Latest / Newcomer Pod / Listen: Bullish on AI, Bullish on Self-Driving, But Apprehensive About OpenAI & Tesla (with Jon McNeill)
Transcript
- 0:00Hi, I'm Eric newcomer and this is the Newcomer podcast.
- 0:07Madeline is out this week. I have a fun Co host special
- 0:10guest with me, John McNeil. Hey everybody, nice to be.
- 0:13With you, Eric, you were on an earlier podcast.
- 0:16Thanks for coming back to our new iteration.
- 0:19John and I go way back. He was the Chief operating
- 0:22officer at Lyft where I got to know him.
- 0:25I what followed you to Las Vegas and wrote a profile?
- 0:29And obviously we talked a lot over the years about the ride
- 0:32hailing industry. You were also top executive at
- 0:35Tesla for a long time. What was your title at Tesla?
- 0:38It it started out as president of sales and then became
- 0:40president of sales and service and then it was a bunch of stuff
- 0:43after that. So at the end, I think I was
- 0:45just president. There was too much.
- 0:48There was too much. One of the people behind the
- 0:51scenes doing a lot of work. Well, Elon is the face of the
- 0:55company and now you're on. You're on the board of GM, which
- 1:01obviously puts you in sort of cruise world a lot and then your
- 1:05day job as an investor helping to what incubate do you use
- 1:10incubator? I never know.
- 1:11Accelerate, incubate. What word people?
- 1:13What is the word you prefer and how you think about starting So
- 1:15the. The name of our firm is DVX
- 1:17Ventures and we describe it as we invent companies because
- 1:19literally we, we come up with the idea and we build the
- 1:24initial product, put it out in the, put it out in the world and
- 1:27start to work on traction. And then once we get traction,
- 1:30we hire a team. So we're we're kind of inventing
- 1:33and then hopefully accelerating and incubating all at once.
- 1:37Do do you miss the operator world or like to be at a company
- 1:41like Tesla where it's, you know, so much work to being an
- 1:45investor? Does it feel like a whipsaw or
- 1:47how do you how do you experience the life of an investor?
- 1:51It does feel funky. One of the questions when I was
- 1:54starting DVX, Reid Hoffman was super generous with his time
- 1:59when and said, hey, come down and let's spend an afternoon
- 2:03together and run me through how you're thinking about doing
- 2:05this. And one of the questions he
- 2:07asked me in that session was will you ever be ACEO again?
- 2:12And I said, why do you ask that? He said, well, because I
- 2:15actually had the idea to set up one of these idea studios, and I
- 2:20did it with Max Levchin. And yeah, the very first company
- 2:24that we that we invented was a firm.
- 2:26And he jumped in as CEO and he said we'd never thought to have
- 2:30a talk about this. And all the sudden I'm giving
- 2:32money back to investors because my partner's gone.
- 2:35He said, you really need to make this decision.
- 2:37And I said, I've been ACEO seven times.
- 2:40Like I've checked that box and I really like working with other
- 2:43CEOs and helping them. And so I, I said, I think I can
- 2:48firmly said, I'm not going to be ACEO again.
- 2:51And I, and that's been really helpful.
- 2:54But I got to say like I do miss the energy and being energized
- 3:00by being day-to-day neck deep in a business with a team solving
- 3:04problems and being an investor is several orders removed from
- 3:10that. And it's it's hard for me some
- 3:12days, I have to admit, like I just.
- 3:13Thankfully, thankfully for us as an investor, you're much more
- 3:17free to opine on the world that probably once we're.
- 3:21We're have the company do this in an operating role, that's for
- 3:24sure. Well, first, yeah, I want to
- 3:28talk about Open AI. So for context, you know,
- 3:31literally like a year ago, I did AAI draft with my friends Max,
- 3:35Shawn, James Wilsterman, who I hope Co host through Royal
- 3:38Valley with on one of our podcasts.
- 3:41And in the draft, I basically insisted that I be able to pay
- 3:44for the right to go first because they wanted to sort of
- 3:47randomize. And I was like, no, the whole
- 3:49bet is how much you're willing to pay for open AI.
- 3:51And I took, I think it was a minus $75 billion handicap.
- 3:56So Data Bricks was the second ranked company.
- 3:58So that is a real cost. So I was extremely bullish in
- 4:02this $157 billion valuation would seemingly validate that.
- 4:07So yes, exactly why have I become bearish?
- 4:09I think #1 is just the commoditization of foundation
- 4:13models. I feel like when I made the bet
- 4:16it was, you know, at Cerebral Valley, it felt like we're all
- 4:20pivoting off Chachi, BT and open AI and like everything is a
- 4:24reaction to the magic there and people trying to find the magic
- 4:28for their own companies. And I just feel like that's
- 4:31really changed where people are much more excited about
- 4:34applications and much less excited about any of the
- 4:37foundation models. And then the second is just sort
- 4:40of the, the weaknesses that I knew about Sam Altman now really
- 4:44manifesting in open AI. You know, it wasn't clear, maybe
- 4:48sort of my, my view on some of his shortcomings wouldn't
- 4:51translate into open AI. But then you see so many people
- 4:54leaving, you know, the board trying to oust him and just like
- 4:57overall dysfunction and all these really wild prom, you
- 5:01know, ideas like a $7 trillion project to build chip
- 5:05infrastructure and just, you know, sort of sort of a crazy
- 5:09idea. So, so I think it would it would
- 5:11those it was the sort of weakness of foundation models as
- 5:14a business and and the weaknesses of the the CEO leader
- 5:18who's so much of the value was was hung on.
- 5:22Yeah, like I, I, I think, you know, I the framework that's
- 5:25dominated tech for the past 50-60 years is Moore's Law where
- 5:31you've got essentially a every 18 month, the doubling in the
- 5:35capacity, processing capacity of the chip.
- 5:38And that starts to define all kinds of businesses and business
- 5:41models around that. Because the chips get more and
- 5:44more powerful and you can do more and more with software and
- 5:47it's enabled things like obviously the PC and the mobile
- 5:50phone. But underlying that is kind of a
- 5:53business framework, which is OK, you've got 18 months.
- 5:57If you've got a an advancement happening every 18 months as a
- 6:01business, you got 18 months to make your profits.
- 6:04And then that gets commoditized and sort of gets lower placement
- 6:07on the shelf and you've got another product that gets top
- 6:10shelf placement and you get to make money off of that.
- 6:13To your point on the foundational models, these
- 6:15things are leapfrogging each other roughly every 30 days with
- 6:18new releases and what I. Haven't understood you need 18
- 6:21months, you have 30 days, right? You have 30 days, so now you
- 6:25have 30 days to collect rents. And I don't understand how you
- 6:30can get a margin profile that's attractive in a business that
- 6:33commoditizes that fast. And so I have, you know, I, I, I
- 6:38understand that that roughly 15 times revenue doesn't seem like
- 6:43a crazy valuation. I understand that, but it also
- 6:47it it lacks an an understanding of what the quality of the
- 6:51margins are. The multiple, if you just look
- 6:54at revenue, seems reasonable compared to Facebook and Google
- 6:56IPOs. This is a fast growing company
- 6:58with a lot of revenue. But then if you try to think,
- 7:00you know, both Facebook and Google were profitable, Facebook
- 7:03was making like a billion dollars in profit when it went
- 7:06public. And, you know, the open AI is
- 7:09losing, what, $5 billion probably this year, more
- 7:12probably next year. Yeah, it's spending a lot on a
- 7:14product that is way less differentiated than Facebook and
- 7:18Google were. Exactly.
- 7:20And I think you've got, you've got a couple of a couple of very
- 7:23big kind of questions to ask, which is what, what's the margin
- 7:27profile in the business that would justify that valuation?
- 7:30And the second is you've got a cap X angle and or issue in this
- 7:35business that you don't have in a lot of other businesses where
- 7:38you've got to keep building clusters, rolling out clusters.
- 7:41We saw, Sam said last week as he was kind of poking Microsoft a
- 7:46little bit to say, hey, are you guys going to build my next
- 7:49cluster or not? Right, right.
- 7:51And if they don't, of course, he's got now, you know, $6
- 7:54billion in the bank plus another $4 billion credit line.
- 7:57So he could probably take a pretty good stab at building
- 8:00that cluster. But it just speaks to the CapEx
- 8:03that is necessary in this business to keep maintaining
- 8:07your foundational model lead. And I think this, I think this
- 8:12race is going to skinny down to a few players fairly quickly.
- 8:17But again, coming back to your valuation question, I think all
- 8:20of these would be really interesting to start to get some
- 8:23visibility into. But investors today don't have
- 8:26that. We don't have visibility into
- 8:27margins, what the future margin profile looks like and we don't
- 8:31have visibility into how much capital is gonna be necessary to
- 8:35create those margins. And sort of the, the regular, we
- 8:39don't have access to the exact terms of like the Microsoft
- 8:42deal, which means that Microsoft gets to like gobble up a lot of
- 8:45profits that certainly Google and Facebook didn't have those
- 8:49types of deals. I, I wanted to say, you know, in
- 8:52the piece, I talk about Brad Gerstner at altimeters, his bowl
- 8:56case for open AI, which he delivered at Madrona's IA Summit
- 9:02in Seattle recently. And you know, his case is very
- 9:05much tied to ChatGPT, just like, forget foundation models.
- 9:09Yes, they're, they're going to be commoditized, but ChatGPT is
- 9:13the iconic product. You know, Google had Alta Vista
- 9:17and Yahoo and like, that didn't matter.
- 9:18It's it was Google search. And this is Chachi BT It's in
- 9:22the lead. It's making billions of dollars,
- 9:25growing quickly. This is, you know, I sort of
- 9:27joked in the piece, Chachi BT is what every student has learned
- 9:30to cheat on their essays with and they're gonna take it with
- 9:33them when they go to the workplace.
- 9:35I mean, people are very dependent.
- 9:36It's funny, I was looking at Google Trends.
- 9:38If you look at Chachi BT, it's like during the work day, people
- 9:41are using it all the time and it sort of columns on the weekends.
- 9:43Like people are really using ChatGPT.
- 9:46So what do you, Yeah. Do you have a take on that
- 9:49argument, whether it's a big enough consumer product to to
- 9:52justify how people are thinking about Open AII?
- 9:55Think you, you describe Brad's case really well.
- 9:57I mean, he describes open AI, ChatGPT and open AI as having
- 10:02reached consumer escape velocity.
- 10:04And so you've got a multi billion dollar business, you've
- 10:07got a, you know, $5 billion business heading to an $11
- 10:10billion business over the next year.
- 10:13As his analysts kind of built the slides that he used in the,
- 10:16in the presentation at, at Madrona, but also last week in
- 10:20his investor, his LP summit, His argument is this is growing way
- 10:24faster than Facebook grew, than Google grew, etc.
- 10:27And so I, I hear that like I do think there is on the consumer
- 10:31side, there are folks that reach the breakout that become the
- 10:34verb and open AI has become the verb for AI, right?
- 10:38And then they do get position with in the consumer markets of
- 10:42significant share and they get to collect rents.
- 10:45We've certainly watched Google do this in search.
- 10:47And I, I got to put their mirror on myself.
- 10:50My own consumer behavior is that I am searching a whole lot less
- 10:55and I'm asking Google and anthropic and perplexity but
- 10:59questions a whole lot more cuz I just don't have to waste time
- 11:02poking in to link. Sorry Google, not you're not
- 11:05asking ChatGPT or. I'm asking chat, yeah I'm asking
- 11:07ChatGPT and I'm not asking Google because I used to.
- 11:12I just am tired of clicking through links to try to find the
- 11:15right answer and I get the right answer.
- 11:17Served up most of the time now. Very few hallucinations with
- 11:22either chat. Interesting, I don't feel that
- 11:23way. I I want to be able once they I
- 11:25like the perplexity idea that I don't use it regularly where I
- 11:29at least want to know why it's saying what it is.
- 11:31And it's very hard for me. Just like, OK, open AI is giving
- 11:34me some factual answer, but now I'm sort of lost.
- 11:36I'm like, was it right? Or like, what was it not?
- 11:40You know, I'm, I think I'm someone on Google search where
- 11:43I, I want to find the original link, you know, like so much not
- 11:46even like the Business Insider rewrite of the original link.
- 11:48You know, I want to find the original original.
- 11:51And, and I feel like to me the chatbot sort of get you so far
- 11:55away from that. All right, net, net at 157
- 11:58billion. Are you a bear or are you a
- 12:01bull? I may end up eating my words,
- 12:05but if I'm an investor charged with finding 5 to 10X
- 12:10opportunities, I just don't see the rent story here that creates
- 12:16a trillion dollar company. And so I don't know how to get 5
- 12:19or 10 times more money at that valuation.
- 12:21And so I guess, you know, we voted with our feet.
- 12:23We were, we were asked and flattered to be asked by a
- 12:28number of investors in this round if if we wanted to
- 12:31participate and we didn't. Because lots of SPVS running
- 12:34around. SPVS run around, you know, Koslo
- 12:37was real real today, almost a half a billion dollar.
- 12:40The SVP that my God or SPD that went in so but we didn't cuz I
- 12:45couldn't figure out how you're gonna get 5 or 10 times return
- 12:47on this. And I still understand.
- 12:49I know how Apple collects rents and it's a margin story and I
- 12:52know how NVIDIA collects rents and that's a very similar margin
- 12:55story. In fact, double the gross
- 12:57margins of Apple. So I understand why those
- 13:00companies are valued the way they are.
- 13:01I just can't, I don't have the visibility to understand that
- 13:04now with with open AI and I could be wrong.
- 13:07Over time, you just can't ignore the executive turnover.
- 13:10When I see turnover like that in in amongst executive teams, like
- 13:17this much, this deep, this rapidly, that points to
- 13:20something else going on that I'd really want to understand as an
- 13:24investor. Is that cultural?
- 13:26Is that just merely opportunities that people are
- 13:28seeing on the outside as a disagreement of strategy,
- 13:30whatever that is? I'd like to understand that
- 13:33because there are very few companies that over time have
- 13:35that level of Isaac turnover and are successful.
- 13:38Not saying there aren't any. Yeah, I'm curious what you have
- 13:41to think about this. This sort of fits both honestly,
- 13:43like some of the Elon world stuff and this open AI
- 13:47situation. You know, I, there was this
- 13:49founder mode discussion where Paul Graham wrote the essay
- 13:51saying, you know, you need a founder who's really in charge
- 13:54of their company and talks a lot about Brian Chesky and he
- 13:57compares that with Manager Mode. I, I made a snarky tweet, you
- 14:01know, when that was all happening, comparing Uber stock
- 14:04price and Airbnb, you know, Uber being the manager mode and
- 14:07Airbnb being the founder mode. And Dara has performed better
- 14:12over the last few years anyway. But but a subtle point that I've
- 14:16been making is that a lot of times you want these sort of
- 14:19founder mode execs right below, you know, the CEO.
- 14:24And that we placed so much focus on the CEO and not a lot of the
- 14:28execs they have out running the business.
- 14:30And that in the Uber case, there were these sort of founder mode
- 14:33execs who understood the business and been there pretty
- 14:36early on. Like, how do you think about
- 14:38this sort of founder mode discussion as someone who has
- 14:41been sort of a #2 at a lot of these companies?
- 14:44Well, like I was. I was a six time founder before
- 14:49I met Elon and so I and so the the whole discussion of founder
- 14:53mode totally resonates with me because I because I got to
- 14:57experience it first hand. But then like to answer this
- 15:00question when I got to when I got to know Elon and Elon asked
- 15:04me to join the team at Tesla, I I said to myself, like electric
- 15:10vehicles wasn't my thing like I didn't invent this.
- 15:12This was I wasn't the founder, but I found myself like just
- 15:15literally a few days in having absorb that this the whole EV
- 15:22opportunity in space and Tesla's a company and literally adopted
- 15:26it and made it my own. And so I pursued, I found I was
- 15:29pursuing that on a daily basis with all the fervor that I
- 15:33pursued my own companies as a founder.
- 15:37And I watched Gwen doing this at SpaceX and we might want to talk
- 15:41about that, but I because I think she's a good example of
- 15:44this too. And, and again, I had this like
- 15:47similar conversation with Reed Hoffman where he described he
- 15:51had founded LinkedIn and he was operating as the founder.
- 15:54And then and then Jeff Weiner came in five years after the
- 15:58founding, roughly, if I have the story right.
- 16:00And Jeff took LinkedIn to a whole other level and Reid
- 16:04started almost immediately to call him a Co founder.
- 16:07And when I asked Reid why, he said because he acts like a
- 16:10founder. He acts every like every the
- 16:12founder I am. And there are cases where, like
- 16:15you said, Dara came in and really took Uber on a valuation
- 16:20journey that is a significant departure from Lyft.
- 16:24Satya has come in post Bill Gates and Steve Ballmer and you
- 16:29could argue like the value that he has created is incredible,
- 16:33just absolutely incredible. And there are others Brian
- 16:38Nichols, who's it was at Chipotle, is now at Starbucks.
- 16:42You see these CE OS who were able to come in and make the
- 16:44mission their own and therefore operate like a founder, but they
- 16:48also have operator chops so they can scale these businesses and
- 16:54and I think those those leaders are super interesting.
- 16:56So I think it's, it's just a more nuanced framework than just
- 17:01saying there's founder mode and manager mode and.
- 17:04And this is. Only two things, like I don't
- 17:06think that that really works out in the real world.
- 17:09And as a shareholder, you want to find the Sachas and the
- 17:12Bryans and the Quinns to come in and the Jeffs to come in and
- 17:17scale your company because they create so much value.
- 17:20All right, well, let's that the Tesla self driving date.
- 17:26Do you know what is he calling a demo day?
- 17:29We Robot is what it was called. We.
- 17:31Robot Little twist on Isaac Asimov's iRobot, that famous
- 17:37book. But it was, yeah, the the event.
- 17:39And a lot of sci-fi borrowing overall at the event.
- 17:43Anyway, yeah. Do you want to give us a rundown
- 17:44of what what stood out to you? This was late last week I think.
- 17:50What stood out to me was #1 like, I will, I have learned not
- 17:56to discount Elon, not to bet against Elon because he finds a
- 17:59way to win. And there has been kind of this
- 18:02multi year build up to this cyber, cyber cab, robo taxi
- 18:08event and full self driving getting to really it's, it's the
- 18:13potential of its name, full self driving.
- 18:15And so I have like a lot of people was really looking
- 18:17forward to this to say, OK, I wonder what he's got up his
- 18:20sleeve now because it's been in tech circles, pretty well known.
- 18:26People move around these companies a lot between Tesla
- 18:29and Waymo and Cruise. And so you sort of get a sense
- 18:33of what everybody else is doing because of this employee
- 18:36transition between the three. And you know, it was pretty well
- 18:40known that about a year ago, they ripped up their current
- 18:44script of, of deterministic neural Nets and went to a end to
- 18:50end LLM. And you could see in the Tesla
- 18:54tracker, this is this website that tracks each release and
- 18:58self reported from Tesla owners how many miles between
- 19:01takeovers. So in other words, are the cars
- 19:04doing something you're not comfortable with, you take it
- 19:06over. And that's, it's called an
- 19:07intervention. And so the key metric in the
- 19:10business is miles between intervention.
- 19:12And just to set the set the story, like if you're, if you've
- 19:16got a permit from the state of California, you got to report
- 19:18this data to them. And granted, there's a little
- 19:22bit of noise around the, the, the definition of an
- 19:24intervention, but all this data is published for Waymo, Cruz,
- 19:31Zoox, Aurora, you name it, that are on the roads in California.
- 19:35And you'll see that over the last year, Waymo and Cruz have
- 19:4110s of thousands of miles between interventions.
- 19:45And Tesla was at like 12, right? We're talking, you got to get on
- 19:49a logarithmic scale to even plot those companies on the same
- 19:53graph. And Tesla's now up to 14.
- 19:58And and they're making progress. You can see it with every
- 20:02release. But that product, that progress
- 20:05has to be orders of magnitude to be able to get a job.
- 20:10And so I thought we were going to start to hear the technical
- 20:14side of that. Story on the robotaxi night and
- 20:20we didn't hear any of it it was a 19 minute presentation half of
- 20:24which was robots dancing around and it was almost like they had
- 20:28lost the script and and you know I so I think I like a lot of
- 20:33investors was was disappointed this some were dismayed the fan
- 20:39the fanboys are definitely still.
- 20:40The stock market didn't like it, right.
- 20:42It was down like 8% or something because.
- 20:44The market is struggling to see, OK, what's the path?
- 20:47As you probably know, like the FSD system in a Tesla today is
- 20:51what's described as level 2. And that means as a driver you
- 20:55can't take your eyes off the road.
- 20:57And Level 3 is where you can take your eyes off the road.
- 21:00So now you're hands free and eyes free and level 4 is you
- 21:04don't have to pay attention at all and the car can just drive.
- 21:07And that's where Waymo and crews are.
- 21:09So you've got like Waymo and crews at level 4, you've got
- 21:12Tesla down at level 2. And, and I think what investors
- 21:15were looking for was show us the road map to go from level 2 to
- 21:19level 4 to at least get equivalency with these other
- 21:23companies that already have millions of miles of, of
- 21:27driverless, driverless experience.
- 21:29And when investors didn't hear that, you watch two things
- 21:33happen. One is the Tesla stock went
- 21:34down. And the second is the Uber stock
- 21:37went up because I think folks read their shorting and they had
- 21:42to cover their shorts of Uber and Lyft to say, gosh, we
- 21:44thought this was going to be a big announcement with timelines
- 21:48that were imminent, and this was really going to damage Uber's
- 21:50business. And they didn't hear that.
- 21:53So they reinvested into the rideshare market and disinvested
- 21:59from the Tesla story, which I think is really interesting.
- 22:03To paint the picture of the day for people.
- 22:05I mean, Elon has been hyping for years self driving, you know,
- 22:10they are going to be a self driving company.
- 22:11They're going to replace Uber and Lyft.
- 22:13That's why you would have people shorting these companies.
- 22:16It's like, OK, this is going to be the big reveal, especially at
- 22:19a time where Elon's focus on the election, he's focused on XAI.
- 22:23You know, it's SpaceX, which we'll mention in a little bit,
- 22:26is having great moments. Tesla is big public company that
- 22:29is fanboys are heavily invested in is going to have its moment
- 22:32and he's going to present everything.
- 22:34And what do we really see? We see sort of like proto, you
- 22:38know, we see like a bar with two doors and no steering wheel.
- 22:41We see these cool robots that are almost certainly operated by
- 22:46human beings. But what we don't see is some
- 22:48huge leap in self driving. Tesla is the ultimate sort of
- 22:52gets a lot of mindshare relative to to where the the tech the
- 22:56tech is. I would describe this, this is
- 22:59up to Elon to confirm this or not, but I would describe the
- 23:02Rocket business as his first love.
- 23:05And I would describe Tesla as as a business that he was an
- 23:11investor in, but a reluctant CEO.
- 23:13He had to come in to save the company, not only with money but
- 23:17with leadership. But I think his first love is is
- 23:23being at at Starbase, watching those things go up and watching
- 23:28them come back down and pushing the engineering teams in his
- 23:32role there to. And Tesla's the cash cow.
- 23:34That's the thing that makes this all work.
- 23:35That's how he buys Twitter. Well, it's a it's, it's
- 23:38certainly the personal, it's the personal asset that allows his
- 23:42personal investment into SpaceX, etcetera.
- 23:45But now you could argue that on its own, SpaceX is a cash cow.
- 23:48Yeah. I mean, they've got dominant
- 23:49market share, delivering payloads to space.
- 23:52And now they've got Starlink, which is the only product of its
- 23:55kind and no competitors. And, and so they've got 2 cash
- 24:00flowing businesses that are really attractive now at SpaceX.
- 24:03So I think SpaceX has got is, is a fantastic cash flowing
- 24:07business in its own right. But Tesla stock allows Elon to
- 24:10do things like X and other cash consumptive consumptive dreams
- 24:15that he's got. John McNeil at DVX, thank you so
- 24:18much for coming on the show. This has been a lot.
- 24:19Of fun. It has been a lot of fun.
- 24:20Eric, good to talk to you.