Nvidia, Amazon & Microsoft Eye $60B Investment in OpenAI, Tesla vs Waymo Robotaxi Race | Jan 29, 2026

0:13Welcome everyone to the information's TITB. My name is Akash Pasfritz. It is Thursday, January 29th. We've got a big scoop for you about how much NVIDIA, Microsoft, and Amazon are considering investing in OpenAI's latest funding round. We will have on our OpenAI reporter for you in just a minute. We're then getting straight to earnings season. First up, we will talk Meta and Microsoft with Tomasz Tunguz from Theory Ventures. Meta investors seemed bullish on the results. Microsoft investors didn't feel so confident. We'll also break down ServiceNow's quarter. We've then got Ross Gerber from Gerber Kawasaki coming on the show to talk about Tesla's quarter and the decision to do away with its two high-end models.

0:54And finally, we are speaking with the co-founder of Armist, the big cybersecurity startup that ServiceNow said last month it is acquiring. It's going to be a fun show, so let's get right on into it. The Information published exclusive reporting that NVIDIA, Amazon, and Microsoft are in talks to invest up to$60 billion in OpenAI. I want to bring on our OpenAI reporter, Shree Mupiti, to help us break it all down. Shree, welcome back to the show. It's great to have you here. It's great to have you here. So OpenAI wants to raise$100 billion. $60 billion of it, in or around there, you say, could come from these three strategic investors.

1:32Help us understand what the breakdown is and then where the rest of that$100 billion could come from. Totally. So the strategic investors are NVIDIA, which can invest anywhere up to$30 billion. So that can be somewhere between 20 to 30 billion is where we expect it to land. The next investor is Amazon, which would be a new investor in OpenAI. We had previously reported last fall that Amazon is in talks to invest more than 10 billion. What we've learned is that it could potentially even go up to 20 billion or even more than that, given other components of a deal that might be negotiated alongside that.

2:06And then lastly, existing investor Microsoft is also in talks to invest, and that could be somewhere in the several billions of dollars. Less than$10 billion is what we know, but that would add to already Microsoft's large stake in OpenAI, given that they've been a partner to OpenAI from the last couple of years. And then the other part of this, which we've previously known, is that SoftBank is looking to add another$30 billion, which kind of gets OpenAI to its$100 billion total almost, right? Exactly. So SoftBank is also in talks to invest up to$30 billion. And what we will say though, or what I've been hearing so far is that these companies, these strategics will be, are in talks to invest up to, but we don't know actually where it will land.

2:49And one of the pieces that I had heard also is that it is actually unlikely that all of these strategics will come in for that full amount. And so it's not clear if OpenAI will actually hit its full hundred billion. If, for example, the strategics and SoftBank aren't able to make that sort of like threshold of where they need to be, OpenAI will then turn to other institutional investors, such as like Thrive Capital or other financial investors that could then potentially fill out the round. Now, when we think about what the amounts might end up being, certainly in the case of Amazon and Microsoft, we know that OpenAI has these cloud deals that it sort of, I don't want to say quid pro quo, but it's like, you scratch my back, I scratch yours.

3:34type of thing in some cases. Do we have any sense for if this depends on how much OpenAI will end up spending on those cloud deals? Yeah. So Amazon, for example ,'s commitment could depend on other deals that both OpenAI and Amazon are negotiating. So these deals are separate and distinct, but because they're all sort of tied and interlinked, they're being negotiated around the same time as well. And so the other deal that Amazon and OpenAI might be landing as well is related to OpenAI's cloud server rental deal. OpenAI had already, for example, signed on a deal worth$38 billion with Amazon over the next seven years.

4:15And so that could potentially expand. And then the other deal is actually a commercial deal for OpenAI to sell, for example, its products like ChatGPT for work to Amazon. And so if, for example, all those deals land in the best way, then maybe Amazon's investment can grow much larger than what the 10 might be, 10 billion might be. And then the other point I wanted to add is related to NVIDIA is that NVIDIA, we had previously reported that is in talks to invest 100 billion dollars in tranches last fall. And so that deal is actually separate from potentially this NVIDIA deal. But of course, everything's interlinked.

4:54So maybe things might change and deal terms might change depending on sort of how things are struck. Now, you talked a little bit in the story about how much funding OpenAI needs to fund all of its ambitions. And you've also reported extensively on the cash flow or cash burn projections rather for OpenAI. This$100 billion round, I mean, this is a big round. I'm sort of wondering how much more funding they would need on top of that. I mean, do you anticipate there's even more funding rounds coming down the line? Honestly, these companies' projections change so quickly. We talked about this when we were reporting on Anthropics financials.

5:33And so perhaps this is the last federal before the company raises, but things, again, change so quickly. And so with large competitors like Google and Microsoft sort of having large balance sheets, perhaps OpenAI will continue raising, or maybe they'll do so in the public markets. And so it's not clear to me when it will be the last raise. But all we do know is that last summer had projected, I believe,$430 billion on compute spend over the next four years, and then also burning about$70 billion until 2030. And so I think that it just shows that they definitely still need a lot of money. Great.

6:15Well, that means that there is more reporting for you to have fun with. Sri, I want to thank you for coming on. That is Sri Mupiti, our OpenAI and Anthropic reporter here at The Information. Okay. Meta reported quarterly results last night. It is expecting revenue in the current quarter to accelerate to roughly 30 % growth, up from the 24 % growth rate that it posted in the fourth quarter. CapEx continues to soar. Microsoft revenue grew just about the same rate as it did last quarter. Joining me now to break it all down is Tomas Tunguz, general partner at Theory Ventures. Tomas, welcome back to the show.

6:51It's great to have you here. Great to be here. Thanks, Akash. Let's start with Meta. What was the headline for Meta for you? Well, stock is up meaningfully today. You can see that the average revenue per person continues to increase somewhere between 10 % to 16 % quarter over quarter. I think that's the exciting part. you see them investing pretty significantly within the world of AI. And there are some rumors that the underlying AI models that they're building around next generation LOM are starting to perform pretty well. There is pretty significant capex here and a fair amount of debt. I think overall, one of the reasons that the stock jumped is that the CFO on earnings yesterday said that they expect earnings to actually be higher year over year.

7:36And that's because the overall revenue growth of the business is large enough to sustain significant capex spending if you also include the debt. One of the things that was interesting to me about last night's call was just how much of a focus they put on the AI business uplifting the recommendations and the ads, the core ads business, which obviously is performing very well. And I think in previous calls where Mark Zuckerberg has really emphasized the personal super intelligence angle to Meta's ambitions, and that was still there, but it felt like it was a little bit of reassurance and saying it doesn't really actually matter.

8:13I mean, we're still going for it, but like the ads business is going to benefit one way or another. Absolutely. Some of the research have published, some of the researchers of Meta have published research on the impacts of reinforcement learning on overall ad performance. And it's quite significant. I mean, you can see it, right? Like quarter over quarter, ARPU, 10%, 15%, 18%, 16%. Those are pretty astounding numbers for a business ad scale. And so I think it tells you that there's a lot of juice left to squeeze from the lemon on the ads business. One of the questions I wanted to ask you was, Zuckerberg last night talked about, you know, he really wants to have control over building the best in class models.

8:54And that's obviously why he's gone so hard at hiring talent and why they're hyping up this new family of Lama models whenever they do come out. But Lama is an open source family of models. And so I wonder if you could just help us sort of square the fact that it's open source. And yet he's saying, I mean, we want to really own this stack of the most competitive models. I mean, are these contradictory? Is it different? Help us walk us through it. Right. Well, so there's open source and closed source models. Both of them exist. And then you have the data center business, right? So you can host a closed source model on meta infrastructure.

9:32or you could host an open source meta model also on meta infrastructure. Both of those businesses are working really well today. The very largest inferences businesses, Anthropic and OpenAI, are predominantly closed source. But you have many NeoClouds that are offering open source models. I don't know which direction meta will go. The first four LLAMA models were open source. but there's no, at least I haven't seen any direction to say, to indicate whether or not Meta's new models will be open or closed. But from what you're hearing, I mean, they're coming out pretty good so far from what we're hearing.

10:07Yeah. I mean, I think it's a testament to the team. Also, many of the techniques are for the current generation are well known. You can look at the Chinese open source models really hitting close to state-of-the-art benchmarks almost every week. And so as a result, I think Meta should be able to catch up relatively quickly. And then the question is, can they put themselves in a position to set the pace? That I think can take more time. Okay. So what about Microsoft? I mean, Azure growth was 39%. That was about the same as it was last quarter. The stock is down. What do you think investors are reacting to?

10:43Well, yeah, the stock is down a lot, 10%. I think you're seeing 72 % of software stocks this morning or trading down. And Microsoft is top three in terms of overall decrease. There's a difference in the way that Meta and Microsoft were perceived.

11:03Meta's core business is accelerating, is growing faster. Azure is still relatively constant. One of the limiting factors there is just data center capacity. And Amy Hood mentioned it on the call. They're sold out until 2026. They can continue to grow at this current rate. I think probably the major risk here is actually the open AI concentration, right? 45 % of the backlog, the remaining performance obligations are tied into open AI. And there were one or two questions in that session yesterday talking about can open AI actually generate the kinds of revenue needed to satisfy these RPOs? That's a big question that's on the stock.

11:43He pauses. Tomáš pauses. Nobody knows the answer. I mean, we ran an analysis yesterday of a chart of just like how fast OpenAI and Anthropic are growing in terms of billions of revenue. And you look at other businesses generating tens of billions of revenue, it's taken them 30 to 50 years to get there. So whether that's Salesforce or Adobe or SAP, and these two businesses are doing it 15 years or less. So there's clearly a lot of momentum. And the numbers that we're talking about in terms of well over a trillion of OpenAI data center commits. I mean, we've never seen those kinds of numbers over five years of intended planning.

12:21And so I think that injects a little bit of uncertainty into the stock. And then you have a broader drawdown in the software market combination of the two led to the result today. What about the Office 365 co-pilot numbers that we got yesterday? So we got a little bit of color. I think it was 15 million paying users, I think, if I'm not mistaken. Whatever it is, I mean, it's, you know, some proportion of the current Office 365 users have subscribed to Copilot. I mean, it's growing quickly. My question for you is if you think that that growth can sustain. Well, it's a cross-sold product. I don't think it's a product that people go out and buy.

13:02It's more of a product that is sold within an existing software suite like Power BI. is. I do think Copilot as a brand, and I'll include GitHub in this category as well, had the early lead and is now starting to be outmoded, right? Look at what Claude is doing with Cowork or with Chrome or Google releasing the Chrome automated browser features, the integration of AI within email. You're seeing a lot more innovation within the world of GitHub with the rise of Cursor, there hasn't been as much innovation or the momentum hasn't continued out of Microsoft the way that it has everywhere else. And so I think it's reasonable to have concerns about an increasing churn rate or at least a decreasing co-sell rate within the co-pilots broadly.

13:51Before you go, I do want to ask you about ServiceNow, which is, look, we had you on this time last quarter and you and I were talking about how ServiceNow had posted very strong results And still, I think the stock had fallen the day after earnings one quarter ago. And it seems to be kind of the same story today. I mean, ServiceNow stock is down for the year, over the past year considerably. There is that broader drawdown in software stocks that you talked about. But, I mean, if you look at the multiple for ServiceNow, a year ago, it was trading sort of in the top five, top ten SaaS multiples in terms of forward revenue.

14:32Now it's like much closer to the median. And the results have still stayed somewhat the same. What do you think? It's pretty amazing. I mean, eight consecutive quarters of beating expectations. McDermott yesterday talked about it being a rule of 55 company, right? It's 21 % of revenue growth, 35 % free cash flow margin. Like it's really good. The fundamentals of the business are there. You have three major acquisitions. There's some acquisition integration risk with Moveworks, Armis, and Beza. And so clearly there's some execution risk there, but they're being aggressive in terms of M &A to position themselves within the world of AI.

15:11There might be some pressure from some newer startups within the world of ITSM who are growing very quickly, but none of them are less than 10 million in ARR. And so I think it's very difficult to look at ServiceNow and say, there's a meaningful startup competitor that's coming after them. You do have Atlassian that's pushing up against them from the mid-market Jira Service Desk with 55 ,000 paid customer accounts. And so maybe there's a potential competitive threat there, but I don't think there's anything acute. and so it's it's it's a bit of a head scratch right you're right the p-e ratio has fallen pretty significantly and so i don't know why the stock is being punished even though you look on almost every single metric the company is doing very well we have the co-founder of armis actually coming on the show in just a couple minutes towards the end of the show i wonder what questions you might have for him maybe not necessarily about service now because i plan to ask him about that, but just about the business that he's in and how you see it going.

16:13I mean, I think, so within the world of security, within the world of ITSM, within the world of human resources, the basic workflow is ticket management. And I, as an employee, might ask, what is my 401k policy? I might ask, I can't get into my computer. Will you help me reset my password? And then the last one might be, I think I have a virus on my computer. And all of these workflows are tickets. People have problems that are then processed. They're starting to be processed via AI. And so conceptually, there's no difference in the technology. We're selling to different buyers. I would ask him, how quickly are all those workflows converging into a single system?

16:52And I have to believe he will probably say very quickly, and that's one of the reasons why ServiceNow bought Armith. Great. Well, Tomas, I want to thank you for coming on. It's always a pleasure that is tomas tunguz a general partner at theory ventures here on ti tv okay tesla reported fourth quarter earnings revenue fell three percent year over year net income also fell markedly the company announced a two billion dollar investment in xai and also said it would stop production of its model s and model x high-end cars over time joining me now is ross gerber president and ceo of gerber kawasaki ross welcome back to the show it's great to have you here great to be back great to be back so another exciting earnings call from elon musk and company let's talk about the model s and the model x what did you make the decision to stop man i was up at 2 a.m crying i love this car it was like the car that i found tesla you know like in 2013 when they released the car, I think it was 2013 or 14, and we went down and saw it, I was like, this car is going to change the world.

18:03And we invested in Tesla, you know, and it's been one of my best investments in my career over the last, you know, 12, 13 years that we've owned it. And, and so to, and the production of the Model S is just like a crushing blow to what I tweeted yesterday, the end of an era in Tesla. Like Tesla is not an EV company that's going to make self-driving EVs. It's now a robotics company, all in on essentially robot cars and robot humans. And I think for me, the challenge is, first of all, I invest in this as a climate play and I see the business of EVs as a great business. And then secondly, if it's going to be a robotics play, it seems like it will be years before any revenue or monetization of these investments.

18:53I mean, it could be five years before they even see any real revenue from this stuff. So they seem to not want to make the transition smoothly. And so I'm just concerned. Long-term, I get Elon's vision. I don't know if I buy it, but I get it. But it's not the company that I used to own. It's sad. Okay. But so, Ross, this is a, this, I mean, we had you on a couple months ago and I asked you the same question, which is that you're still a Tesla shareholder. You're grappling with this transition. And what you had told me at that point was, look, it's no longer an automotive play. It's an Elon play.

19:30And when you lose faith in Elon, that's when you rotate out. Where are you in your sort of reflections on this? Oh gosh, you know, it's been, as I said, a long night. Up at 2 a.m. Yeah, you know, I fell back asleep. Don't get me wrong. I'm not going to like die on this issue. But yeah, yeah. No, but I really did wake up like, oh my God, I should buy this line from Tesla. Like I'll make the Model S and Model X and I'll rebrand it because the only reason they're not selling cars is because of Elon. So you're right. Tesla has become a bet on Elon because there's no valuation fundamentals that make this company like a$1.5 trillion company.

20:12There's no numbers in my mind that make this stock worth more than$200. So if you're going to pay over$200 for the stock, you're betting on Elon, 100 % on this vision. And I'm not a short seller or a hater in that sense. I'm not going to tell you that Elon's not going to be successful because that would be stupid. he's been successful at a lot of things that he's tried and he's been unsuccessful at many of the things he's tried like Cybertruck but you know I'm not going to bet against Elon I'm going to hold some of my stock because God willing he this is the future you know I want to be a part of it if he succeeds because there are very few people bold enough to do what Elon's doing and as an investor there's a place for that in your portfolio as long as you put it in the speculative place.

20:56I'm just not willing to pay$400 a share for it. Now, there was some discussion yesterday about, not on the call, I mean, I was seeing people talk about the fact that SpaceX will likely go public later this year. And could you see a number of Tesla shareholders just sort of make that their Elon play? And as a result, you see them rotate out of Tesla into SpaceX. is that a risk for Tesla do you think yeah it is a risk and we're getting a lot of calls about SpaceX for sure and I keep telling people you know this thing has played out so long ago I can't even tell you you know to pay a trillion and a half dollars for a space company you know that does 15 billion in revenue is just insanity like a hundred times revenue for a business that's not really a great like business in the sense of profitability so SpaceX has been wildly successful Well, same thing, if you're going to pay 23X just because it's Elon's company, I wish you the best of luck.

21:55And it's not something I'm going to do. But that said, I think there's a pretty decent amount of Tesla shareholders that want to buy SpaceX stock. And I don't know if they'll sell their Tesla stock to do that per se. I think they'll just add it to their portfolios. And we're seeing that on the Tesla Aquilite side, like Ron Barron and Cathie Wood are all in on both companies. and I would expect the Tesla faithful to do that. Are you going to buy SpaceX when it goes public? No, and we already own SpaceX through some of the barren funds. And so we've done that long ago, and it's not a big play for us.

22:30We own some XAI at my firm, which we bought originally as X, which has now paid off. So I can find it ironic that Tesla is paying twice as much for their shares than I paid for mine. But that made me double my money, so I'm not going to cry about it. So that said, I think XAI is probably a much more interesting investment than SpaceX, to be honest. And I want to ask you quickly about the CyberCab. There was a lot of talk last night on the call about the vision of the model where you can sort of rent out your Teslas as part of the fleet for the CyberCabs down the line. how confident are you in the cyber cabs from a technology perspective and then what do you think about the whole fleet model elon is pitching so these are both two separate like shows on its own right you know like i can talk about robo taxi and the risk and rewards and human behavior and i'm in the camp that humans don't change very easily and we all have our own cars and they're all different every time a yellow car comes by my kids yell yellow car you know everybody wants their own vehicle and trying to change human behavior, not to mention just how long it would take to make millions of these robo taxis.

23:48It will be many years into the future. So I just don't see humans changing. So I think we're going to be driving for the rest of my life. Personally, I'm not giving up driving. If you think I'm riding around in some gold cab everywhere I want to go, that is not the way I see my future versus I wanted a Tesla Roadster that would drive itself. You see what I'm saying? So I think a lot of people aren't giving up their cars. So then this idea of it's going to be like Airbnb is just like a fantasy because where I live, there's Waymos everywhere already. Like you can take an autonomous ride any minute here in Santa Monica.

24:25Like they're everywhere. They've already blanketed the cities. So you are already seeing this. And what we're seeing is it's just eating up Uber's business. It's not convincing people to sell their cars when you can buy a Tesla for$500 a month. And if you figure that out, it's about 50 cents a mile to drive a car, the average amount of human drives a year. And right now, taking a cab like a Tesla Robotex, it costs$5 a mile,$3 to$5 a mile. So until those numbers make sense economically, people aren't going to do this either. So, you know, I'm not that convinced we're all going to be cab riders in the next five years.

25:04I'm just not convinced at all. And in fact, I will not be one. I took my first Waymo this week, actually, in San Francisco. I was in San Francisco. Pretty cool, huh? And I took it twice. And the funny thing, to your point about behavior, the funny thing for me was the Waymo was fine. I mean, you know, I took it and I didn't think about it. and then when i was taking an uber to sfo uh yeah i was sitting in the front seat and the driver at some point he told me he said yeah you know it's on self-driving right now and then i got scared for some reason it's weird he was sitting there and it's like i i was like i want you to be driving if you're sitting in the car but the waymo i was totally fine with it anyway well you just hit tesla's biggest problem is that teslas are not good enough at driving to sit in the back seat and trust it.

25:54Exactly, yeah. And a lot of that is because they refuse to add more sensors and radar to the system, and they'll solve autonomy next month if they just change their hardware. That's what I believe. Their software is way better than Waymo's, but their hardware sucks compared to Waymo's, and that's why we feel confident with Waymo, because the Waymo drives confidently, but full self-driving drives like a 14-year-old. And yeah, you want the monitor, like monitoring, and when I use it, I have to monitor it because I don't trust it. And I'll get in the back of a Waymo and literally go to bed. You see what I'm saying?

26:26And so Waymo's won the race. So the idea is, oh, well, Tesla can just like scale so quickly once they solve this. That is also not true. There's a whole infrastructure around taxi that I've gone around and followed Waymo's and gone back to their main base and seen their main base and everything. And dude, that's a full operation. So this idea, it's just like Airbnb is exactly how Airbnb works, which is it's not really mom and pop people just renting out their houses. What air does it become? Yeah, there's a centralized fleet. It's a business. So people are buying fleets of houses and renting them, and they're like bad hotels, essentially.

27:03So essentially what they're talking about is creating fleets of Model Ys that will be like bad cab services instead of taking one that's centralized. So I think Tesla is better off just managing the cab service themselves because they can control quality. Remember, quality control is why Uber sucks, because you don't know what you're getting with the driver. You don't know how bad it's going to smell, how bad the driver is, how long they've even been in your city. In a lot of cases, never even driven. So I think there's a big disruption that's going to happen, but it's going to happen to Uber.

27:35You know, and I think Waymo is the winner here. And I think Tesla can compete if they can get this to work. But Elon even said it in the conference call when they asked him, why isn't full self-driving working? And he basically said, because where Ross drives, it's super freaking hard. And we haven't mastered it because people crash there all the time. And he's 100 % correct. Where I drive every day, West Los Angeles, he brought up Santa Monica Boulevard in Wilshire, which is right the center of where I'm always driving. And he was talking about these intersections that are like hell. And they are.