're-- you pick, either pick or just use one of those base-level models, but you're doing the extra data value-added user interface to go after a particular application. That, I think, will... is where we'll see the most activity. Uh, if somebody's bold enough to say, "No, I'm going after foundational piece," you know, "Screw you, Bill Gates, you don't know"-
- NKNikhil Kamath
[chuckles]
- SPSpeaker
... I say, "Hey, good for you!"
- NKNikhil Kamath
Right.
- SPSpeaker
You know, I, I could be wrong.
- NKNikhil Kamath
Right.
- SPSpeaker
We are in r- you know, this is all very new, uh, stuff.
- 22:06 – 24:10
Sectors to Focus on with AI in India
- NKNikhil Kamath
Are there any arbitrages in place, like a unique advantage India, either by virtue of population or the demographic diversity has, that, that will help a young entrepreneur focus on a particular facet around building this?
- SPSpeaker
Um-
- NKNikhil Kamath
Or is there a legal arbitrage? Is regulation better here than there?
- SPSpeaker
I think it's very possible that the health advice piece, that the US will overregulate that and, uh, you know, just see the, the risk and not, not the benefit, and that middle-income countries, and I'd put India at the top of the list, will engage in a dialogue and, and drive very aggressively to get the benefit. But yes, I would expect, uh, definitely in health, definitely in agriculture, a lot of the most exciting, fast-moving things, uh, will be here in India. Education, I think everybody's going to be doing it, and, and including India.
- NKNikhil Kamath
I've seen this happen in Indonesia. Do you think the business of delivering medical advice and prescribing people, uh, not with a human on the other side, but a model per se, do you think the potential for that is big in India for young entrepreneurs to focus on?
- SPSpeaker
Yes. Um, you know, there's going to be a quality bar.
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
Uh, and, and yes, and, you know, maybe at first it's okay pregnant women or people dealing with diabetes, um, because it's easier then to think through all the possible questions and make sure the quality is right. But eventually-
- NKNikhil Kamath
Mm-hmm
- SPSpeaker
... you bet, your, your lifelong medical agent that takes all your sensor data and, you know, how you felt, you know, when you eat this or drink that, uh, it'll be there, and it'll be reading the latest literature, um, and it'll be super helpful to you.
- 24:10 – 25:43
Bill's advice for people in early 20's
- NKNikhil Kamath
Okay, last question to you, Bill. If you were twenty-five years old, living in India, uh, no, no crazy skill set at hand, and about a hundred thousand dollars of money to start a business with, what would you do today?
- SPSpeaker
... I'd probably build, pick some AI thing and just use the Google, Microsoft platforms and, and go on top of that and try and-- because I'd be in touch with the customer so well and improving my data, keep ahead. Um, you know, and I'd say: "Okay, who's the most demanding customer for this?" Uh, and go after that. It might be fun, you know, like clearing legal backlogs, you know, that one kind of fascinates me, uh, because it, it's a mechanism that helps all, uh, business activity. Um, I'd envy the guys who get to work on the foundation models. I'd say, "Damn!" Uh, even I sort of envy them because when I was young, I thought, "Gosh, you know, when is AI really going to make the breakthrough?" It took a lot longer, uh, than I expected. I'm glad now I get to play a supportive, uh, role, help shape it a bit. Satya, uh, engages me, and now, you know, we're using AI in our foundation work, you know, whether it's drug discovery or all these things. Uh, so we can be way more ambitious than we were before it came along.
- 25:43 – 29:03
What are the possibilities with AI?
- NKNikhil Kamath
I happened to meet Satya and Sundar when I was in the US, and, uh, we were having dinner at a certain place together, and these guys seemed to talk about AG- AGI in a very precocious manner. Like it's much closer than anybody is thinking it is. Do you think that there, there could be one day when suddenly we are surprised, like a couple of years down the line, and it's there?
- SPSpeaker
Well, there, there's so much-- there's not really a really good definition. These models already today, when you have lots and lots of documents, you know, more than a human could read in a lifetime, it is able to find things and reason across these documents in a superhuman way. And so when you think about a lawsuit or a drug filing, and anything that gets to be over, say, ten thousand pages, human cognition-
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
-even a [chuckles] great specialist person, you get ten thousand pages, I dare you.
- NKNikhil Kamath
Mm-hmm.
- SPSpeaker
Uh, and these things, it's not just keyword search. This is semantic representation-
- NKNikhil Kamath
Mm-hmm
- SPSpeaker
... of that level of complexity. So we're already superhuman-
- NKNikhil Kamath
Mm-hmm
- SPSpeaker
... in a dimension that is kind of surprising. I mean, I thought robots, you know, would be taking over the warehouse, you know, ten years before they'd be studying our legal and regulatory documents. And we know that we have brute force scale up. We have a lot of external algorithms, you know, to help the AI know how hard to work. I mean, most people don't realize, today, it works exactly the same amount where it puts out, "The next word is..." And it doesn't think in advance. It doesn't think, "Okay, I have to write this poem. Let me figure it all out." [clears throat] No, it, it, it, it's generating word by word. And so getting it to go faster when it's easy or take more time, like math problems, you know, we, we all laugh because it, it, it's actually quite bad at certain type of recursive math things, like Sudoku puzzles, which are trivial. Uh, in fact, if you say to the AI, "Put this into algebraic form," and then give it to a tiny little solver, it's perfect. But its current algorithm is this single linear pass, print the most probable token, single-- And so there's a lot of control architecture work that's being played around with to deal with these quality issues. You know, clearly, humans, we think, "Okay, for this problem, it's just a joke, blurt something out. Uh, they're asking me for medical advice or marriage advice, I better, you know, really, um, think through the implications." And, and so the control system, uh, as well as the, the extra scale, is going to drive improvements at, you know, some rate that could be quite rapid.
- NKNikhil Kamath
Right. [upbeat music] No, thank you so much, Bill, for having taken the time with me today, and, uh, look forward to seeing you in May.
- SPSpeaker
Fantastic.
- NKNikhil Kamath
California. [chuckles] Thank you.
- SPSpeaker
[upbeat music]
- 29:03 – 31:30
Outro
- SPSpeaker
Hi, guys. We are in Delhi today. We're trying to shoot a new, a very interesting guest, and, uh, we're going to do something fun.
- NKNikhil Kamath
What else are we here for?
- SPSpeaker
What else we here for?
- NKNikhil Kamath
What else we here for?
- SPSpeaker
Did you know that your baby boy would save a son?
- SPSpeaker
No, uh, actually being serious, serious.
- SPSpeaker
Did you know-
- SPSpeaker
[laughing]
- SPSpeaker
-that your baby boy would someday... Pretty good. Pretty good.
- SPSpeaker
This is our beautiful, beautiful venue. We have Mark, Thad, Sheldon, everyone setting up lights. This is where everyone's going to be sitting. We're going to have a live screen monitor here, which I'll probably be looking at. A boom rod mic, and that's what we're going to place on top of the two speakers.
- SPSpeaker
Yeah.
- SPSpeaker
Hi, Chetan. This is the first time we're working together. What are your thoughts? You don't look stressed at all.
- SPSpeaker
... [laughing]
- SPSpeaker
[laughing] That's called a skimmer that we're going to put over the glass. That's our spot, and that's how big the skimmer is. I'm also worried about other things, so let's see then how it looks, and then take a call. We're going to change the rose. It doesn't look so nice, does it? [upbeat music]
- NKNikhil Kamath
Is that okay?
- SPSpeaker
Yeah, that was fantastic. [upbeat music] How does it feel?
- NKNikhil Kamath
Good. [laughing]
- SPSpeaker
[laughing] Nikhil. Little bit.
- SPSpeaker
A little bit more.
- NKNikhil Kamath
Huh?
- SPSpeaker
Say something great. [laughing]
Episode duration: 31:30
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode ckNEdxQ0Tc0
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome