Inside the mind of Elon Musk & is the future of EV market Chinese?

0:02Hello and welcome back to the Times Tech Podcast with me Katie Prescott and you Danny Fortson That's Katie in the city and Danny in the valley for those who know. Now, Danny, a couple of weeks ago, we were talking about Elon Musk's grand ambitions for interplanetary travel. I do recall this, but I don't recall. What did we conclude? I think we thought it was a great idea. Keep going, Elon. But ever the interesting character in tech and beyond, we thought it'd be good to dig down a little deeper into how he's been spending his time on Earth over the past few months. Yeah, so today's episode is going to be slightly different from usual, because we're going to be joined by John McNeil.

0:44And he served as the president of Tesla for a 30-month period, about two and a half years, when revenue went from$2 billion to$20 billion. And we're hoping that John, having worked at the kind of the right hand of the king, can help us go inside the mind of Elon. Yeah, because he did work at Tesla at this really key period of growth, but it was a while ago. So he was there between 2015 and 2018. And since that, he's been the COO of Lyft, the ride-sharing company. He's a board member for General Motors, and he was the vice chair of the board for Cruise, which is the now defunct robotaxi business that General Motors used to own.

1:24And then, of course, there's AI infrastructure. He's quite into that as well. So there's going to be tons to get into here. Yeah, fascinating, meandering conversation awaits us. John, welcome to the show. We have so many things we want to ask you. Yeah. And I think, I'm afraid, we're going to start with Elon Musk. You don't say. If we may. Does everybody do that to you? It's a good cocktail party trick, for sure. I'm sure. I'm sure. This is an interesting one. I mean, you worked at Tesla. Gosh, I mean, it's quite a long time ago now, right? 2015. It's already quite a long time ago, yeah. 2015.

1:57So we had one product in the Model S. So yeah, it's been quite a long time ago. And what was it like working with him? I had to say, like, so surprisingly, maybe what you don't hear a lot in the press is the guy is hilarious. He I can't tell you how many times I was crying in meetings because I was laughing so hard. He's super funny. And all the things you know about Ilan, he's like super focused and really, really incredibly driven, doesn't suffer fools at all. Those are all true. But he's pretty fun to travel with. And when he's in his comedy mode, pretty fun to be around. I remember there's some leaked emails from Elon around kind of what is okay to do inside Tesla.

2:38And it's really all around efficiency and cutting through the nonsense, et cetera, including just like if you're in a meeting and you deem this meeting useless, then you can just walk out and that is okay. Did you actually see that happen? Because I'm always interested in kind of you hear our company culture is X, Y, or Z, but then what that looks like on the ground. Yeah. Like, so those emails were all, as you said about efficiency, maybe the one of the more famous ones is about communication where it's just about, it gives you permission to go talk to somebody who's in the direct, who's basically a direct line of fire into the problem.

3:14And so rather than going up the hierarchy and back down the hierarchy, forget all that, just go directly to the person you need information from and skip like getting permission from your boss, their boss, all that sort of stuff. And so that was done regularly because people had permission just to be super efficient communicators and kind of ignore the hierarchy and the politics. And that kept the politics like really at a minimal level. And I think the point around the meetings that he made was if you're not if you find yourself in a meeting room and you're not a decision maker, get up because only decision makers are people having direct input into that decision should be there.

3:53Otherwise, it's a waste of your time and you can't be productive. for that hour or two hours or whatever. So get up and walk out. And oftentimes he, I mean, the finance team was famous for having a bunch of people trudge into a meeting, like dozens. And he would literally point at each one. You a decision maker or do you have input for me? No? Okay, you can leave. How about you? Next? And he'd go around the room until it was whittled down to maybe five or six people. Right. And the point got across, like super fast. A little slower in the finance team, but it got across. I mean, it's kind of, um, these are my words.

4:29It's a little bit of a, a kind of a management meat grinder. This feels like a lot of people get in there. It's super, super intense. And then people get burned out or they leave or they just can't take it anymore or whatever. Did you find it like that as well? Because again, I mean, we can talk about Elon's political role, but he has created this company that went from zero to the most valuable car company in the world and kick-started the ev revolution none of that is easy yeah none of that happens without him i don't think but to your point like the i think why was there the average turnover or the average kind of life expectancy of a senior exec so there were like seven six seven people on the senior leadership team on that team the average life expectancy was 14 months so when i entered my third year i already been through like two management teams you're a lifer exactly all of a sudden like i'm i've been here three years i'm looking around the room uh and there's a couple like long-termers on that team but i was on the third cfo on my way to the fourth like to your point is it's super hard, I think, to keep the pace for the long term in those roles and in the roles that report to those roles, because those people are kind of the next people up.

5:47And so it's intense for sure. When I decided to leave, it was not because I was exhausted, but because I was really having a problem with the leadership style that involved cruelty. And I think a number of leaders self-select out of that because after a while it just becomes something maybe something for my in my case you just don't want to be a part of it we had sam altman on the podcast a few months ago and obviously the two of them have fallen out quite publicly but altman said that elon musk has changed a lot since he first knew him do you think that's the case i i think so like he's he's clearly very different from when we worked together and so when we worked together we were trying to actually believe it or not build his profile because we had a demand problem at tesla and we needed to create demand outside of the super early adopters and we were going from producing a few thousand cars a quarter to 10 000 cars and then 20 000 cars and so we were literally working on how to build his followership on twitter believe it or not in 2015 and he was shy i would say, and not all that comfortable with us using him as the amplifier for a lot of the messaging.

7:01But it was, it's, it's clearly a different Elon now and that we see. And I think he's, he has changed in a lot of ways. And I think the world stage changes you for sure. And so that was it. That was an actual strategy that the top team was behind. Okay. We need to create Elon as the totem, as the kind of leader of this movement as a way to generate more fans of the company. Yeah. Like when I, I remember when I was getting to know Elon and he said, Hey, if you take this job, one of the responsibilities you can have is demand. And there's a constraint there that I think you need to be aware of. I said, what's that?

7:36He said, the marketing budget for this company is exactly zero. Have you done that before? And I said, no, I haven't. Um, so we got to figure out like how to do that. And we knew that like, we were at a big, big time disadvantage versus the rest of the car companies we're competing with. Like it's super competitive industry. And we were small, so we didn't get the benefit of like supply chain pricing on volume, we were buying tiny volumes. And so we were paying through the teeth for everything that we bought. And we knew that like, the big car companies were spending$3 ,000 a car on marketing.

8:09And we had to make that up. Basically, we were spending$3 ,000 more on parts. So we didn't have any marketing. And so we decided that we were going to do earned media and we were going to do a lot of organic stuff to earn that media and then really use Elon as the megaphone. And so I forget how many followers he had when we started, but it wasn't many back in 2015. And we knew the space folks, that whole cohort were really into him. The SpaceX folks. The SpaceX fanboys. And so we started to build a bunch of posts that would build followership in space. And then once we got space, then we moved a lot of that messaging into cars.

8:52And we got the early adopters and then it really started to wind up. It was our only way to market was we would come out with a new feature like all wheel drive and we would have him talking about all wheel drive on Twitter. And then we'd just be repurposing it everywhere. So you'd be like, you created the monster. I don't think I created the monster. I don't want to take any credit for that. But we did get a company, you know, we grew it from$2 billion in sales to$20 billion in sales in 30 months, largely on the back of that. And it's that kind of story that is the reason that, you know, he's been touted as a good person to lead the doge and strip out costs from government.

9:30What have you thought when you've been watching him do that role? And the distraction, I guess, that is also created for him within Tesla. Yeah, I think like when you're taking costs out of one of your own companies, it's a way different environment than you're taking costs in Washington where there are a lot of long knives out for you. And in almost every kind of different vector, whether you're threatening people's budgets, you're threatening their constituents, you're maybe threatening their power, which is the currency in Washington. I think that's a completely different ball of wax. I'm not sure he was prepared for all the long knives that were going to be chasing that effort.

10:05nor did I think he was all that sensitive in the way he handled people in the cabinet and their agencies and their domain as well. So like, I think there's, there's probably ample blame as both sides do a postmortem on that, but, you know, delivering a hundred to 200 billion and there's real questionable accounting around that against the goal of 2 trillion can be characterized as a victory. You mentioned earlier, cruelty and that being part of the management style or management ethos. And that seems to have really been amplified in the way he's carried that out. Can you talk a little bit more about that cruelty aspect?

10:46How kind of central is it to the way he runs his organizations and treats his people? And then I'd like to, I think we can move on. I don't know if you agree, Katie, to kind of what's happening in the EV market today, because that's also completely fascinating. Yeah, I don't think it's central at all. I think it pops up. And I think Walter Isaacson in his book did a good job kind of describing why, like I have empathy for him. He is chased by a lot of demons that come out of a really rough childhood in a really rough teen years. You know, there's the old adage that those who are bullied become bullies.

11:22And that would pop up every now and again. And there's a story in Walter's book about a guy who was just, he was not even involved with a machine at the Gigafactory, sort of getting fired randomly and then getting rushed out or chased out of the factory. And both the co-founder of Tesla, J.B. Straubel, and I are reacting to that in real time, saying this isn't right. And he wasn't even involved. And so I don't think it's central, but it is something that pops up. And it pops up. And again, I've got empathy for why it pops up. Right. Should we move on to talking about the EV market then? and I'd be interested to hear your reflections on where the industry is at at the moment.

12:01I think it depends on where you are in the world. Like I was in Shanghai and if you're standing in China, like every car that goes by is an EV. Every car, like when you're in a car on the highway, every car is an EV. And so if you're standing in China, like it's an EV world for sure. And similarly, if you're standing in the Scandics in Norway, it's an EV world in particular. It is slowing in Europe and it certainly has slowed in the US. I never would have thought that the drivetrain of a car would become a political issue, but it has. And the reality is that EVs were sold and Teslas were sold in zip codes that were 80 % plus Democratic zip codes.

12:44And so that is a real, real challenge when the hot potato sort of becomes EVs and potentially if subsidies get taken away, that will impact demand. And so the U.S. story is way different than the China story. And I think that politicians and policymakers need to be careful because there is obviously a world market for this. It's kind of where the auto industry is heading. You can't stop it. You can certainly slow it down. But if the U.S. ends up ceding a trillion-dollar annual industry to other shores, particularly China, that's something that I think policymakers are going to look back on and really regret.

13:23And how much of a threat do you see China being? I think they're a huge threat. They are producing fantastic cars in maybe the most highly automated factories in the world, certainly the most highly automated factories I've seen. And they really are software first. It's amazing. You can control your house and your driving experience from your phone. They're that integrated. And the Chinese government has made this a priority. They made this a priority publicly in 2015. And now sitting 10 years later, they've achieved their goals and they've subsidized both labor and plants and machinery and the industry along the way to the tune of hundreds of billions of dollars.

14:06And that's put them in the low cost producer position. If you go to places as far flowing as Mexico City or Tel Aviv, you see BYDs all over. BYDs are everywhere. Everywhere. And they're good cars and they're super well priced because they combine that know-how with the government subsidies and they've got a really attractive price on that product. I just looked up some figures as you were talking and I think it's one in two new cars as an EV in China. So 50 % market share and rising. And rising fast, exactly. Versus the US where it's less than one in 10. So it's about 8%. percent. When you step back and look at kind of what is happening and that kind of hockey stick growth of EVs in your, you know, if we gave you kind of truth serum 10 years ago when you were at Tesla and that was the vision, would you expect it to be where we are today, where this is actually a thing?

15:01Yeah. They, uh, like they, they helped us establish a factory, uh, just outside of Shanghai and they built that factory in eight months. And we're talking 1.5 million square feet. Wow. Dead serious about getting the infrastructure in place. But I would say it all boils down to, believe it or not, one policy decision that was made in that. First in cities, and then now it's become more national. But Shanghai and Beijing were the first cities to do this. The policy change that they made was, if you want to get a license plate, If you want to register your car, you buy an EV, license plate guaranteed.

15:41Buy a gas car, get in the lottery, and you have a 1 in 100 chance of giving a license plate. Brilliant. That moves the market. Exactly. Wow. The market moves immediately. And it started with the inner ring of like Beijing or Shanghai where they said, here's the policy, and they kept moving it to the outer rings, and now it's become a de facto national policy. so when you see new car sales it's way more than one in two uh that's going ev but when you see registration now i think 50 percent of the footprint uh or the park uh the park car footprint is ev in china as a result of that it's really striking here in the uk how good the infrastructure is becoming i mean i remember interviewing the boss of ford a few years ago and that was the big problem it was like you know are we going to have the charging infrastructure to allow people to drive around the country.

16:27And I was walking in North Yorkshire a few weekends ago, which is, you know, it's kind of rural. Would you say, Danny? Exactly. I would say. And the little villages around there all had charging points. For sure. You know, policy here is we're going to have to have, we'll get rid of the combustion engine in the next few years. And it has moved the dial. Yeah, I was like in a little village in Vermont this weekend, like tiny village, maybe 260 people. And there are two DC fast chargers in there. And then we were traveling up the length in Vermont. And so then we went to the next village and there was a DC fast charger and it just kept going.

17:01And like you're saying, like, I think the infrastructure has now caught up. It was behind for sure, but it's now caught up. And that's certainly going to encourage more adoption as people see that.

17:20I really want to talk about self-driving cars. Yeah. Because you were on the board, vice chairman of Cruise. Yep. I rode into Cruise like two years ago, maybe. And I had a good experience. And if you look at what's happening today in San Francisco or around the country, really, Waymo is now doing 250 ,000 rides a week. Yeah. Paid, self-driving taxi rides. But could you just talk about the Cruise kind of rise and fall? because Cruz was one of the early pioneers. GM comes in and buys them and kind of lets them operate as their own kind of entity within the mothership. They launched the service in San Francisco and then it kind of all goes wrong, culminating in this very bad accident where a woman was hit by one car, knocked into the path of a Cruz.

18:12Then that car was like, okay, I need to pull over and ended up dragging her along the street. And soon thereafter, things were suspended. And then Cruz effectively, as far as I can tell, has been just disintegrated. It doesn't really exist anymore within GM. What went wrong? So now you've outlined like what happened the day before my first board meeting at Cruz. I joined the board the day after that. I know it's terrible. And it was really tragic. And luckily, that woman is OK now. But what that did was that started to open the aperture a bit to say, OK, we've been pursuing RoboTaxi. And several of us joined the board around this time.

18:50And we just started to ask questions about the RoboTaxi business. How big could it be? How profitable could it be? And, you know, there are people that tout these crazy numbers like this is a trillion dollar industry. I'm never going to have to drive a car again. And, you know, my Tesla is going to be doing taxi rides while I'm at work. And then it's going to come pick me up. And when you start to pull the string on that, you say, okay, explain this to me. How is it a trillion dollar industry? Let's take Uber just reported earnings last week. They reported 10 billion in revenue, 4 billion of that in ride share, the rest in eats and delivery and other stuff.

19:23So$4 billion in a quarter. Take that times four. That's$16 billion. That is for the 80 % market share owner in North America. So that's a$16 billion revenue company or market, if you will. And say they're 80%, so you can goose that 20%. Now you're at 20 billion. It's$20 billion in China. Now we're at$40 billion. It's$20 billion in Europe. Now we're at$60. And the rest of the world probably rounds to another$10. So now we're at$70 billion in revenue. It's not a trillion dollar market. And it is a low to no margin business. And the real trick is when you start to add in the dollars for who's going to buy the fleet, which was the next question.

20:01Hey, are we buying the fleet? Is GM buying the fleet? Right. Does GM become a taxi company? And when you look at that, that's a 30 to$50 billion downstroke. And if you're earning single digit EBITDA against that, it is decades before you ever get paid back. And so we started to ask, is RoboTaxi the business we want to be in? And then you come to, well, what else is the alternative? And the alternative is obviously, hey, who wouldn't want a car that is their chauffeur? Like who wouldn't want a car that you sit in the back and drive? Like Ben Horowitz famously rides around in Silicon Valley in the Bay Area in the back of a Sprinter van where she's got an office and he has a driver.

20:45And he can work while he sits in traffic. That can become available to all of us. Yeah. And so we started to say, should we take the cruise capability, absorb it into GM so we can put this in every one of our cars? Is that a better business than RoboTaxi? And at the end of the day, that's how we made the decision. It is a better business than RoboTaxi. It's a higher margin business. It's a much bigger TAM. And we think that this is going to be a moment in time like we experienced 15 years ago when the world almost immediately transitioned from flip phones to iPhones. Going from driving my car to not driving my car?

21:25Exactly. I feel a bit left out of this conversation because obviously here in the UK, we just don't have them. I mean, you talk about going in them all the time, Danny. I mean, we have got a driverless car company called Wave, which has just got some investment from SoftBank. It is fantastic. I've been in one of their cars, but you need to be in there with a safety driver. Right. What's your thought on when we might see a rollout here? Well, I think the pendulum swings back and forth as you're working in autonomy. It swings back and forth between hardware limitations. Are the chips fast enough?

21:56And software limitations. Has the software seen enough? and can it make decisions fast enough? And the pendulum swings back and forth. And right now we're kind of at a, I would say a hardware constrained chapter of that problem. So take Waymo, which is Danny pointed out, is given 250 ,000 rides a week. But it's only in certain weather conditions and only in certain speeds. They can't go up to 65 miles an hour because the chips aren't fast enough to make decisions at 65 miles an hour. And so we'll get unconstrained on the hardware, side with faster chips and then it will swing back to can the software make decisions in snow in ice in high speeds at airports etc but it's just it's it's it's kind of inevitable it's fun watching this come down the uh the curve because it just keeps getting better and better and better and every time i'm in in a waymo in the bay area i just feel like it's it's it's gotten markedly better each time i ride and that can be just week to week so i think we're going to see this kind capability and personally autonomous vehicles over the next few years i won't put a number on it because my old boss was famous uh for missing deadlines but it'll be in the it will definitely be within the decade i would think so both katie and i have uh young kids my kids are six and eight okay and rite of passage at least in the states you turn 16 you go out and get your driver's license yeah 17 here but yeah of course yeah exactly and it was like i got i had i had my appointment at the dmv the day after my 16th birthday i passed my test i'm like see ya in my old 76 78 station wagon i grew up in london public transport was a bit better but yeah yeah do you think so it sounds like what you're saying is like maybe that's that is just not going to be a thing by the time that they get to be you know of driving age i think for your six-year-old it's definitely going to be a thing and probably for your eight-year-old too like they're going to they say oh it'd be quaint can we go to a park dad where i can drive a car in the driver's seat but that sounds like kind of a quaint thing a stick shift or whatever totally i think i think driving is going to be like a you know we have like horse parks around the bay area as you know and i where you go to drive ride a horse uh and i think we're going to have parks where you go to personally drive a car uh for the most part kids will say like why would i do this i can sit in the back I can either be entertained or I can get work done or whatever, have a conversation, sleep.

24:29It's going to be a completely new experience in a car. But you still think there'll be car ownership? I do because at rush hour, I want to commute and so does the writer. And so we have a demand spike at rush hour that disappears during the day. And so I'm going to want my car to take me to work while you want a Waymo to take you to work. And that's why the personal autonomous vehicle makes sense. I'm probably not going to want to rent my car out and let people spill coffee in the back and have to clean that thing and maintain it. And barf in it, for sure. Yeah, I was going to say that, but then I pulled back.

25:08But yeah, exactly. I don't want that stuff happening in my car. So I do think there's still going to be plenty of demand for car ownership. And remember the car park around the world, around the Western world. So whether you're looking at the UK or Europe or the US is 12 years old. And that is 17 million new cars a year that are feeding that. We got a long time before every car on the road is autonomous and electric, but I think we're going to see that transition happen for sure. So is this kind of the core of Mary Barra's, you know, CEO's vision for GM? It was like, we are going to like, you know, with power windows or whatever, really nice speaker system, our cars in the future, at some point, they're all just going to have this self-driving capability, hardware, software, et cetera.

26:02It's just going to become a standard thing. This is the future. Yeah, Mary's got this really focused vision for GM, which is zero accidents, zero congestion, zero emissions. And so that all leads to this AV future. And that's definitely the path that we're on. I mean, they have the best eyes-on, hands-off system, Super Cruise. when you look at any reviewer in the rankings, that will then become next hands-off, eyes-off, which will then become, after that, no need to be in the front seat. And so they're on this journey, for sure. And the cruise team and that talent has really helped to accelerate that journey.

26:43Right. I think it's probably coming at this from a completely different perspective, because in London, I'm seeing people I know stopping buying cars, because public transport is very good, kids can walk to school, If you want to go further afield, you can rent one quite easily. I wonder when you're thinking about public transport of the future, what's your view on electric air taxis? Like when they might take off. So to speak. Sorry. You could resist that. When you start looking at different things flying around the sky, like Elon and I used to talk about this. almost everybody on the management team at Tesla had some, either had a pilot's license or had pilot training.

27:25And once you learn to fly, you learn how quickly things can fall out of the sky and how dangerous it is. And so after experiencing just that danger as a private pilot, all of us said, we don't want to touch this. Like, it is going to be so dangerous until this gets figured out. But the reason that airlines and air traffic work in the Western world so well is we keep those planes two miles apart from each other. They have plenty of time to react if something goes wrong. And you start filling the skies with EV dolls. That is a very different ball of wax. And so I think I would love to see it. But I think it's going to be a really choppy ride getting there because we're going to go through some accidents that are going to be really hard to swallow along the way.

28:17Enough transport puns. Exactly. We moved from choppy ride into AI infrastructure, which we talked a little bit about infrastructure earlier. But tell us about your new venture, VistaShares. So this came out of an interview I did actually on CNBC, and we were talking about super cycles. And most investors are lucky enough to live through one super cycle, like the industrial revolution, the internet revolution. And we've got like, depending on how you count, four to five going on right now. We have an electrification super cycle going on. We clearly have an AI super cycle going on. We have a biotech super cycle.

28:52We have a robotic super cycle that combines several of these. So one of the hosts on CBC said, well, how do I invest? And I said, I'm sure they're ETS focused on this. And then we got off camera and I said, before you do that, let me just check on those. So I got in the car, left the studio, got in the car, started to look at an EV ETF. And having lived with the build sheet of what goes into an EV for years, it's kind of like tattooed on my brain. And so I opened up the first EV ETF and it says the top five holdings are Apple, Google, Microsoft, NVIDIA, and AMD. And I looked at it and said, there's none of that in an EV.

29:30There's no revenue from any of that stuff. Like you can't put a$25 ,000 NVIDIA chip in a car. You just can't. And then I looked at AI infrastructure ETFs and same thing. Like it was all NVIDIA for the most part because they're a market cap weight. So all the big names. All the big names. And I said to myself, like, what if you could build an ETF that was based off of the supply chain? The build sheet. Like what really goes into an AI data center? What really goes into an EV? So further down the food chain. Exactly. And one of the principles I knew was that the end manufacturer saying cars makes a 13 % margin.

30:06It's a volume business. But the supply chain, we always joked around all the supply chain CEOs and company owners flew around in jets. We did not. Because they had 40 % to 50 % gross profit pools. So there's been this adage when you make stuff, the profit pools in the supply chain. So I thought about how could we create an ETF that would give people access to those profit pools? And when you look at the build sheet of an AI data center and what it costs to build an AI data center, only 5 % of it is the NVIDIA chips. The rest of it is the cooling, the power electronics, the network infrastructure, the fiber.

30:42There are four times the amount of fiber in an AI data center than a regular data center. And so we were dumb enough to put together an ETF that was not market cap based, but was based on the build sheet of AI data centers, a robot, an EV, etc. And it turns out that that is a superior investing approach. When we back tested it, six months, a year, three years, five years, it crushed the ETFs that were in the different sectors. And so then we launched VistaShares to be this kind of contrarian approach where we would democratize access. Individual investors could get the kind of really sophisticated access to the supply chain that here to date only hedge funds could.

31:25Because they had the researchers that could go build up the build sheet and access those stocks. And we've taken that and made that very publicly available. And so now we've got the number one performing AI infrastructure ETF, and we've got other ETFs that are performing super well. But that's the story of VizpaShares. It literally started in an interview, and we pulled the string and started to say, could we do this differently? And that was last December? That was, yeah. Last December, we launched these ETFs. We launched the first one end of December, launched the second one in April, and we just launched a third one last week.

32:00I know hindsight's a wonderful thing, but don't you wish you'd started this sooner? I do. It feels like we're kind of peak. But we started it as soon as we had the idea. And so if AI is a nine inning game, we're at like one out in the top of the first. I remember when I started fooling around with the internet in the mid 90s for the first time, I wouldn't have guessed that that was going to be a 35 year super cycle, but it is a 35 year super cycle. And we're just getting started on AI. And the funds, there's an AI infrastructure one, there's an EV one, and you're launching a third? What's that one?

32:36It's a resilience fund called QUSA Quality USA. And it's a resilience fund that's looking at now the new investment that's pouring into defense in things like drones and others. Oh, I see. Right. And the same idea. Same idea. Same idea. Because I was looking at what the top four, you know, kind of hyperscalers, meta, Apple, Amazon, Microsoft are spending on data centers. And between the start of 2024 and the end of this year, just those four companies are spending$600 billion. Exactly. On data centers. Exactly. Which is just like, it's like canals, railroads, data centers is the new thing. Totally.

33:19Yeah, you're exactly right. Like when I think about I'm sitting in Boston today and they spent$14 billion on the big dig to build a tunnel underneath Boston. 14 billion. That sounds like a really trite, doesn't it? Compared to 600 billion. So to your point, this is the new infrastructure. And that's just four companies. And you can imagine as the rest of the corporate base starts to now transition into these tools, Jensen Long's right. When you ask the models for an answer, that's called inference. And the inference demand is going to go through the roof as we transition into this. So data center adequacy and electrification in terms of the grid adequacy is going to be a huge issue.

34:02What's the big dig, guys? No, it's a tunnel. They took an elevated highway. They cut right through the heart of Boston, and they buried it in a tunnel. And it was supposed to take five years. It took close to 15. It was supposed to cost a couple of billion dollars. It took close to 14. It became this like totemic thing that just kept getting more and more expensive and taking longer and longer and longer in one of the biggest cities in America. It was just like digging this giant trench underneath the city. Thank you. It did become the face of every failed public works project. On the NVIDIA point, what's your sense of where the company goes?

34:40Do you think someone else can challenge NVIDIA for chip supremacy and AI? I mean, a lot of people are trying at the moment. A lot of people are trying. And they've got a definite lock on a critical combination. GPU hardware chips and the software that developers use to code to those chips. and those two together are critical. I think Jensen had this insight that the rest of his chip competitors didn't have, and that was that developer software really mattered. And so when it started to develop their platform, everybody was kind of looking sideways at it, and now I think his competitors see, oh, my God, that was a stroke of genius, and it keeps their position locked on the GPU side.

35:25But to your point on challenging, now as we get to inference chips, So the GPUs are great at training and they're also great at inference, but you don't need all the GPU, the power of a GPU to do the inference side of the equation. And you can use other style chips to do that. And I think you're going to see some challengers. Like I, there are a few CEOs I wouldn't bet against. I wouldn't bet against Elon. I wouldn't bet against Jensen at NVIDIA. And I wouldn't bet against Lisa Su at AMD. And she is laser focused on that. And she's brilliant and an incredible CEO. And so I do think you're going to see AMD given NVIDIA run for its money on the imprint side.

36:03And this is a little bit off the beaten path, but it's related to what we're talking about. I don't know if you saw last week, Eddie Q at Apple. He's went, you know, long time. He's been there for 35 years or something. One of the top execs. And he said something really interesting. You said that search on Safari, their browser fell for the first time ever because people are starting to use chat GPT and perplexity and these bots to actually look for answers. So I don't want links anymore. Exactly. That's the yellow pages. I just want just tell me the answer. Yeah. And I don't know if you have any thoughts on that, because obviously it plays into the kind of infrastructure point and the need for, as you say, for inference for the question answering.

36:50But I just thought that was super fascinating. I don't know if that's like a canary in the coal mine for what's coming. I definitely think it is. Like if you think about like who wants the blue link journey where you're looking for a specific answer and you got to go click on multiple blue links and read content. Magazines who depend on advertising. For sure. Exactly. and so like i think like the times and other and other publishers are gonna have to figure out okay how do i show up in the in the gpt world because that's where i go for search now because it's it's instant it's instantaneous answer without the blue links and the reading and the research but yet like i i do find like when i get the answer i want to go deeper and then i'd like to go back to the publisher and i'd like to read the best publisher thank you Thank you.

37:37Appreciate you. But again, the time's going to figure out, OK, like how do we go work with Perplexity or OpenAI or Claude and Anthropic? How do we go work with these places so that we show up for that second click? and I think publishing becomes really a second click business and it's going to be important to be the publisher that's there that has a reputation for veracity and depth and rigor and provides the best answer behind the answers I should say we have at the Times has actually done a deal with OpenAI One of the things I like about the new ChatGPT model and Plexi model is when you ask it a question it says I'm checking these sources and you start to see logos of sources Yes.

38:21Yeah. Yeah, exactly. And I think it's quite reassuring in some way. It is. And it's a step in that direction. But what I really want is at the end of the answer to say, OK, like if you want the best source that we looked at to get this answer, click on this. Go to Times. Read it in the Times or read it in the Atlantic or whatever. Thank you, John. That's such an interesting conversation. Thanks so much for taking the time today. Absolutely. I feel like we could keep talking for ages, but I'm sure you have more important things to do with your life. I got to go figure out how to keep up with all this stuff.

38:52Chat TBT is really good, I hear. Thank you so much. We really appreciate it. You bet. Good to talk to you both.

39:18you