Vinod Khosla: How AI Impacts Healthcare, Energy, Geo-Politics, Media and Climate Change | E1045

1. 0:00 – 0:17

   Intro

   1. VKVinod Khosla0:00

      Technology is very fun to talk about disruption. It's not a lot of fun to be disrupted. I do think we'll have to pay special attention to the people being disrupted, and their needs and desires and wants. (instrumental music plays)

   2. HSHarry Stebbings0:13

      Vinod, I am so excited for this. I've been absolutely loving

2. 0:17 – 17:57

   Artificial Intelligence: Hype or Breakthrough

   1. HSHarry Stebbings0:17

      your Twitter content recently. And so first off, thank you so much for joining me again today.

   2. VKVinod Khosla0:22

      Happy to do it. Always like talking to entrepreneurs.

   3. HSHarry Stebbings0:26

      That is so kind of you. But I want to start today on the, kind of, topic of the day, which is obviously AI, artificial intelligence. And Khosla and you have been one of the biggest proponents and, uh, conversationalists of the space in the last 10 years, really, a- as one of the first investors in OpenAI. And I just wanted to start with the question of, do you believe the recent hype in AI is due to technological breakthroughs and innovation, or is it the result of hype cycles keeping up with innovation from years ago?

   4. VKVinod Khosla0:57

      A little bit of both. Uh, let me give you a perspective, um, going back a long time. In 1982 at Sun, we adapted TCPIP and the internet as our native network. We were the only ones. Uh, by 1996, it was clear that exponential growth was going to continue, and we invested in Juniper, which, by the way, was a 2,500X return, a $7 billion return for us on a $3 or $4 million investment. Uh, when the internet really took off, and, uh, I remember investing, uh, meeting Marc Andreessen to invest in Netscape, the browser company, on a Saturday morning, uh, in Mountain View, um, when it was three people. It really took off, but the exponential was clear much earlier. The reason I bring this up is I wrote my first blog on, uh, AI, do we need doctors and do we need teachers? Uh, a little tongue-in-cheek, but, uh, encapsulating the idea that expertise could be free in 2011, 2012. It was pretty controversial then, but it was a similar moment to when we had, uh, adopted the internet and TCPIP in the '80s at Sun. And then when ChatGPT came out, it became like the browser moment for the internet, and, uh, it's now obvious to people the extreme capability that we've developed here and will continue to develop and is progressing at a rate where even I can't keep up. But anytime there's a rich area for development, bubbles form, and I worry a little bit that that will happen here. There's obviously real technological progress, real expansion of the opportunity set for real building businesses, but also a lot of hype that follows, you know, a l- or sort of a rich area to mine. Sort of the gold rush phenomena. What I'm fundamentally optimistic about is when you look at bubbles... In the 1800s, 1830s, there was a bubble in England when you got the right to build a railroad between two towns, you could issue scripts almost immediately. But... And it did, the bubble did burst. But more railroad was built after the bubble burst than before. My point is, I think the progress of AI applications in utility for society will go on a steady pace based on technical capability and entrepreneurial interest in it, not what Wall Street or, uh, investors think about it. So, I think investors are important, because they fund these developments, but investors cause bubbles, uh, the reality is the underlying technology and its applications.

   5. HSHarry Stebbings4:08

      You mentioned that OpenAI and ChatGPT being the equivalent of the browser in terms of the consumerization and the access that it provides consumers. You tweeted before, "The daylight between what we'll be able to do with AI and what we'll actually do with AI is human resistance." Can I ask, what did you mean by this human resistance, Vinod? And how does it show itself?

   6. VKVinod Khosla4:31

      Well, there's two types of things. You know, first, a genuine concern that AI will displace jobs, and what do we do with that? If I'm right, the level of job displacement over 20, 25 years will be large. And we've dealt with that kind of job displacement before, but over multiple generations. For example, in agriculture when, uh, it was 50% of all jobs in the US in, in 1900 and went to 4% by 1970. If the change is happening rapidly, then job displacement may cause us to have rules and regulations that slow down progress. Another example could be, uh, physicians resisting AI being able to do physician types of tasks. The AMA, um, is a very good organization, but, uh, you know, there is protection built in in certain areas, whether it's in unions or physicians or... Pick your favorite expertise, media, uh, where artists or music, production music producers may object to some of these. That's what I mean by human resistance, because people are being disrupted. You know, honestly-Technology is very fun to talk about disruption. It's not a lot of fun to be disrupted. And so, one has to be very considerate of the people being disrupted and accommodate them if we're going to make better f- progress. I do think we'll have to pay special attention to the people being disrupted and their needs and desires and wants.

   7. HSHarry Stebbings6:13

      I- I want to approach this from a very fi- first principles interviewer style and ask very basic questions, so I hope that's okay. What should we be doing, do you think, for those that will be disrupted? As you said, it's fun for us at the forefront of innovation, if you're being disrupted, less so. What should we be doing and how should it be handled?

   8. VKVinod Khosla6:32

      So, let's start with the math. Let's say the next 50 years of GDP growth without AI and some of the newer technologies would be 2% a year. Per capita income may go from 70 or 75, uh, wherever it is today, or 65, I forget the exact number, to about 175,000. If we change GDP growth to 4%, th- that per capita income would increase to 475,000 or so. My point is, we will have adequate resources, adequate, uh, abundance to share much more broadly. I wrote, uh, a blog in 2016, um, in Fortune magazine that said AI will cause... And this was s- six years, uh, seven years ago. AI will cause great abundance, great GDP growth, great productivity growth, everything economists measure, and increasing income disparity. It was the first time in 2016 I wrote about needing some sort of universal basic income, uh, which means redistribution of income in a much better way than we have today. The reason I gave you my calculations of the effect of changing GDP growth from 2% to 4% over 50 years is we will have enough to share. So, my suggestion would be, some large percentage of per capita income growth, or GDP growth be reserved in a social security pool for our universal basic income or something like that. There's many ways to redistribute, some more acceptable than others. But I think happiness in our society will be a function of equality, not just GDP growth. And I think we'll have to consider that, and AI will make that p- abundance possible. So, I'm optimistic it can be done. Bill Gates has talked about a robot tax. Um, there are multiple ways to approach this. But I think reducing the rate of increase, uh, and putting the excess, uh, into a pool for s- like a social fund would be the right approach because increments are much easier to share than, uh, somebody saying, "Well, let's increase taxes on everybody." I'm fine with tax increases personally. I do think, uh, we can afford to have a better tax system, more equitable tax system. But there are maybe better ways to do this, uh, and increase sharing of the benefits of these technologies, um, through other mechanisms.

   9. HSHarry Stebbings9:24

      Do you think AI does more to harm or to help income inequality?

   10. VKVinod Khosla9:29

       Well, AI is a tool for productivity growth. Um, capitalism does result in more inequality, a- almost always. Everybody's better off, but the income i- inequality increases. That tends to be what history has shown us. And this will, AI will only accentuate that. And so sharing the benefits of AI in a different way will be important. I think it's not by constraining AI or not, uh, using AI to benefit society. It's by taking the product of these technologies and sharing it more equally. The increased abundance should be more equally shared.

   11. HSHarry Stebbings10:16

       I'm fascinated. Again, I said I loved your tweets. I really have loved them. One of them said, "The choice is, do you let AI do the jobs or not? Some will and some will be laggards." When you think back on that, who do you think will embrace it, and who do you think will be the laggards?

   12. VKVinod Khosla10:34

       I worry that the Western political system, uh, and capitalism is by permission of democracy. Uh, so the Western political system will pay more importance to the short term, which is things like job displacement. While in China, you might get Tiananmen Square-type tactics to adapt it and win economically as a country, as opposed to each person being fairly treated.

   13. HSHarry Stebbings11:08

       I- I- I totally understand that as a concern. I think the hardest thing is political incentives, which is every politician is incentivized to remain in power, and actually encouraging job displacement will not help them in the next election. It might help-

   14. VKVinod Khosla11:21

       Right.

   15. HSHarry Stebbings11:21

       ... the company economically in 20 years, but the incentive mechanism of politics doesn't align to the GDP and per capita growth over the long period.

   16. VKVinod Khosla11:32

       Absolutely true. And if the change is slower, it'll be easier to accept than if it's super rapid. I'll give you an example. I'm pretty convinced instead of displacing physicians... Uh, and long time ago, I wro- m- wrote a blog called, uh, 20% Doctor Included. We could quadruple the number of interactions between a physician and patients. In fact, it's already two in Australia. There's many more interactions with your primary care physician, um, every year for the typical patient than we do in the US.... so you could increase the level of service, increasing the level of attention with AI taking excess of that. Uh, there's projected to be, under current course and speed, a shortage of primary care physicians in this country. I think we can have an excess even when we increase the number of interactions by 5X because AI can take 80% of that increased load. And, and that's a good example of how one can work cooperatively to improve the level of service in healthcare and primary care while, uh, not disrupting physicians.

   17. HSHarry Stebbings12:47

       Do you believe then that AI will be a competitor or a complement to the doctor? And do you believe doctors see it in the same way?

   18. VKVinod Khosla12:55

       You know, I wrote about it, uh, uh, probably eight years ago. I wrote this paper, 20%, Doctor Included. It's about 100-page document. And I wrote about the human element of care being provided by doctors and the detailed expertise being provided by an AI. You can't expect your oncologist to be caring, talking about your cancer, and have read the 5,000 most recent articles, uh, on your cancer. They can't keep that knowledge in their head. So, that's the kind of partnership that'll be possible. You know, you don't want a computer- a computer telling you you have cancer. Uh, you want a caring human being. I jokingly said to the dean of the Harvard Medical School about five or six years ago, they ought to take- change their selection criteria for more empathy, more mirror neurons, and less IQ only, which is what it takes to get into Harvard Medical School. I said, "Take the model of the UCLA film school and look at caring humans as criteria number one for admission to the, uh, M.D. program."

   19. HSHarry Stebbings14:12

       What other ways do you think AI will most significantly change healthcare and medicine in the next 10 years?

   20. VKVinod Khosla14:19

       Well, for example, we can custom design proteins for particular purposes. I'm pretty excited that for a large number of genetic defects, and there's probably 6,000 or so diseases that are single mutation genetic diseases, we will be able to design, with AI's help, custom solutions because we've essentially solved the protein folding problem. Uh, we can fully characterize the design and behavior of DNA, RNA, proteins. And that's all really, really exciting, so that'd be a huge impact on the pharmaceutical side, not the expertise side of healthcare. Almost every area, whether in healthcare or in media, will be exciting.

   21. HSHarry Stebbings15:07

       Can I ask you, you mentioned before a- a brilliant thing, which is, "I've never seen innovation come from large institutions." And it's really applicable across, uh, medicine, across, uh, education, across music, but do you think in this next wave of innovation due to AI, that holds true? Or does AI actually allow incumbents or the dominant parties to just become stronger?

   22. VKVinod Khosla15:33

       I believe innovation will be spread broadly, uh, in different areas. Now, large language models, uh, may or may not be constrained to the large platform companies of ri- right now, it appears that'll be the case. There'll be two or three major providers of large language models. But for example, uh, we have an effort in symbolic logic to see if in three or four years, it could change the equation of how you train these systems and the, uh, and the amount of data required and all those kinds of things. We have an effort at developing curriculum for AI instead of curriculum for humans. Almost everything AI's training on is designed for humans, not for machines. So, there's a number of large innovations, and it's hard to tell today which ones will play out, work out, um, or be important in three or four years. So, when we originally committed to OpenAI, uh, four or five years ago, uh, it was very unclear where the world would lead and what the capabilities would be, and when they would d- develop. It was clear to me that there was a reasonably high probability of pretty significant advancement at some point in time. Hard to say whether it would be 2022 or 2025, um, but that didn't bother us. The shape of the curve and the exponential really interested us. I think there's other approaches to AI. Um, you know, Josh Tenenbaum, for example, working a lot on probabilistic programming, others working on formal techniques like symbolic logic. Do they add to, replace, complement, um, uh, LLMs? I think all those are possible, uh, and within LLMs, large innovations are possible. So, I would say the future is not predictable, but it is discoverable, and what I'm really optimistic about is the best minds in the world are now working on this problem. And I think that's more a precursor of progress than any single thing, like, "Hey, we suddenly have a trans- new transformer-like paper," or "'Attention Is All You Need', the famous paper." It's more about the talent working on it that makes me excited.

3. 17:57 – 20:51

   Predictions and AI's Potential

   1. VKVinod Khosla17:57

   2. HSHarry Stebbings17:57

      What do you think are the barriers that we need to overcome for AI to have the impact we all want it to have on healthcare?

   3. VKVinod Khosla18:04

      Well, in healthcare and other areas, uh, AI is doing some things really well, other things not so well, and I think that's where development needs to be...... rapid. For example, in healthcare specifically, guardrails is a very, very important issue. Safety is a very important issue. So, a lot more attention has to be paid to those factors. Explainability is a huge issue. Those things will need to be developed by somebody applying AI broadly to, say, primary care or oncology or pick your favorite area, cardiology. But all of those, uh, I'm pretty convinced will happen in the next five years. Uh, and we'll see the capability develop to provide very, very low cost, much more frequent and accessible care. And accessible both im- a- accessible 24/7 and accessible at a very cheap price to every s- m- m- human being on the planet, I think.

   4. HSHarry Stebbings19:14

      Vinod, y- we've spoken about healthcare. You've written... or we'll talk about education, but you've, uh, written extensively about both many years ago, ahead of time, seeing things before anyone else did. To what extent do you think about market timing when investing? Because, yeah, I think it's, um, Marc Andreessen's quote, "There's no such thing as a bad ti- a bad idea, only a bad time." Do you think about market timing?

   5. VKVinod Khosla19:39

      Um, you know, you think a little bit about it, but I mostly feel I'm not smart enough to predict it. In retrospect, you always feel dumb about some things and really smart about others. You know, we invested in OpenAI and we invested in Commonwealth Fusion about the same time, and I'm really happy about both, but, you know, both felt like longer term bets than they've turned out to be. And progress has been more rapid, both in AI and in fusion, uh, very different areas, uh, than I would have expected. Suddenly, I feel smarter, maybe even overconfident, uh, about my predictions. Um, you know, today, I sit here and say, for example, the amount of compute power AI systems need, uh, the expense of training. Could one have analog computing, and we are working on that, uh, dramatically change that equation by 100X, both cost and power? And I think that's entirely not only possible, but likely. When's a little bit harder to predict, though we are chipping, tapping

4. 20:51 – 29:17

   Energy Solutions and Global Impact

   1. VKVinod Khosla20:51

      out chips this year.

   2. HSHarry Stebbings20:52

      Now, I, I, I, I have to ask, you mentioned fusion there and you mentioned Commonwealth. I, I thought it was fascinating when I thought about kind of the future of energy and when I was doing a lot of work prepping for this. Because you've said before, "Forcing noneco- noneconomic solutions is the wrong approach to climate and solutions will need to be at what I call the Chindia price. My bet is on fusion and geothermal." I just wanted to break this down 'cause I didn't understand some of it, if I'm honest. Which is, why is forcing noneconomic solutions the wrong approach first?

   3. VKVinod Khosla21:25

      Well, energy generally is such a large-scale problem that there's no way, you know, we will spend the money to scale it. It's great for the first 1%, 2%, 3% of the market, whether it's the electricity market with solar and wind or the hydrogen market, or pick your favorite. I believe subsidies should be declining with scale. So they can be high when it's 1% of the market because the tax on the overall energy costs would be low, but they have to start scaling down and no subsidy is being designed that way. We are... as it gets to 2% or 4% or 8% or 15% of a segment, whether it's hydrogen to replace natural gas and coal or, uh, sol- uh, solar or fusion to replace, uh, electri- electricity from coal, it has to scale down. If it's not, uh, it's going to become too expensive. Do you get clean energy and no healthcare spending? Uh, what are you going to trade off? Uh, those are the kinds of practical things at scale that matter. Uh, having said that, look, early subsidies help get technologies developed when they're subscale, but by the time they are 5% of market, they really shouldn't need subsidies at all. Also, practically, if you're a politician in India, you're not gonna tell people they can't have clean water but their kids will be better off because there'll be a less warm climate. It's a very hard political trade-off to make. You know, coal consumption's been increasing in some of the developing parts of the world. You can't defy the laws of economic gravity at scale. You can do it in small startup sort of situations, but not at scale. And for it to be contributing positively to economic development in countries like India or China, uh, it has to be competitive globally with other sources or it won't be adopted. And this leads to a very important point in climate. We don't need incremental solutions. Um, we need large breakthrough technologies will make, uh, these new cont- uh, new technologies competitive at scale. So we can take small risks and have incremental decline in costs, but that doesn't solve the problem because it won't be broadly adapted. Or we can take larger risks and if successful, really have solution fusions, that kind of thing. Deep geothermal, uh, 4 or 500 degree centigrade geothermal would be that. Finding a new geothermal site incrementally doesn't really help.... it's not scalable, it needs to be scalable by 100 X more than exists today, at least. And, and so my interest in technology is that scale, and are economic and don't defies the law of, uh, economic gravity when scaled. Aviation fuel's a very simple example. And while we are talking about it, you know, take it, these take a long time, so sometimes the venture community ignores them, but they're profitable. Uh, if I look across our portfolio, some of the investments we made more than 10 years ago are, you know, there's a whole number of them, QuantumScape, Verdagy, LanzaTech, Impossible Foods, that are five to 10 X multiples, um, though they took a longer time to get there. So, Climate 1.0 was successful, if one was patient and picked the right technologies, things like fusion, new plant proteins, um, I'm very excited about new fertilizer production techniques, new steel production techniques. We are taking radically different approaches in cement, steel, um, those are really, really exciting breakthroughs, and there's no question they will be cheaper than its fossil alternative. Already, I can say the cement we are producing will be 70% less carbon, and our first plants will be operational this year, and cheaper than traditional Portland cement. Uh, and that's when things start to scale.

   4. HSHarry Stebbings25:58

      Can I ask you, on fusion, wha- what is at the core of the bullishness, and how does fusion bankrupt entire industries and destroy massive companies, like you've said before?

   5. VKVinod Khosla26:09

      There was one breakthrough that enabled fusion to happen, which was high temperature superconductors. It wasn't possible to build a fusion reactor with a 20 tesla magnet, which is what Commonwealth Fusion is doing. With that magnet, the size of the reactor scales to one over tesla to the power of four. So if you quadruple the power of the magnet, which one can do with high temperature superconductors, the reactor itself becomes 1/250th the size, so 250 times smaller. Because of that, uh, you can now both be economic, but also experiment much more rapidly, build them much more rapidly. Commonwealth Fusion was able to build a 20 tesla magnet for fusion and demonstrate it in three years, three and a half years. It has taken the large ITER project with smaller magnets 25, 30 years to get to the same place, so 10 X faster. I believe demonstrating positive energy by '25, maybe '26, will be also 10 X faster. Not only has that happened, there's a dozen other credible projects in fusion, um, that have been started mostly because Commonwealth instigated this change towards fusion, and I have this theory you need instigators, and it's got, uh, sparked up a field that was slow and not attracting the best talent to attracting the best talent now. And there's probably a dozen startups and that's what makes me optimistic. Uh, at least one or two or three will be successful in this.

   6. HSHarry Stebbings28:00

      How does fusion, how does geothermal, how does new energy resources and innovations within kind of their extraction, how does that rearrange international relations, do you think?

   7. VKVinod Khosla28:11

      Well, if we can reduce our dependence on oil, we can dramatically reduce international tensions and strate- what are called strategic interests. There's so much fighting over strategic resources like energy, the minerals and metals is the other area, um, we can eliminate that tension. Imagine the following, if you take every coal plant in the country and its f- its boiler, the coal boiler or a natural gas boiler is replaced every five to 10 years, they deteriorate pretty quickly. If you could replace it with a fusion boiler, you wouldn't have to build 5,000 new power plants in the US. You'd just have to build 5,000 new boilers, which is, uh, less than the number of Liberty ships we built in five years during the Second World War. So, that's how you achieve these at scale and completely change industries and our source of power, by being clever about

5. 29:17 – 33:37

   Societal Changes and Immigration Policies

   1. VKVinod Khosla29:17

      it.

   2. HSHarry Stebbings29:18

      I had Larry Summers on the show not too long ago, and he said specifically with regards to kind of energy and its impact on, you know, international relations that China is a prison, Japan is a nursing home, and Europe is a museum. Can I ask, do you agree with that statement? And how do you think about that with regards to the discussion we just had on how energy changes international power flows?

   3. VKVinod Khosla29:43

      I think what Larry was probably talking about was demographics. Um, Europe is an aging demographic. China, uh, ch- China and Japan are rapidly aging. I think, if I read right, by 2040, which is 15 years, 17 years, years from now, um, the average age in Japan average will be 54 or something like that. The world needs younger people to have a vibrant society, and what is called the dependency ratio, the number of people working to support each person not working, whether they're elders or kids.... that dependency ratio in Japan, China, most of Europe, is heading in the wrong direction. The US has a reasonable demographic, India has a great demographic, you know, the Mid East and Africa have good demographics. But this, uh, anti-immigrant sentiment will hurt these, uh, European countries from, uh, eh, really renovating, you know, invigorating their societies, uh, that's a complex problem.

   4. HSHarry Stebbings30:55

      Why are you so bullish on India moving forward? And will India develop into a tech giant, do you think, in the next five to ten years?

   5. VKVinod Khosla31:04

      What I, I love about India is the demographic is young, with a huge focus on learning and education. Now, uh, India's pretty significant problems, I think, leaving a population like the Muslim population behind in India would be a huge disaster and could completely change the optimistic projection I have for India if, if you leave a part of the population behind. And that becomes a source of violence, rioting, things we've seen. But generally, a well-educated population and a course set of technologies, um, and English as a language, a common language, western values, democracy as a value, all those make India a pretty good candidate. As does, uh, uh, you know, Brazil's, uh, similar in many ways. Um, so India's not the only country with that kind of advantage.

   6. HSHarry Stebbings32:09

      Do you think the US and the rest of the world will be increasingly reliant on Indian-born tech talent, given what you just mentioned there in terms of the amount of young people and their talent?

   7. VKVinod Khosla32:21

      I think the great thing about technology is you need to develop it once and deploy it 100 times. You know, you go discover an oil well, you run out of it, you can't scale, build a million of those. If you build a, a, a bipedal robot, you can build a million of those. So yes, technology talent is going to be important to make the world a better place, I'm very optimistic about the role technology can play. Now, there's dangers to that too, but I'm very optimistic, whether it's in healthcare, we've talked about it, climate and energy, we've talked about that, music and media and content generation, uh, lots of places. But the talent can be anywhere, and I think India will be an important source or can be an important source with the right policy. Everything's subject to right politics and right policy in the country.

   8. HSHarry Stebbings33:18

      You mention music there, and content. I, I have to ask about that. You know, you, you've been a long supporter of, uh, kind of content creation by AI, specifically in music for many years. And you tweeted before, "Music won't require personal musicians in 15 years, and your favorite artist will be a personal algorithm."

6. 33:37 – 36:15

   The Intersection of AI and Creativity

   1. HSHarry Stebbings33:37

      Can you take me inside your mind here, Vinod, in terms of how you see the future of music with AI at the heart of it?

   2. VKVinod Khosla33:44

      So we invested in this area also five years ago, and some founders from Australia came to me and said they wanted to build a top 10 music hit generated entirely by an AI. We're not there yet, but, um, you know, Splash.AI will let me be a rapper. In fact, I was able to sing a rap song at my daughter's wedding, um, because of AI. I'm a terrible singer and a terrible, terrible rapper. I do think, you know, the Taylor Swifts of the world will still be there, I'm not suggesting they will disappear, uh, artists and phenomena will be there. Their creativity will be dramatically enhanced by AI. It's like getting a much more powerful tool at their beck and call. They can try 100 acts, more variants of things they like or don't like or wanna guide the experience to. So, a Taylor Swift will always be an experience. But the Splash.AI has become one of the largest music games on Roblox, where every 13-year-old can also create their own music. And I think, at some point, you might want music for your brain, the way your brain works and the music you like, for a quiet walk, or when you wanna get energized before a athletic event. It'll be personalized to you, and can be personalized. It may be an adaption for something Grimes sang, or something from the old days of Elvis, or it could be something brand new and completely original. I think people are very narrow-minded when it comes to imagining how music might be used and enhance peoples' lives, uh, with the help of AI. Um, I think it'll be a powerful tool, both for musicians and individuals, empowering them to do things that they couldn't. If I don't have the voice, doesn't mean I shouldn't be able to play that music.

   3. HSHarry Stebbings35:51

      How do you think about the challenge of verification? With the ability to create s- content that can be very similar to others, verification becomes the most important thing, making sure that it is authentic. How do you think about that and media rights in a world where content supply could really be infinite?

   4. VKVinod Khosla36:12

      Look, that's a tricky problem. Uh, funnily

7. 36:15 – 39:07

   Challenges and Opportunities in AI Technology

   1. VKVinod Khosla36:15

      anticipating this world, my wife runs a non-profit called ck12.org. It's meant to design con- free content, open source content for everybody, and a personal AI tutors for every student, uh, that are near free or free. And multiple years ago...We changed our open source license to include a prohibition against derivative works. So, we looked ahead a few years and said, "Our content will not allow for derivative, uh, k- work, including derivative algorithms that use the content to train on." So, we were forward-looking. Now, the law in this area is not clear at all. Um, so in Splash.AI, we've trained the music from scratch. There is no YouTube music or any artist's music used in training our AI. And it's completely IP free of any artist. I'm excited about music models that don't rely on any IP from any other source. We authored our own loops and stems and st- w- voices and all that, and we will also license voice, uh, voices from artists. If Grimes is willing to license her voice, which she has, then it'll become a thing our users can use in... as they personalize their music with her voice. I think it'll dramatically im- enhance productivity and creativity, both in music, but also in content, in storytelling, in, um, uh, you know, your brand logo, uh, just almost everything.

   2. HSHarry Stebbings38:00

      Where do you think the real money to be made here is? Is it in the aggregator, your Spotify of the world? Is it in the rights holder, your Universal? Is it in the, that company that creates it, the creation tools? Where does value accrue in the future of music as, as you kind of described it there?

   3. VKVinod Khosla38:17

      Look, in the end, human beings love originality. So, where that originality comes from, little hard to say. Will AI get more creative than your typical artist or will artists be dramatically enhanced to have their creative, creativity be amplified? Hard to say. But I think that's where, uh, uh, uh, value will accrue from. I'm not sure agg- aggregation will be as important. It might be. But definitely at every part of the stack from the silicon hardware that runs these things, um, to aggregation and anything in between, there will be value creation.

8. 39:07 – 47:19

   Quick-Fire Round - Reflections and Personal Experiences

   1. VKVinod Khosla39:07

   2. HSHarry Stebbings39:07

      If it's okay with you, Vinod, I'd love to move into a quick fire round. So, I say a short statement and you give me your immediate thoughts. Does that sound okay?

   3. VKVinod Khosla39:16

      That sounds fine.

   4. HSHarry Stebbings39:18

      So, what single piece of content... A very hard question, apologies for this. But what piece of content has had the most impact on you?

   5. VKVinod Khosla39:26

      A few years ago, I went to a Taylor Swift concert and the concert was, concert was great. Uh, and she's on her new tour. But standing there watching the experience and the performance, I could think of five ways, uh, in which, uh, you could enhance that experience. For example, with the display behind her where she was much larger than life, you could see fully holographic displays. You could see music enhancement in many ways. If I were to pick one thing, I would say on the entertainment side, that had a very large influence on me. In the AI field, um, seeing some of the early examples of, wow, that's now possible, even if it's not reliable yet, uh, that was pivotal for me and got me convinced AI was at a stage of a Cambrian explosion. I remember vividly going to the MIT Plasma Fusion Lab, uh, with Bob Mumgaard, the founder of Croy- uh, Commonwealth Fusion before Commonwealth Fusion existed, and he had me stand on a chamber and he said, "Below you was 600 million degrees a few seconds ago." Th- those kinds of moments are pivotal, um, and it in- inspires you to say, "How can we make this even more pragmatic, more scalable, more real, more creative?" The whole bit. Sorry, not a rapid fire question, but... (laughs)

   6. HSHarry Stebbings41:05

      Vinod, it's, it's a joy to hear you describe it, so please, do not worry. You mentioned Marc Andreessen at the beginning with that Saturday morning meeting. We see Marc Andreessen today, a larger-than-life figure. What was Marc like back in those early, early days when you backed him with Netscape?

   7. VKVinod Khosla41:21

      You know, it's hard to remember 1996. Um, I know it was Castro and El Camino, uh, h- in Mountain View. He was definitely intense. I remember coming out of the meeting calling John Doerr and saying, ju- "Let's just do this." Remember, the internet was this ambiguous thing. It was one of these triggering moments I call instigations for me. That's what he was, uh, like. It was... it didn't take very long to decide we should be backing, uh, uh, Marc and Netscape.

   8. HSHarry Stebbings41:54

      Tell me, how do you see the future of transport?

   9. VKVinod Khosla41:57

      You know, so I have my favorite project. I believe, as an optimist, we can replace most cars in most cities in the next 25 years, as ridiculous as that sounds. I will only suggest the internet sounded ridiculous in its scope and st- uh, impact in 1996. I think email sounded ridiculous in 1990. PCs in every home sounded ridiculous in 1985. Computer in your hand, the iPhone sounded ridiculous in 19- in 2006. So, we've gone through these moments. ChatG- uh, GPT founded, uh, sounded ridiculous five years ago. So, I do believe we can replace most cars in most cities. We have a little startup called Glidewise.... which uses self-driving the wrong way. I think the Waymo approach and the Cruise approach is nice. It adds functionality. It's not transformative of cities. Self-driving, uh, cars used in public transit will make public transit cheaper and faster and point-to-point and on demand. That's what we're building at Glidewise. We bid on three projects so far. None of them projects we were invited into, these were public transit bids, including the one connecting the San Jose Airport to the new Google campus to the Apple campus. We bid, lobbed in a bid over the transom, um, uh, uh, over the transom and won it outright to build a public transit system. People say, "That's crazy, uh, for startups to do." But we won three bids now, uh, in a row. We haven't lost one yet. I'm sure we will lose something as a startup in public transit. And remember, these are 50-year revenue streams. Very, very long revenue sources if we can get there. Hmm. Of course, that's always a risk. So I'm very excited about transforming the nature of cities because the one thing fixed in cities is street width. Anything that allows much more throughput, more passengers per hour per foot or meter of lane width is what's needed, uh, because that's the one thing in cities that'll never change. And I think gradually we will have... go from Uber can't pick up and drop off on Market Street to h- cars can't drive on Market Street. Up and down, there's public transit or one lane. We designed a public transit system to be bicycle lane width and have more throughput than light rail. That's exciting. I get very, very excited about that kind of an idea.

   10. HSHarry Stebbings44:47

       You've mentioned some of your incredible investments with Khosla before in this show.

   11. VKVinod Khosla44:51

       Mm-hmm.

   12. HSHarry Stebbings44:52

       When you reflect on the ones that you haven't done, which one stands out to you as one you didn't do, but wish you had done?

   13. VKVinod Khosla44:59

       Oh, t- there is so many. I remember in 2007, a journalist asked me about, uh, Twitter, and they interviewed me and they sourced their questions on Twitter in 2007, I think. Um-

   14. HSHarry Stebbings45:14

       Wow.

   15. VKVinod Khosla45:15

       I think maybe in early 2007 at Davos or something. Um, and I said, "Huh. Interesting." But couldn't see the potential of it. That was duh. My biggest mistake ever was I was at Sun and Cisco routers was software running on Sun machines and Sun's logo, uh, or tagline used to be, "The network is the computer." And I didn't see and nobody at Sun saw routers as more, a- as maybe a huge thing, even though we had adapted networking from day one. By the way, the first computer to have standard networking from day one. Um, in fact, in 1982, we had hardware encryption on every computer, uh, because it was networked. Too early matters, I guess. Um, but, uh, yeah, I've made so many mistakes. I, you know, I like to say I have made more mistakes than almost anybody I know. But I'd rather try and fail than fail to try, which is what most people have done.

   16. HSHarry Stebbings46:24

       That's exceptional. I love that. Um, and I'm probably gonna get killed for asking this from your team, but I have to anyway. People chase money, honestly, Vinod, most of the time, uh, in a lot of society. Um, you've been incredibly successful. Does money make one happy?

   17. VKVinod Khosla46:41

       All the research is very, very clear. Money helps up to a certain point in society, and then is not hugely correlated to happiness. I do think passion for projects makes you happy. Passion for a vision, passion for an area, whether it's music or it's a scientific breakthrough. That I think makes people's life meaningful and happy, and that's sort of what I strive for.

   18. HSHarry Stebbings47:10

       Vinod, I've loved this. I can't thank you enough for putting up with my questions. You've been incredible. So thank you so much for joining me today.

   19. VKVinod Khosla47:16

       Thanks very much. I very much enjoyed it.

Episode duration: 47:19