Tencent’s OpenClaw Obsession, The Information’s Next GP List, Former Tesla Exec on Elon’s Ideology

27 Mar 2026 · 47 min · 16 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

The episode covers four stories: Microsoft’s hiring freeze, Tencent’s OpenClaw push, The Information’s 2026 “Next General Partners” VC list, and an Editor’s Cut debate on private equity using AI, plus a book interview with a former Tesla executive.

Guests and backgrounds

Aaron Holmes (Microsoft reporter, reports on Azure and North America sales hiring freeze). Juro Osawa (author on Tencent OpenClaw reporting; with Asia Bureau chief Jing Yang). Julia Hornstein (venture capital reporter; compiled 2026 Next General Partners list). Ken Brown (senior finance editor; wrote Editor’s Cut on PE + AI). Martin Peers (co-executive editor; discusses AI optimism vs security/accuracy risks). John McNeil (former Tesla president; later COO of Lyft; CEO/co-founder of DBX Ventures; board member incl. GM and Lululemon; author of The Algorithm).

Key claims and examples

Microsoft told managers to pause new hires in Azure engineering and North America enterprise/midsize sales until June to improve gross margins; Copilot teams still hire. Tencent launched eight OpenClaw-related products in a month, encourages internal “horse racing,” donated after OpenClaw founder Peter Steinberger complained, and had a U.S. meeting to propose more support; risk cited when OpenClaw updates broke Tencent products. VC list criteria: led major deals early (AI like Together AI; defense like autonomous shipbuilder Cerronic and investments in Antwerp Industries); examples include Lisa Hahn (Lightspeed; Character AI and Thinking Machines Lab) and Max Rimpel (General Catalyst; Mercore and General Intuition). Editor’s Cut: PE firms will use OpenAI/Anthropic across thousands of profit-making portfolio companies to boost profitability; debate highlights security risks, AI inaccuracies, and “token pricing” implying demand may be overstated. John McNeil argues Tesla pivoted to autonomous cars and humanoid robots after China trip; Optimus manufacturing and volume targets pulled forward; he says “automate last” (DoorDash example; Tesla Model 3 line automated too early) and stresses architecture before agent automation.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

Microsoft's Hiring Freeze

0:45 to 3:53

Discussion on Microsoft's decision to freeze hiring in key divisions due to financial pressures.

“These are the folks who are most likely to become the next GPs running some of the top VC firms in Silicon Valley.”

Tencent's OpenClaw Strategy

3:53 to 7:46

Insights into Tencent's aggressive development of OpenClaw products and its implications.

“Okay, so between now and June, they've got to get their house in order so that the total year looks just fine and just rosy.”

Next General Partners in VC

7:46 to 13:40

Exploration of the 2026 list of next general partners in Silicon Valley venture capital.

“The bet on OpenClaw started after one of Tencent's new product managers became obsessed with the open source software.”

Criteria for the GP List

14:00 to 20:26

Learn how the list of general partners was compiled and the factors considered.

“I wonder, let's just take a step back from who's on the list.”

Introduction to Private Equity's Interest in AI

20:26 to 21:18

Discover why private equity firms are focusing on AI and its implications.

“That is Julia Hornstein, our venture capital reporter, here at The Information.”

Private Equity's Strategy and AI

21:18 to 23:09

Understand how private equity firms plan to implement AI across their portfolios.

“Okay, so Ken, you started off your column this week by talking about the market dynamics for compute and the market dynamics for AI and how So sometimes those line up, sometimes they don't.”

The Risks and Realities of AI Implementation

23:09 to 24:11

Examine the potential risks and limitations of AI in business contexts.

“I mean, if it doesn't work, it doesn't work.”

Evaluating AI's Effectiveness in the Market

24:11 to 28:00

Discuss the effectiveness of AI and its demand in the current economy.

“who love to come into, are using it to really take over these companies, cause massive havoc.”

AI in Private Equity and SaaS Challenges

28:00 to 29:10

Exploration of AI's role in private equity and current SaaS market challenges.

“I mean, these are companies that it would actually make a lot of sense, I think, to have AI to help them, you know, address these risks.”

Market Dynamics and Software Demand

29:10 to 31:08

Discussion on the evolving software market and private equity's exit strategies.

“And these PE firms have to exit at some point.”
Show all 16 chapters

Tesla's Strategic Pivot

32:11 to 33:56

Insights on Tesla's pivot towards autonomous vehicles and robots from a former executive.

“Our next guest is a former president of Tesla, one of Elon Musk's direct reports.”

Comparing Tesla to Competitors

33:56 to 36:52

Discussion on Tesla's competitive landscape, including comparisons with Waymo.

“And so he pivoted the company two years ago.”

Merging Tesla and SpaceX

36:52 to 38:43

Exploration of the potential merger between Tesla and SpaceX and its implications.

“I think like one of the things I read about in my book, The Algorithm, is one of Elon's organizing principles is simplicity.”

Insights from 'The Algorithm'

38:43 to 40:58

John McNeil shares insights from his book on Tesla's operational framework.

“So let's talk about the book that you are launching or publishing this week.”

The Automation Debate in AI

40:58 to 42:00

Discussion on the importance of understanding processes before automation in the AI era.

“I mean, you know, and it's not even going to be me.”

The Journey to Effective Automation

42:00 to 45:50

Learn why automating processes too early can lead to failure and how to effectively implement AI.

“But what they did instead was they wanted to learn the business so they could optimize the flow and then have something that could scale.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:13Welcome, everyone, to The Information's TI TV. My name is Akash Pasricha. It is Friday, March 27th. We are kicking off the show today with some exclusive reporting. Microsoft executives have told managers in its cloud and sales groups to suspend new hiring. We'll hear more details about that from our Microsoft reporter. We've also got some more exclusive reporting from our Asia reporters. Tencent is making a big bet on OpenClaw as it continues to compete in the China AI race. We'll hear from our Asia Bureau on that. Next up, the information published its 2026 list of the next general partners.

0:49These are the folks who are most likely to become the next GPs running some of the top VC firms in Silicon Valley. We'll bring on our VC reporter to talk about how she made the list. Also, it is Friday, which means we've got another edition of The Editor's Cut. This week, we are talking about how private equity is stepping in to fund the AI boom. And we will wrap with a conversation with a former Tesla president, one of Elon Musk's direct reports, who wrote a book about what he learned in that role. It's going to be a fun show, so let's get right on into it. Microsoft has frozen hiring in some of its most important divisions like Azure Cloud and also North American sales.

1:28It is the latest data point that big tech companies are increasingly paying attention to their bottom lines as big capital expenditures continue to mount. Joining me now is Aaron Holmes, our Microsoft reporter, one of the reporters behind that story. Aaron, welcome back to the show. It's great to have you here. Hello. Oh, okay. Which groups exactly have frozen hiring?

1:50Ken Brown:So we know that this is happening not across the entire company, but it is happening specifically in the engineering groups that work on the Azure cloud business, as well as multiple sales organizations across North America, but that sell to large enterprises as well as midsize and smaller companies. So this is pretty widespread right now. And specifically, you know, So our understanding is that these units have been told to not move forward with any open roles between now and the end of the fiscal year, which is in June. And what is it exactly about these groups and the way that their performance is going that they've decided to freeze hiring there versus other parts of the company where I understand hiring is still continuing to grow?

2:36Ken Brown:Well, what we know is that the executives who have told managers to stop hires in these groups have pointed to the need to improve margins. Specifically, you know, in Azure, we're told that one of the top priorities between now and the end of the year is to get the gross margins in better shape. And essentially, you know, we got this email from an executive out to managers saying that until we have credible plans locked to address the gross margin gap, pressure on headcount will continue to cascade. Basically meaning that, you know, that these organizations are trying to get their finances right between now and the end of the fiscal year.

3:18Ken Brown:And that is also coming at a time when, you know, Microsoft stock is under a lot of pressure and investors are essentially, you know, with their selling of shares, telling the company that they want to see better margins from Microsoft. But yeah, I mean, at the same time, we're seeing other orgs like the engineering groups that are building Copilot, which is, you know, the AI add-on that Microsoft sells in its software, those teams are still hiring, which probably reflects that, you know, Microsoft is really kind of eager to race other AI companies and keep improving those products. And remind us, the Microsoft fiscal year, it ends in June, is it?

3:57In June, yes. Okay, so between now and June, they've got to get their house in order so that the total year looks just fine and just rosy. I want to ask you a little bit about AWS and Google Cloud. They're the other big cloud players here. Have we heard anything similar about hiring freezes there?

4:19Ken Brown:I'm not sure about hiring freezes specifically, but we do know that both of those companies are looking really closely at how they expand headcount. We know that at AWS, there was recently layoffs and my colleague, Kevin McLaughlin, just wrote about how, you know, the company is now telling employees to use AI agents to essentially automate work that was previously done by laid off employees. And I think the same is certainly true at companies like Google and Microsoft, where, you know, even if AI is not replacing cut jobs one-to-one, there is a growing expectation that employees should be using the AI tools, especially that these companies sell themselves to automate more of their work and to essentially do more with less.

5:03Ken Brown:And we're definitely seeing that across the entire software industry right now. Now, what has Satya Nadella and other Microsoft leaders said about managing the broader CapEx commitments with margins? What are they saying? So it's definitely an interesting time because we are seeing Microsoft and these other companies continue to spend exorbitantly on, you know, CapEx for data centers, also on data center leases, which are a bit more expensive, but help them, you know, free up more capacity more quickly. At the same time, I think that all of these companies are sort of reckoning with the fact that their workforces grew a lot in the last five years, especially in the, you know, post-COVID tech boom.

5:48Ken Brown:And I think now, from what we're hearing, a lot of senior executives at Microsoft are questioning, you know, have we reached peak headcount and do we need to keep growing headcount significantly in the years ahead? Or can we essentially, you know, make do with the amount of people we have now and continue to grow the business using other scaling factors like AI tools or, you know, potentially even cut headcount in the years ahead? So I think we've started to see that already play out with Microsoft's headcount kind of plateauing in the last couple of years after that rapid growth. But this is a very real discussion that's happening among senior executives.

6:25And last question for you, Aaron. I mean, among the sources that you talked to, I wonder what the general vibe is among Microsoft employees or people in the Microsoft orbit. I mean, the stock price is certainly one indication of how investors feel. But when employees hear about certain groups, and these are not small groups. I mean, you know, Azure is the flagship business. When they hear about hiring freeze, are they spooked? Are they sort of, do they understand it? What's the vibe?

6:52Ken Brown:Yeah, I mean, so I think a lot of Microsoft employees are a bit anxious and people have kind of become used to headcount reductions on an almost annual basis. We haven't seen any layoffs yet this calendar year, but last year there were about 15 ,000 roles cut, and a lot of people are bracing for maybe something similar to come again. This, you know, current calendar year. I think in general, people are aware that Microsoft is somewhat under pressure from investors. And one of the kind of easy levers that tech companies have to make investors happier is to cut costs. And headcount is usually kind of one of the first and easiest places for them to do that.

7:31Ken Brown:And that is definitely something that's making people a little bit nervous right now. Great. Well, Aaron, I want to thank you for coming on. That is Aaron Holmes, our Microsoft reporter here at The Information. The Information's Asia Bureau published exclusive reporting about Tencent's big bet on OpenClaw as the company competes in the China AI race. The bet on OpenClaw started after one of Tencent's new product managers became obsessed with the open source software. Our Asia Bureau chief Jing Yang spoke with Juro Osawa, one of the authors behind that piece. Here is that conversation.

8:06Juro Osawa:Hi, Juro. So what's exactly been happening inside Tencent since OpenClaw became a global frenzy in January this year?

8:16Jing Yang:So Tencent has been doing a lot with OpenClaw and AI agents in general. So just this month alone, Tencent has launched eight different OpenClaw-related products and services. And that's on top of like five they had before that. And we've also reported exclusively that there are two more AI agent projects in the making. So there's a lot happening and different teams within Tencent are racing to come up with new projects.

8:47Juro Osawa:That sounds pretty messy and chaotic. I mean, is this something that Tencent's top management is encouraging deliberately or is more of like a bottom-up initiatives coming from different teams and units within Tencent?

9:02Jing Yang:So top management is definitely encouraging this, endorsing these projects. And Tencent's actually quite famous for making different teams compete inside the company to work on similar products. And people call this horse racing. So it's possible that the best product could come out of this kind of internal competition. But, you know, people may not need multiple AI agents for the same tasks. So, you know, some of those projects could end up competing against each other over users and, you know, also internal resources. Right. So but this right now, this makes sense because, you know, the AI agent battle is still kind of in the early stage.

9:45Jing Yang:So, you know, people don't know what kinds of agents will be the right ones. You know, what's the right approach. So they're doing, right now, experimenting with a lot of different projects.

9:57Juro Osawa:I see. And with Tencent so aggressive building open-claw-like or open-claw-based agents, how has this effort been received in Silicon Valley and especially in people who are actively contributing and maintaining the open-claw community?

10:19Jing Yang:So there was actually a little bit of drama because OpenClaw's founder, Peter Steinberger, complained on social media that Tencent was benefiting from OpenClaw without giving back to the open source community. But just days after that happened, Tencent made a donation to OpenClaw, and then last week, a senior Tencent executive ended up having a meeting with Peter Steinberger in the U.S., and they are proposing more support for his foundation. So things have kind of moved very quickly. I see.

10:55Juro Osawa:Well, this is certainly setting Tencent onto a course where they are going to be, you know, I guess they are positioning themselves as, you know, a contributor and an advocate for open source technology. But what could potentially be the pitfall or the risks for such a big company like Tencent, whose services are used by over a billion users? For me, it sounds like you are essentially building a lot of applications on technologies that you do not have oversight or control over. It kind of feels like handing the key to your safe, to another person.

11:37Jing Yang:Well, there was a moment, like even just last weekend, that OpenClaw had an update. And then some of those Tencent products based on OpenClaw stopped working. And so Tencent team had to run around and fix the problem over the weekend. And so that could be an example of relying on a different framework, external framework that's open source and that's not owned by Tencent. But at the same time, this is important to Tencent and it's an opportunity because Tencent in China, Tencent has been a little bit kind of behind in the AI race. and its own AI models aren't very competitive. And Alibaba and ByteDance are the big rivals, but they are ahead of Tencent in AI models and also chatbot applications.

12:39Jing Yang:So this open-claw boom in China, it's creating a new opportunity for Tencent to kind of try to get ahead in this race. And if they can offer the best way to use open-claw in China, integrating it with WeChat and offering a bunch of services, maybe that will help Tencent in this overall race. So I think that's the idea. Right.

13:05Juro Osawa:I guess that is the hope and the game plan here. But at the end of the day, the proof is in the pudding. We're going to have to wait and see how many people and companies are actually going to adopt these agents and how effective they actually will be. Thank you, Jiro. Thank you. That was Jingyang and Juro Osawa from the Information's Asia Bureau. This week, the Information launched our 2026 edition of The Next General Partners. Our venture capital reporter, Julia Hornstein, went to her sources to find out who is most likely to ascend to the most powerful posts in Silicon Valley venture capital over the next few years.

13:44It is an exciting project that the Information publishes every year. And so to talk about who she picked and why she picked them, I want to bring on Julia to talk to us about her reporting. Julia, welcome to the show. It's great to have you here. Great to be here. Okay, let's talk about the next GPs. I wonder, let's just take a step back from who's on the list. How did you even go about choosing who's on the list? What were the criteria that you landed on? So we talked to roughly two dozen sources in the industry, and that can range from anywhere from founders to limited partners to investors who are a lot of these people's peers.

14:20And we really tried to index on people who have led major deals on behalf of their firm. Many of the deals that came up were in Together AI, which was one of the more popular companies on our list. But we also looked at people who were really trying to seize the moment in venture. And this year, unlike other years that we published the list, that was really defense tech. So we saw some people on the list who had led major investments into autonomous shipbuilder Cerronic as well as some investments into Antwerp Industries as well. So this is kind of like the power law in play here. It's like the people who got into the biggest deals.

14:57I mean, those are the people who most likely will have the momentum then to get the next big deals. And that's ultimately what would lead to a GP promotion. Yeah, I mean, we're really looking for people who get into these deals early, who can at least initially point to some on paper returns. But ultimately, we really want people who are returning capital to limited partners. And this year, that looked like a lot of M &A. We had some people on the list who had invested in chip company Grock, which was bought by NVIDIA in December, which we reported. But we're also looking for people who are getting in early to these major deals like Sironic.

15:34Okay. And again, we'll get into the list in a second, but I wonder then themes across the company. You talked about a lot of AI companies. Was it only AI companies? Sounds like there was a little bit of defense sprinkled in there. What did you notice? Yeah. I mean, defense was just one of these areas with a ton of overlap. I mean, like hard tech, deep tech, whatever you want to call it, has just become all the rage this past year in Silicon Valley, especially, you know, given the current presidential administration's interest in adding not only AI into the military, but also this ready-made Silicon Valley tech, hard tech into the military.

16:10Eric at Altimeter, Mustafa at NEA, they really stood out as people who had invested in autonomous warfare, as well as hardware companies like Sironic, as I mentioned, Andrel as well, but also Chaos Industries and Castellion, which are two other up-and-coming players in the space. You mentioned Eric. I mean, I was looking at some of the interesting names on the list. And we should say there's people on the list from Andreessen, from Lux, from Lightspeed, NEA, like you said. Tell me a little bit about Lisa Hahn and Max Rimpel. Those were two names that stood out to me. Yeah, I mean, Lisa is a partner at Lightspeed.

16:48She invests in early stage enterprise software companies where she's, you know, she's worked at a firm for a couple of years, I think since 2022. There she sourced and led the firm's investment into chatbot maker, Character AI. I think she invested in 2023. That investment at the time was valued at roughly$1 billion. But then, you know, Character was bought in that kind of complicated purchase by Google a few years later. She also invested in Thinking Machines Lab, their, you know, seed round in 2025, which the information covered extensively. So I think that she's really proved herself as someone who can source these deals super early and have compelling exits despite a tough IPO market.

17:36And what about Max? Yeah, I mean, Max is a partner at General Catalyst. I think some of his buzziest deals are he helped out with General Catalyst's seed investment into AI staffing company Mercore. He's also invested into Aru, which is this predictions research startup that I think was founded by teenagers, which is just so crazy to think about. But he's also invested into a world model company, General Intuition. And since, you know, I think that what we saw with Max is since General Catalyst led Merck Core's seed round in 2023, I mean, this startup's valuation has just soared. I think that it was last reported at$10 billion, including the investment as of October.

18:19So I think he was another person where we were able to see this trajectory, at least on paper, in terms of returns. I wonder, as you talk to all of your sources and maybe even spoke to some of these investors personally, as you were doing this project, did you learn anything about the state of venture capital? and even though I think about the likelihood of promotions, past lists that we've done, there are certainly people who have been promoted, but then there are people who have been poached and they become GPs at other firms. And I mean, I just wonder if you learned a little bit about whether or not there's room at the top for more GPs given the difficult state of venture capital.

19:03And I know LPs are kind of tightening their wallets too. I mean, what did you hear? Well, I mean, big funds still have a lot of pressure to deploy capital. I wrote a story about this a month or so ago, but all of these big funds have been raising. At a really impressive clip, I mean, Andreessen Horowitz earlier this year raised$15 billion. Kleiner Perkins just raised$3.5 billion. Founders Fund, Thrive Capital, and others are reportedly raising billions of dollars in new funds. So although it's a really tough venture market for people who are sold below GPs, people who are working with smaller amounts of capital.

19:40At the top, there's just so much pressure to get this money out. And really, the only way to do that is by growing the share of general partners at the funds. So I think that there's still a lot of opportunity for people, as we have on our list, who are in these really top funds to make a name for themselves and deploy billions of dollars of capital. In addition to that, as people see that these people are putting money to work, there is opportunity for them to get poached by other funds. Someone who is going to be on our list, who we ultimately couldn't include, was promoted from partner to general partner on Monday.

20:18Who was that? I think it was Morgan Hitzig, but don't quote me on that. Well, look, Julia, it's a great list, and I want to thank you for coming on. and a very exciting project. I encourage everyone to check it out. That is Julia Hornstein, our venture capital reporter, here at The Information. It was a busy week of news with the social media trial and OpenAI's Sora pivot making headlines. But this week on the Editor's Cut, I want to talk about a slightly different topic that we've been covering here at The Information. Private equity has its sights set on AI as firms look for ways to apply the technology to their colossal portfolio of tech companies and beyond.

21:00The information senior finance editor Ken Brown wrote this week in his column about why that is a bigger deal than you might think. I want to bring him on to talk about that, and I also want to bring on our co-executive editor, Martin Peers, who I know has thoughts on all of this. Welcome to you both. It's great to have you here. Hey, Josh. Okay, so Ken, you started off your column this week by talking about the market dynamics for compute and the market dynamics for AI and how So sometimes those line up, sometimes they don't. Where are they at right now, and how are you thinking about that? Well, one of the points of the column is they kind of lined up a little bit.

21:40In the last year or so, you've had huge demand for AI infrastructure, right? They're building data centers as fast as they can, raising billions of dollars. It was unclear whether there was going to be demand for AI, the result of all what these data centers would produce. And so my point was this week, this has become clearer. there is more demand for AI itself, which then justifies the spending that we've had for the data centers. And this is because the PE companies have come knocking. This is so right. There were two deals in the works, they're not final, where groups of PE companies are aligning themselves with OpenAI and Anthropic to use their technology, to use the AI in their portfolio companies.

22:25And these are thousands of companies that they own and they are profit-making entities. They want to make these companies more profitable and sell them. And they're going to use AI to do that. And so that's going to create demand. And my view is it's going to create more demand throughout the whole economy. So, Ken, this is kind of interesting. I mean, you have these PE companies. They're sitting on, they own all these different companies, all different flavors of companies. And we know that private equity companies have a reputation of making large-scale changes across their portfolio. I mean, I guess my question is, this sort of still depends a little bit on how good the AI actually is, right?

23:04And if there's actual ROI from implementing these changes? Well, yeah, of course. I mean, if it doesn't work, it doesn't work. But they're going to be trying it in a lot of different ways, right? Because this has not been a great time for private equity. They have not been able to sell companies they own. They own a lot of software companies which have gotten hammered in the markets because everyone thinks AI is going to crush them. And these private equity firms don't sit back and just let things happen. They're going to be very aggressive in doing everything they can to spruce up, improve profitability and everything for these companies.

23:41And so my view is AI, they're just going to dive into AI and be very aggressive and try a whole bunch of things. and it's going to be a bit of a guinea pig situation, but on thousands of companies. And, you know, yeah, of course it could not work, but I think there's a lot of good evidence that AI is working in enterprise and companies. And so I think this is going to be a big test. Okay, so Martin, what do you think? You think it's going to work?

Read the full transcript

24:07Aaron Holmes:I think that Ken is right. There is a fair amount of evidence that it's working. It's working really well for the hackers who love to come into, are using it to really take over these companies, cause massive havoc. The people who are really looking forward to AI are the security firms who just see this as a massive boom for their business. When all these companies deploy AI at scale and it causes chaos, then the security firms are just going to really enjoy themselves. Look, I'm being a little, you know, going a little bit too far, But my point is that there is this incredible optimism right now about AI, but we are ignoring a few really important factors because no one likes to poo-poo these sort of, you know, popular things.

25:00Aaron Holmes:The first thing is that there is a real security risk with AI. There's no doubt about that. It is turbocharging, you know, any kind of breaches that occur. The second thing is that it often does not really work very well. I mean, it does some incredible things. On the other hand, when you ask to pull any kind of information, you cannot rely on it. Half the time, it's not accurate. But the biggest point that I really want to point out is everyone goes on about how demand for AI is off the charts. That's true. Do you know if you hand out anything for free, it will be popular? the open ai and anthropic are burning tens of billions of dollars over the next five years which tells you they are not charging the right amount for their service so as soon as they start charging the actual uh amount that they they need to actually make money i think we will have a much more realistic idea of what the uh demand is like okay ken what do you think Well, I think there's evidence that it's working in areas like service and in other areas that are getting adopted quickly.

26:20And so these are specific applications, and they may or may not run through software companies, SaaS companies that these companies, these smaller customers actually use already. So, I mean, I think the hacking issue certainly is an issue. You know, you're installing all this stuff that's out there on the web. It's a risk. I don't think these companies are going to do that or allow that. They're not stupid about these things. But I disagree with you, Ken.

26:49Aaron Holmes:I think companies are a lot stupider than you realize. I think there's enormous evidence that companies do things that hurt themselves because they see everybody else doing it. Well, you know, so I see, I mean, I'd say, you know, these PE firms are going to have, they're going to, what's interesting here is these guys own thousands. of companies and so they're going to build up a central repository of knowledge and these companies would not have been able to do it on their own uh you know a lot of these are midsize companies they wouldn't have been able to do it on their own without you know uh the god the guidance that they're going to get and so i i think they're going to not i think they're not going to screw that up it may not work and it may not work quickly but the demand is going to be there the other thing is if you look at the prices of tokens uh they've gone up a lot so a lot of this stuff is not free.

27:38And, you know, ChatGPT may be free, but the stuff that businesses are using is not.

27:44Aaron Holmes:That's true, but they're still not charging the right price. Well, Martin, let me push on you for a minute here. I mean, one class of companies that PE has loved are cybersecurity businesses. I mean, you know, they do these roll-ups of these observability companies all the time. I mean, these are companies that it would actually make a lot of sense, I think, to have AI to help them, you know, address these risks. So, I mean, there is a category of startups that you're, that I think, or not startups, companies that I think makes a lot of sense for PE companies to inject AI into, no? Of course. I mean, obviously, the security firms are the best positioned firms from this entire thing.

28:27Aaron Holmes:So, yes, if they can use AI to help them, the question really is, how much help will it give them? And I keep reading about people using AI for various things. Some of it clearly works. But I also see some of the other things that people are using for, and it's just made up demand. And I do get a sense that people are trying AI because they hear everybody else is doing it and they have to be seen to be kind of hip. But my point is it doesn't always work that well. And there's a real problem if it causes security issues. And then, as I said, it will long term, it will become expensive. So Ken, now put this in context of what we're seeing with the SaaSpocalypse playing out.

29:14And these PE firms have to exit at some point. And the market for exits today looks very different than five, six years ago when a lot of these deals were happening. How hungry are these firms for exits? Will we see them coming? Could this, I mean, the hope is it helps with the multiple, but i guess the question really is i mean is anything going to change the next couple years well i mean that's what they're hoping right i mean these these key firms are stuffed with with these sas companies it was a good business for them to buy regular income they could put a lot of debt on it and it would be just fine they could cut costs they got to fix these companies they got to get these companies up and running i mean look private equity is not anything except profit-making enterprises and the executives there their compensation depends on it so they are not going to sit around.

30:02They're going to cut costs, which is what they know how to do. And they're going to really try to get these companies to caught up with AI so that they don't become irrelevant and die. That would not help their year-end bonuses. So they're going to try a lot of stuff. And it may or may not work. It's just there's going to be demand. All I'm saying is there's going to be demand because there's going to be this stuff is... I mean, maybe it's going to ultimately be a test, right? There's going to be all these use cases. People are going to try to use all this stuff. And if it works, they're off to the races.

30:33If it doesn't work, then they're all going to struggle and have to figure out the next step. Well, okay, Martin, I'm going to give you a hypothesis here. You tell me what you think. So Ken's saying demand increases for the software. Let's say Anthropic. Demand's booming. Anthropic goes public. It finally sorts out maybe how to figure it out economically in the long run. the IPO goes well, SaaSpocalypse. I mean, I don't, you know, but the multiples all start to come up and then maybe there is an exit market in a couple of years for these companies.

31:09Aaron Holmes:I think it really depends for the software firms on the kind of business that they're in. I mean, I was talking to an investor yesterday who was making this very smart point, dashboard firms where all you were doing is you have this sort of fancy dashboard to track projects are probably doomed. And we can think of a few that we use, in fact. But then there are others where they have much more sophisticated software that is not so easily able to be replicated by AI, which will probably have a big advantage. And if those firms also are already integrating AI into their products, then they will probably be fine.

31:54Aaron Holmes:So I don't think you can make any kind of blanket statement about the entire industry. I think it really does depend. Great. Well, Martin and Ken, I want to thank you for coming on. That is Ken Brown, our senior finance editor, and Martin Pierce, our co-executive editor here at The Information. Our next guest is a former president of Tesla, one of Elon Musk's direct reports. He went on to be the COO of Lyft. And today, John McNeil is the CEO and co-founder of DBX Ventures. He also sits on the board of GM, Lululemon, and a number of other companies. This week, John published his book entitled The Algorithm, where he unpacks the formula that Elon Musk and company have used to build Tesla and SpaceX.

32:37I want to bring on John to talk about what he is seeing in the market right now. John, welcome to the show. It's great to have you here. Hey, nice to be with you. Well, congrats on the book launch. It's an exciting week for you, and I'm excited to talk about it. I have to tell you that I saw you're also on the board of CrossFit. Yes. So you're an avid CrossFitter. I am. And when that opportunity came up, it was hard to turn down. Okay. Well, have you heard of High Rocks? That's what I'm training for right now. I have. Yeah. Those are tough. Good luck. Well, CrossFit is not easy. I will tell you that much.

33:12Not for the faint of heart. That's for sure. Yeah. All right. Well, we can talk about that later on. Look, I want to talk about the book, but I also want to get your thoughts here at the outset on the current state of Tesla's business. It's been a while since you left the company. I mean, what do you think about the company's current strategy right now? Well, I think Elon came back from a trip to China about two years ago and said to his team, hey, look, the car business has been won by China. And so we're going to go, we're going to pivot and we're going to go all chips in on the two things that are going to win in the long run.

33:47And that is autonomous cars, so robot cars and robots in the factory. Humanoid robots could be the biggest consumer product of our lifetimes. And so he pivoted the company two years ago. You might remember then he canceled the$25 ,000 car. He canceled really any investment into superchargers, and the whole supercharger team was let go. and that was because he was making this hard pivot to what he saw as the next chapter for Tesla. And so I think, you know, now the challenge is maintaining that car business and the cash flow from the car business long enough to get through to the other side, which is volume production of robots, but also a volume rollout of robotaxi.

34:39Right. So let me ask you, do you think that Tesla can catch up to Waymo in that game? I think the chances of Tesla on the humanoid robot side are much stronger potentially than on the autonomy side. And the reason I say that is because I'm hearing from former colleagues at Tesla that they're blown away by the progress the Optimus team is making on a weekly basis. And the fact that they pulled forward the manufacturing launch for Optimus and then have also pulled forward the volume expectations into late next year. I think that's surprising a lot of folks, but not surprising the people that are on those teams.

35:22So I think they've got their work cut out for them in catching Waymo, but they may emerge in a market leadership on the humanoid robot brand. So this is kind of interesting though, because humanoid robots, I mean, they're a bit, we're already seeing the autonomous vehicles from other companies on the road. And, you know, Tesla also has its own cars out there. But humanoid robots, it's a bit further out. So I guess what you're hearing is you're actually more confident in that further out bet. Yeah, and it may not be further out. Like I think of both of these businesses, autonomous cars and robots, these are two businesses where we're behind China.

36:04And so if you want to see the movie and watch the movie first, you can see this movie playing out already in China. So China has autonomous cars on the roads in all of their major cities. And you can see what that's done to the ride share business in terms of share. You can also see humanoid robots in action across the economy in China from factories to people's homes. And so this is not future, its current state in China. And so I don't think you have, you've got to really stretch yourself intellectually to say, could an American company release a humanoid robot when the Chinese already have?

36:43Probably. And do you believe it's a long ways out? It's probably not a long ways out. Do you think that Tesla could merge with SpaceX? I think like one of the things I read about in my book, The Algorithm, is one of Elon's organizing principles is simplicity. And if you approach it just from that standpoint, it's easier to run one public company than it is two. So you think he is going to merge that? I don't have any knowledge of that. I'm not sure what side of the cal sheet bet I would take. Probably I would take the over on that. I think the odds of that are high. And the reason I say that is because he does like simplicity.

37:25He'd rather run one public company than two. And the second is that Tesla, if you consider their two big products are going to be human-oriented robots and autonomous cars, at the core of those products is AI. And so it just makes sense to put those things right next to the AI asset, I think. Even though you have a company, I mean, SpaceX, you know, I'm looking at the financial profiles of these companies. And we talked a lot on this show about the financial profile of SpaceX and XAI and how XAI is obviously a lot more capital. Well, not capital intensive. XAI is a bit tougher of a financial position right now.

38:03They're in total burn mode. Right. And so, I mean, so that combination itself, it's not really helping SpaceX. And this combination, I mean, it's not really going to help Tesla on a financial level. Well, I think on a financial level, no. But again, like he thinks about the long term and thinks about like, what is the most efficient investment of his time? And does he want to invest time in the overhead and SOC compliance and everything else of running two companies versus one? And so you'll think about that as his organizing first principle versus the cleanliness, maybe of the P &L or the balance sheet.

38:43Hmm. So let's talk about the book that you are launching or publishing this week. I have it right here, The Algorithm. It was a great read. So you talk a lot about Elon Musk's own playbook. And what I wonder is, were there any parts of your playbook personally that differed from his? And if so, how did that end up playing out? So the algorithm is really the output of a bunch of mistakes we made. And we developed this framework over time. And it's really the entire team that was at Tesla during this time frame where we went from threat of bankruptcy to threat of bankruptcy and product launch to product launch that almost led us off the edge of bankruptcy.

39:31And we developed this framework to say, like, how do we get ourselves out of this ditch and never get in it again? And so I think that framework is the result of a lot of mistakes that were made. Some were his, some were ours. And those differences kind of get sanded down to give a framework where the framework is sort of the operating system of how Tesla operates today. and the question that I got the most of the time was like can you do this without Elon and so the book tells the stories of the people who are on the front lines doing this without Elon without me uh utilizing this framework to drive innovation at a faster pace than anybody in the industry whether that was cars or rockets uh and then I've taken this into our own startups and taken this into uh legacy companies like General Motors who've been around for more than 100 years and started to deploy some of these techniques.

40:29But it wasn't so much about the differences as it was about us being in this super fast learning loop together and learning from huge mistakes like automating the Model 3 line before we'd even produced a car. Now, the interesting part about the playbook is automate. You basically say you should automate things last. It's not the first thing you should do. It's step, I think it's five or six. I can't remember how many there were. Five. Last step. Last step. So we're in this era of AI now where automation is the name of the game, right? It's like, I don't need any people. I can just do it myself, right?

41:07I mean, you know, and it's not even going to be me. It's going to be an agent. I mean, automation is the first thing that people are thinking about. You're a venture investor. How do you think about this playbook then in this era of AI where that is literally what people are trying to do? It's interesting. There's a professor at Stanford, Matt Glickman, who teaches entrepreneurship. And he's coined this phrase, concierge before you automate it, which basically means do it manually before you automate. Because if you automate a bad or flawed product, you're just going to speed your time to that bad answer.

41:41And so like a great example, this way outside of Tesla, I'll give you two, is DoorDash. So DoorDash, five Stanford CS majors, put up a PDF of restaurant menus with a phone number at the bottom. Now, these guys had the talent to hit the keyboards fast and automate fast. But what they did instead was they wanted to learn the business so they could optimize the flow and then have something that could scale. So they took those orders. They went to the restaurants, ordered the food, arranged the payment, took it out on delivery runs. And they started to see how you can optimize. and then they automated.

42:19And I mentioned this incident we had with Model 3. We desperately needed to get Model 3s out because we needed the cashflow. But we automated a line first before we'd ever produced a car. And it turned out that automated line never worked. And we had to build a tent in the parking lot and start to produce Model 3s by hand to save the company. And that's when the rule emerged. Like we cannot do this again. And the big lesson from this is you got to automate last because if you don't have a process in place, you're just going to spend a lot of time automating really bad process and probably creating really bad outcome, which we almost did.

42:57Okay. So now extend that, though, to the big question of today, which is how much demand there will be for AI. I mean, I hear you saying, hey, automation is not the answer. You should actually think about automating less. And here we have companies whose entire pitch is we will automate it all for you. automating last isn't automating less uh you're gonna last okay okay right automate last and you're gonna automate uh but it's just that you're picking your spot to do that um and and i think jobs taught us this uh elon's definitely taught uh taught us this that yeah perfect product basically sells itself so spend the time up front to design the product architect the product tested in the hands of real people and then pour the concrete over the process in the form of automation.

43:45Right. And I mean, I guess it is, I guess this is sort of the interesting exercise to think about is when the product itself is automation, I guess what I hear you saying is even if the product is automation, the process by which you sort of rolled this out in the company or I don't know, build the software, et cetera. I mean, I'm basically just trying to, I'm trying to understand how you might think about the SaaSpocalypse and the idea that we can just build our own software, you know, using these automated tools. You know, if you think that that's a viable argument in this world where you're saying, hey, you've got to think about a lot of things before you just make it on your own.

44:24Absolutely. You got to think about the architecture. You got to think about, you got to know the problem in depth. And that's kind of what I'm saying. I've got a couple of friends who started now a very successful AI company called Maven AGI. And Maven outperforms all of their peers in the customer service AI space. So they deflect 95 % of calls that are coming into one of their customers' organizations. But when they started, they didn't start with hands-on keyboard to start. They started to break down the problem and said, hey, when you get a customer support call, how many of the support calls can be answered with straight search index?

44:58so I'm not going to run an expensive AI in this. Turns out a big chunk. Then they said, how many of these questions should I answer with a small model? They do the experimentation and they figure out there's another chunk that can be answered with a variety of small models. And then the super hairy questions get answered by a large model with a human in the loo. So they've got like four layers of architecture that exist now because they just went slow at the beginning and they broke down the problem in depth before they were going wild on the keyboard or having their coding agents go wild. And that's what I'm talking about.

45:35Like understand the problem in depth so you can create a company that's not deflecting 60 % of support calls, but it's deflecting 95 % of support calls. And that's all a function of going essentially slow at the beginning so you can go fast with AI tools and automation at the end. Right, great. Well, John, I want to thank you for coming on. That is John McNeil, author of The Algorithm, here on TI TV. That does it for today's show. A reminder, we are on this stream Monday through Friday at 10 a.m. Pacific, 1 p.m. Eastern. I want to thank you all for tuning in. We really do appreciate your viewership.

46:09Make sure to subscribe to the information on YouTube and follow us on X, Instagram, TikTok, and check us out wherever you get your podcasts. I'm already excited for our next show on Monday. Have a great rest of your Friday and have a great weekend. Bye-bye for now. Thank you.

46:54Thank you.

47:24Thank you.

From the publisher

Former Tesla Executive Jon McNeill talks with TITV Host Akash Pasricha about Elon Musk’s management "algorithm" and why Tesla is pivoting to humanoid robots. We also talk with Aaron Holmes about Microsoft’s hiring freeze in its Azure cloud and sales divisions, Jing Yang and Juro Osawa about Tencent’s aggressive bet on OpenClaw in the China AI race, Julia Hornstein about the next generation of Silicon Valley General Partners. Finally we get into the private equity AI boom in this week’s Editor’s Cut with Martin Peers and Ken Brown.


Articles discussed on this episode: 

https://www.theinformation.com/articles/microsoft-freezes-hiring-major-cloud-sales-groups

https://www.theinformation.com/articles/tencent-bets-openclaw-make-lost-ground-china-ai-battle

https://www.theinformation.com/newsletters/dealmaker/next-general-partners-winning-deals

https://www.theinformation.com/projects/general-partners-venture-firms-2026


Subscribe: 

Sign up for the AI Agenda newsletter: https://www.theinformation.com/features/ai-agenda


TITV airs weekdays on YouTube, X and LinkedIn at 10AM PT / 1PM ET. Or check us out wherever you get your podcasts.


Follow us:

X: https://x.com/theinformation

IG: https://www.instagram.com/theinformation/

TikTok: https://www.tiktok.com/@titv.theinformation

LinkedIn: https://www.linkedin.com/company/theinformation/

More from The Information's TITV

All 304 episodes
Tencent’s OpenClaw Obsession, The Information’s Next GP List, Former Tesla Exec on Elon’s IdeologyThe Information's TITV · 47 min
Listen in VO