On this episode of “Okay, Computer.” Dan Nathan and Deirdre Bosa of CNBC discuss her interview with the Lyft CEO, AI for Uber/AirBnB, and Elon Musk renaming Twitter X. Later, Dan sits down with his “On The Tape” co-host Danny Moses and Michael Dempsey, managing partner at Compound, to discuss machine learning, AI, electric vehicles, and crypto.
0:00 - The State Of Ride Share
10:20 - X Marks Twitter
18:15 - Ad Break
19:40 - Michael Dempsey Joins
27:35 - Self-Driving
34:45 - Women’s Health
39:00 - Crypto
49:10 - A.I. Mania
About: Each week our tricked-out team of tech investors and former operators breakdown the biggest headlines and themes in both public and private markets, with a specific focus on the intersection of web2 and web3. We will be joined by some of the most influential voices in tech, media, and crypto leaving listeners with fresh perspectives on increasingly complicated topics impacting their lives and investment portfolios. Okay, Computer. Podcast is hosted by Dan Nathan and Deirdre Bosa. Follow OkayComputerPod on Twitter.
---
View our show notes at RiskReversal.com
Learn more about Ro body: ro.co/okay
---
Email us at contact@riskreversal.com with any feedback, suggestions, or questions for us to answer on the pod and follow us @OkayComputerPod
We’re on social:
Follow Dan Nathan @RiskReversal on Twitter
Follow @GuyAdami on Twitter
Subscribe to our YouTube Channel: https://www.youtube.com/channel/UCRAOycPjsSgcEyQcuJD_ENA
0:00 - The State Of Ride Share
10:20 - X Marks Twitter
18:15 - Ad Break
19:40 - Michael Dempsey Joins
27:35 - Self-Driving
34:45 - Women’s Health
39:00 - Crypto
49:10 - A.I. Mania
About: Each week our tricked-out team of tech investors and former operators breakdown the biggest headlines and themes in both public and private markets, with a specific focus on the intersection of web2 and web3. We will be joined by some of the most influential voices in tech, media, and crypto leaving listeners with fresh perspectives on increasingly complicated topics impacting their lives and investment portfolios. Okay, Computer. Podcast is hosted by Dan Nathan and Deirdre Bosa. Follow OkayComputerPod on Twitter.
---
View our show notes at RiskReversal.com
Learn more about Ro body: ro.co/okay
---
Email us at contact@riskreversal.com with any feedback, suggestions, or questions for us to answer on the pod and follow us @OkayComputerPod
We’re on social:
Follow Dan Nathan @RiskReversal on Twitter
Follow @GuyAdami on Twitter
Subscribe to our YouTube Channel: https://www.youtube.com/channel/UCRAOycPjsSgcEyQcuJD_ENA
Category
🗞
NewsTranscript
00:00 (upbeat music)
00:01 - Welcome to OK Computer, I'm Dan Nathan.
00:03 I am joined here with Deirdre Bosa.
00:05 She is CNBC's Tech Check host.
00:08 Hey, De.
00:10 - Hi, I'm coming to you from a San Francisco summer.
00:12 We finally got the good weather.
00:13 - Wait, just so you know, and I saw this headline,
00:16 it was on Bloomberg today,
00:17 that in Q2, demand for office space in San Francisco
00:22 was up 10% in the quarter,
00:24 and I think the editorial comment was,
00:27 is the bottom in for San Francisco real estate.
00:29 Does it feel like there's a bit of activity
00:32 going on over there that you've been just kind of waiting for
00:34 over the last few months, quarters, that sort of thing?
00:37 - Dan, you must be a mind reader.
00:39 I've been working on this piece for the last week
00:41 with NBC Nightly News.
00:42 So you will see it on Lester Holt tomorrow evening,
00:45 but that is exactly what we look at.
00:48 Specifically, the generative AI hype and the activity,
00:53 it's all here, it's all in San Francisco.
00:55 You might've heard that people moved out
00:56 during the pandemic, they're all coming back,
00:58 and we visited a huge headquarters for scale AI,
01:02 a big generative AI company,
01:04 but that is really what's providing hope,
01:06 and there are signs of this revitalization.
01:08 - Oh, that's great.
01:09 I mean, I actually did not know that was coming,
01:11 so I think that's interesting, and that's super cool
01:14 that you're gonna be on Lester Holt, the NBC Nightly News.
01:18 Listen, you and I got a lot to talk about.
01:20 We're gonna be really quick here,
01:21 'cause I know this is probably one of the busiest weeks
01:23 of the year for you, Dee, but we have a great conversation.
01:27 I want you all to stick around when Dee and I are done
01:30 with my on-the-tape co-host, Danny Moses,
01:33 and we are joined by Michael Dempsey.
01:34 He is the managing partner at Compound.
01:37 That is a VC firm, and we're gonna talk about
01:38 how they are thinking about investing in AI,
01:42 machine learning, robotics, digital health,
01:45 and believe it or not, crypto,
01:46 and these are areas that they've been involved with
01:49 for a while.
01:50 They are not Johnny come lately,
01:51 then Michael is a brilliant guy,
01:52 so we had a great conversation, so stick around for that.
01:55 Dee, by the time that people are listening to this,
01:58 okay, we have already had earnings
02:00 from Alphabet and from Microsoft.
02:02 This is a $2.6 trillion market cap company
02:05 that the implied move in the options market
02:07 is about 5% in either direction,
02:09 and also Alphabet, a $1.5 trillion market cap company,
02:13 also implied move about 5%.
02:14 Those are some big moves for some big companies.
02:17 Again, you guys are all gonna know what they reported,
02:20 what they guided to, and what the reaction is,
02:22 and I had a great conversation with Gene Munster,
02:25 who is of Deepwater Asset Management,
02:28 where we looked at all of those earnings,
02:29 and then also Meta and Apple and Amazon,
02:32 which report next week, so you can find that
02:34 in the On the Tape feed.
02:35 All right, Dee, let's get to it,
02:36 because you've been doing some really interesting reporting.
02:39 I know that the day that the Lyft folks
02:43 announced a new CEO, you had a conversation with them.
02:46 I feel like it was a few months ago,
02:47 and then on Tech Check this week, you had another one.
02:49 What were some of the takeaways?
02:51 Because again, this is kind of interesting,
02:53 tying back to what you're seeing in San Francisco,
02:55 maybe some signs in life of some hubs like that
02:58 in and around AI.
03:00 What was your take?
03:00 I know you cover Uber very closely too,
03:03 but thoughts there, 'cause I know that in your reporting,
03:05 you spent some time on just their singular focus
03:07 on North American ride share.
03:09 Are we seeing some signs of life as it relates to Lyft,
03:12 and some of the things that we've seen
03:14 as far as Uber and their core business?
03:16 - Yeah, this was really the year that,
03:18 actually, let me correct that.
03:19 The last few years, we've seen Uber pull away from Lyft.
03:22 They used to sort of trade in tandem, as you well know,
03:26 sort of the major players in the ride sharing space,
03:28 but Lyft had just really fallen behind over the pandemic.
03:31 So a few months ago, they replaced their co-founders
03:34 with a new CEO, this guy, David Rescher,
03:36 who worked many years ago at Amazon and Microsoft,
03:39 but he'd been in the nonprofit world.
03:40 So we sat down with him 100 days in
03:43 to sort of look at the state of Lyft
03:45 and ride sharing as a whole.
03:46 And he has closed the gap a little bit.
03:50 They've won a few points back of market share
03:52 by competing more on price,
03:53 but that is a dangerous game
03:55 for ride sharing companies to do
03:56 because those margins are already so thin,
03:59 and these companies are still bleeding
04:01 hundreds of millions, in some cases, billions of dollars.
04:04 So he said, "Listen, we're competing on price."
04:07 And I asked him, "How far are you gonna go?"
04:08 I think that's what Wall Street wants to know.
04:10 Are we gonna get to another price wars
04:12 where these companies have been working on profitability,
04:14 but do they suddenly turn that back
04:16 to compete for market again?
04:18 And he said, "No, we're gonna compete on other ways."
04:21 And I thought the most interesting thing he said was,
04:24 "It's not terrible being number two."
04:26 And he pointed to Pepsi, he said,
04:27 "You know, it's a pretty good business, so we're okay."
04:30 And I thought that that was setting expectations quite low,
04:34 but also really indicative
04:36 of where the gig economy ride sharing has come.
04:40 And you know, just what a journey it's been.
04:42 We've been covering these companies.
04:43 You have been looking at them for so long too, Dan.
04:45 When the competition was so fierce,
04:47 and now it's just sort of, they need each other
04:49 because they're duopoly, and in some ways, an oligopoly,
04:53 because they set prices in a way,
04:56 and you have the CEO of Lyft saying,
04:58 "We're not gonna go too hard on them."
05:00 - Yeah, so it's interesting.
05:01 We're gonna get our earnings from Lyft,
05:03 I think August 8th, and then Uber,
05:05 I think maybe next week, August 1st.
05:08 And this is a tale of two cities here, man.
05:10 I mean, like look at the market cap on a Lyft.
05:13 It's $4.4 billion.
05:16 I mean, they do have a bit of cash.
05:17 The stock's only up 5% on the year.
05:19 It's had a big run off of its recent lows,
05:21 but this one on Uber,
05:23 and this has kind of snuck up on us a little bit.
05:25 This in the last three months,
05:27 this stock is up 60%,
05:29 nearing $100 billion market cap again.
05:32 They guided to gap profitability this year, marginally,
05:35 I mean, by maybe a couple pennies or so,
05:38 but sales are still growing,
05:40 and high teens sort of rates,
05:43 and this is a company where margins are getting better.
05:46 I think from the mid to high 30s,
05:47 expected to be above 40% this year,
05:50 maybe above 42% next year.
05:52 Why have investors, do you think,
05:54 just come around to this story?
05:56 Because it's just taken off in the last few months.
05:58 That's the Uber.
06:00 - Well, you could maybe attribute the demise of Lyft,
06:03 and it's shrinking market share to Uber's gain.
06:07 And I think it's so interesting now
06:08 that Uber doesn't want Lyft to go away by any means.
06:11 They need them because as they approach $100 billion,
06:14 remember that they have all of these
06:16 regulatory battles as well.
06:17 And you don't want to face those
06:19 being accused of a monopoly, right?
06:22 They need Lyft to show that there's actually two players
06:25 that are still somewhat competitive in this market.
06:27 But the regulatory stuff always seems to creep up
06:30 and surprise investors in a way.
06:33 Even when you have a positive regulatory headline,
06:35 the stock will sell off because it's just a reminder
06:38 that this business is somewhat fragile.
06:40 It rests on all of their drivers
06:42 being part of the gig economy, independent contractors.
06:45 Remember a few years ago when California
06:48 was looking at changing that status,
06:49 and Uber and Lyft said, "Listen,
06:50 "we're just gonna have to shut down completely
06:52 "because the business is not sustainable
06:54 "if we have to pay our drivers like employees."
06:56 So there's always that.
06:57 How good is it as a business?
06:59 It's on some shaky ground,
07:00 but I think that Uber, Dara Khosrow Shahi
07:04 has certainly made strides in convincing investors
07:07 that he knows what he's doing.
07:08 He's gonna diversify into food delivery
07:10 and he's gonna go towards a better measure of profitability.
07:14 - Yeah, so another name that kind of snuck up on me
07:17 and had a disappointing quarter in guidance
07:19 and the stock sold off, I think like 10%
07:21 when it reported a couple months ago,
07:23 two, three months ago, was Airbnb.
07:25 And I know that's another name
07:26 that you track pretty closely.
07:27 Again, also up 35% in the last two months or so
07:31 nearing $100 billion market caps.
07:33 It just seems like the mood has changed.
07:36 And these are two companies,
07:38 Uber and Airbnb, they use machine learning.
07:41 This has been a big part of the engines that they've built
07:45 over the last, call it 10 years or so,
07:47 but you're not hearing talk,
07:49 "I'm sure Airbnb is gonna be integrating a chat bot."
07:52 You know what I mean?
07:53 "One of these things and Uber
07:54 "is gonna be using all these stuff."
07:55 So it's funny, we haven't heard these names
07:58 get into the AI hype, if you will.
08:02 Again, you and I have been talking about this for months now.
08:04 All of these companies,
08:05 this is what they spend their R&D on.
08:07 This is what they kind of pile back.
08:10 When they weren't making money,
08:11 this is the sort of investments they're meant to be making,
08:14 but we haven't heard it yet.
08:15 And I feel like we will start to hear this more and more
08:18 from some of these companies
08:19 because they're gonna be asked the question,
08:20 how they're thinking about these technologies
08:22 and how will they use them.
08:24 - Well, for Uber, I think they wanna be careful
08:26 talking about artificial intelligence
08:28 'cause on the streets here in San Francisco,
08:30 you see fully autonomous cars driving around.
08:35 You don't even need a driver.
08:36 And this was something that Uber had bet on
08:38 in the Travis Kalanick era,
08:39 but gone away from in the Dara Khosrow Shahi Uber 2.0 era.
08:43 So that could disrupt its entire model
08:46 if all of a sudden you don't need a driver.
08:47 I mean, we're still far from that, but that's really,
08:50 I mean, when you talk about artificial intelligence,
08:53 you're really talking about autonomous vehicles.
08:55 And I think that Uber has gone away from it.
08:57 So what, they're gonna slap a chat bot on something?
08:59 I'm not sure that that's gonna cut it for investors.
09:02 I'm trying to think of other ways.
09:03 I mean, yeah, to make the algorithm better,
09:05 to make pickup times faster.
09:07 Sure, that's around the edges
09:08 and they've already been doing that,
09:09 but can they partake in this generative AI cycle
09:12 in a meaningful way?
09:14 I'm not sure, but I think you're right.
09:16 They'll probably be asked about it.
09:17 - Yeah, and I think the one thing that as I talk to
09:19 like investors, VCs and the like,
09:21 and people in the tech space in general,
09:24 I think one of the things that people remain
09:26 really optimistic about is we don't even know
09:29 the sorts of things that it's gonna do
09:30 and the changes that it's gonna make.
09:32 And again, I'm not a technologist by any means.
09:34 I definitely consider myself a bit of like early adopter.
09:38 So I always find this stuff pretty fascinating,
09:40 but I do seem to be,
09:42 and I think you can probably tell this over the last few
09:43 months, a bit cynical about some of these large
09:46 platform companies, just like as far as at least
09:49 what investors are allowing for them in the near term
09:52 to accrue whatever value about some of the early
09:55 positioning that they have, you know, that to me,
09:58 I'm probably thinking about fading a little bit
10:01 and then thinking about how some of these other companies
10:03 can get a lot of leverage, you know,
10:05 out of these technologies.
10:06 Once they're a bit more developed and we see better
10:08 commercial applications for them, I mean,
10:10 I guess that's the thing that kind of might excite me.
10:14 All right, we got to talk about this other big story here.
10:16 This is Elon Musk.
10:17 He's finally just basically taking a flamethrower
10:20 to anything on Twitter.
10:22 And I thought this was a really, you know,
10:23 by renaming it X and this happened, I guess,
10:25 over the weekend, he tweeted this out
10:28 or Z did it out or I don't even know
10:30 what they're calling it now.
10:31 And I thought this was a really interesting take
10:33 from Matt Levine over at Bloomberg and he writes this great,
10:36 you know, daily piece.
10:38 And so he says, I guess my question is,
10:41 what is he paying for?
10:43 Musk didn't want Twitter for its employees.
10:45 He fired most of them or its code,
10:47 which he trashes regularly or its brand,
10:49 which he has abandoned or its dedicated users
10:52 whom he is working to drive away.
10:54 He just wanted an entirely different Twitter-like service.
10:57 Surely he could have built that for less than $44 billion.
11:02 Mark Zuckerberg did.
11:03 And I think it's really interesting.
11:04 And so, you know, this story so well,
11:07 you knew tons of people who've worked there,
11:09 who've invested in the company,
11:11 who've been counter parties and in around it,
11:13 who've been competing with them.
11:14 You're out there in the middle of this thing.
11:17 I mean, what is the answer for this?
11:19 Because again, you know, you always hear that
11:21 build versus buy sort of thing.
11:23 It's probably keep broke one of the,
11:25 like the most cardinal rules that exist in Silicon Valley.
11:29 And he seems to be dismantling it brick by brick.
11:32 - Yeah, but you know,
11:33 he's almost like a God here in Silicon Valley.
11:36 You know, he created Tesla, he created SpaceX.
11:39 There is sort of this reluctancy to say that Elon Musk
11:43 doesn't know what he's doing because people have said that
11:46 and live to regret it in the past.
11:48 But I have the same questions.
11:51 I have no idea what he's doing.
11:53 I know that he's trying to build a super app
11:55 and he was early to that, right?
11:56 He wanted to do that before a lot of the Asian companies
11:58 actually did.
11:59 He's envisioned x.com becoming this sort of everything app
12:04 or website or whatever, everything platform for a long time.
12:07 But the thing is,
12:08 is that the internet has evolved very differently in the US
12:11 than it has in China.
12:13 And there's reasons that you can have a super app
12:14 like WeChat, like Grab, like Kakao in Korea.
12:19 In Asian, you can't have it here.
12:22 And I don't know what his answer to that is
12:25 and how he has critical mass.
12:28 Did you see though, last night they were removing,
12:31 or I think it was during the daytime,
12:32 they were removing the Twitter signage,
12:34 which is such an icon here in San Francisco,
12:37 in downtown San Francisco.
12:38 You pass and you're like,
12:39 "Oh, that's the Twitter headquarters."
12:41 Dismantling it, I think didn't get the proper permit.
12:44 So it was just ER that was left on the sign,
12:47 which is ironic.
12:49 - I thought you're reporting,
12:50 I saw you on Tech Check on CNBC yesterday,
12:53 and you were talking about how just,
12:55 the difference between the development
12:57 in some of these Asian countries
12:58 and some of the services that you talked about,
13:01 because they were building on a mobile native
13:03 sort of platform,
13:04 and we're still stuck with some of these legacy financial app
13:08 that weren't built for mobile, if you will.
13:11 And so the integration of mobile and social and e-commerce,
13:15 and then payments is just gonna be a bit slower here.
13:18 There's like massive incumbents
13:19 and they're trying to basically,
13:22 kind of defend their moats for all intents and purposes.
13:24 But the one thing I would say,
13:25 we were on Fast Money last night,
13:27 we were talking about this,
13:28 and I gotta tell you, this new CEO, Linda Iaccarino,
13:30 and I'm not expecting you to comment on this,
13:32 I know she was probably a former colleague of yours
13:34 when she was at NBC Universal.
13:35 I mean, the stuff that she tweets,
13:37 and I think Casey Newton over at Platformer,
13:39 which I love his newsletter,
13:42 I mean, he basically called it just a word salad.
13:45 It meant nothing.
13:46 She's just out there.
13:47 I feel so bad for this woman, to be honest with you.
13:49 I hope she's getting paid an awful lot of money
13:51 because he has just put her out there as a human shield,
13:54 if you will,
13:55 just to deflect all the garbage about what's going on.
13:58 And I think at the end of the day,
13:59 only he knows what's going on.
14:01 But what's really interesting to me
14:03 is when you think about the super app,
14:05 and this goes back to,
14:06 and I mentioned this to Gene Munster
14:07 in my conversation with him on Monday,
14:09 is that, you know,
14:11 think about Mark Zuckerberg and Facebook's early attempts
14:14 at building hardware, building a mobile operating system,
14:17 and they had to abandon it, right?
14:19 Because the iPhone was just a juggernaut.
14:21 And then if you were on Android, you know,
14:23 there were some applications that worked a bit better,
14:26 definitely outside the US, but think about this, you know,
14:29 they have three and a half billion monthly active users
14:32 across the Facebook platform.
14:34 They have WhatsApp that's got over 2 billion.
14:36 They got Messenger that has over one and a half billion.
14:39 They have Instagram that's got over two.
14:41 They have Reels that's got over one.
14:43 I mean, think about this, right?
14:44 And now Threads was the quickest app to 100 million.
14:48 And again, and Twitter has 330 million,
14:51 and they've never been able to monetize that properly,
14:53 right?
14:54 So like to me, Facebook's not far away
14:57 from putting it all together,
14:59 and we know what they wanna do with payments
15:01 in the digital space.
15:02 We know what they wanna do with the metaverse.
15:05 So if there's ever a company, and I'm not, listen,
15:07 I'm not a Zuckerberg fan.
15:09 I'm not a Facebook fan by any means.
15:10 I think it's funny how Elon Musk
15:13 has made Mark Zuckerberg look good
15:15 in the Valley or all over the place,
15:17 but they're really close to a super app.
15:19 - Then why didn't they just put Threads
15:22 as a button inside of Instagram?
15:24 It's almost like they gave up before they even started.
15:26 But Zuckerberg has been talking about
15:28 trying to make a super app for a very, very long time.
15:30 And you could argue that WhatsApp itself
15:32 is maybe the closest thing to it.
15:34 But again, it's just so far from the super app
15:38 as it exists in Asia.
15:40 And again, I don't know that you can put it all together
15:42 because people, like I don't use Facebook anymore.
15:45 People don't wanna open it up,
15:46 but maybe it could just be a tab.
15:48 And that's how WeChat works, right?
15:49 There's no, the whole point of a super app,
15:51 there's no reason to ever leave.
15:53 And I do because I lived in China and used WeChat
15:56 when it was becoming very popular.
15:58 I always am kind of amazed
16:00 at how they want me to switch apps.
16:02 But I think, oh, this was just so much easier.
16:04 Wouldn't it be easier?
16:05 Couldn't you just have all those billions of people
16:07 that you just mentioned in one app
16:09 and have your algorithm feed across it
16:12 and better target them for advertising
16:15 or cross sell or whatever it is, right?
16:17 But it may be the closest, but it's still so far off.
16:20 - I mean, listen, you know what?
16:22 I think that they have basically 40 some percent
16:25 of the planet's population
16:27 and over 50% of those who are connected,
16:30 you know what I mean, using their app.
16:32 You know what I mean?
16:32 - They are amazing numbers.
16:34 - Yeah, and to me, and then we're gonna get a sense
16:37 from Apple next week, if you think about how most
16:39 of the people, at least in North America and Europe,
16:42 access a lot of their apps is through an iOS device, right?
16:46 So you think--
16:47 - iOS maybe is the super app all of us aren't talking about.
16:50 - Yeah, exactly.
16:51 Well, there you go.
16:51 All right, well, listen, we're gonna keep watching this one
16:54 as it unfolds because I think there's very few people
16:56 who have a really good sense of what his plan is.
16:59 But if you think of the absolute destruction
17:02 that's been created from a value standpoint
17:04 and you think about this,
17:05 and I think I said it to you last week,
17:06 I keep hearing people speculate
17:08 that this thing is gonna be in bankruptcy pretty soon,
17:11 which is just amazing if he were to hang the banks
17:14 on all that debt, 13 billion,
17:15 and then all the equity people who came in on that, right?
17:19 Like they rolled their equity at 5420 into this deal.
17:23 I mean, I just, to me, it's purely, it's very astounding
17:27 for that sort of value destruction
17:29 in such a short period of time,
17:31 where it wasn't really in the market's hands,
17:32 it was in his hands.
17:33 That's why he took it private, you know what I mean?
17:35 Because you weren't gonna be subject to market forces
17:38 and it seems like almost everything that he's done
17:40 from his behavior on the platform,
17:42 driving away advertisers, driving away users,
17:45 it just seems really odd to me.
17:47 - Well, I mean, one thing you could say about that
17:49 is at least he's destroying the rich people's value, right?
17:53 He took it private and he got all of these rich individuals
17:57 and venture capitalists to invest at 5420.
18:00 He's not taking it down with the retail investors.
18:03 - That is a fact.
18:04 All right, Deirdre Bosa,
18:06 she is the host of CNBC's Tech Check.
18:08 I really appreciate you joining me again.
18:10 We'll see you next week.
18:12 Stick around for my conversation
18:14 with Michael Dempsey and Danny Moses.
18:16 - Hey, Dan. - What up, guy?
18:18 - You're into this FinTech.
18:19 What's all this I'm hearing about Current?
18:21 - You're gonna like this guy.
18:22 Current is a FinTech company
18:23 that's completely disrupting traditional banking.
18:26 - Wait a second, does that mean
18:27 I don't have to drive to the bank anymore?
18:30 - Yeah, exactly.
18:31 I'm a new Current customer
18:32 and I manage all of my finances from one easy to use app.
18:35 - Well, I gotta get this app, but where can I learn more?
18:38 - It's super easy.
18:39 You just go to current.com/okay, O-K-A-Y,
18:42 and download the app.
18:43 That's current.com/okay.
18:46 Current is a financial technology company, not a bank.
18:48 Banking services provided by and Visa debit card
18:51 issued by Choice Financial Group, member FDIC,
18:54 pursuant to a license from Visa USA Inc.,
18:57 and can be used everywhere Visa debit cards are accepted.
18:59 Hey, it's Dan here.
19:00 Are you tired of struggling with your weight?
19:02 I know I was, and that's why I turned to Roe.
19:05 They offer a revolutionary medication approved by the FDA
19:07 specifically for weight loss.
19:09 It's paired with lifestyle coaching,
19:11 so the weight comes off and stays off.
19:13 I started with a free online visit
19:15 and connected with a US licensed healthcare professional.
19:17 They prescribed the medication that helped me drop 30 pounds
19:20 in the last four months.
19:22 And now Roe is offering a discount just for our listeners.
19:24 To get our special offer, go to roe.co/okay.
19:28 You'll pay just $99 for the first month
19:30 and $145 per month thereafter.
19:33 If prescribed, medication cost is separate.
19:35 That's roe.co/okayay.
19:39 All right, welcome back to "Okay Computer."
19:41 I am here with two very special guests.
19:43 One that many "Okay Computer" listeners might know,
19:47 that is my friend, Danny Moses.
19:48 He is the co-host of the "On the Tape" podcast
19:51 that Guy, Adami, and I do.
19:52 We've been doing it together, Danny, since January of 2021.
19:56 This is your first appearance.
19:57 - On "Okay Computer."
19:58 Everyone should be very nervous that I'm on "Okay Computer."
20:01 Well, I'm dipping into an area
20:02 where I really have very little.
20:04 - So last night, you were our co-host
20:06 on CNBC's "Fast Money for the Full Hour,"
20:08 which was really fun with Guy and me.
20:09 And now here we are on "Okay Computer."
20:11 But we have actually somebody
20:12 that we've wanted to have on for a long time.
20:14 Michael Dempsey, he is a managing partner at Compound,
20:18 a VC firm here in New York City.
20:21 You're an investor in Compound.
20:23 - I am.
20:24 - Your good friend started that fund.
20:25 - David Hirsch.
20:26 - David Hirsch, who I know, who's also a brilliant guy.
20:29 Talk to us a little bit about why Michael,
20:32 Michael, welcome, first of all, to "Okay Computer."
20:34 Thanks for joining us here in our studios.
20:35 And talk to us a little bit about your experience
20:37 with Compound and what we're gonna be talking about
20:39 with Michael today.
20:40 - Sure, so years ago, I wanna say five years ago,
20:43 I invested in Compound One, I wanna say roughly.
20:46 I met Michael then.
20:48 You think he's young now.
20:50 You should have mentioned five years ago.
20:51 And I was always blown away by not just how smart you were,
20:56 but how cynical and seasoned you already were at that age.
20:58 And when I met you, you were very involved in many areas,
21:02 crypto being one of them,
21:03 which I was very interested in learning more.
21:04 And I think the way that you approached crypto
21:06 was very smart.
21:07 You were always skeptical,
21:09 but smart on how to approach it.
21:10 And I know that's in your fund to some degree
21:11 in Compound One, you have some crypto investments,
21:13 we can talk about it.
21:14 But more importantly, to me,
21:16 the thing that Compound does better than any other firm,
21:19 if you wanna call it VC,
21:20 it's almost private equity at this point
21:21 because they have so many companies that have matured beyond,
21:24 is that you want it to stay small.
21:26 And your funds, if I'm not mistaken,
21:28 were 50 million each,
21:30 three funds that were part of the Compound One unit,
21:32 but roughly. - Around that, yeah.
21:32 - Roughly.
21:33 And that gives you the opportunity
21:34 to have meaningful positions in smaller companies
21:36 when they're starting up.
21:37 So let's start with that right now,
21:39 kind of the process, how you guys think about investing
21:41 and being able to spread out,
21:42 knowing that some of these investments
21:44 aren't going to succeed.
21:45 And also knowing, give yourself the firepower
21:47 and the patience, which you guys have, by the way,
21:49 to do follow on offerings and to be patient
21:51 and find the right areas.
21:52 And now I think it's paying off for you guys
21:53 in terms of having waited a little bit.
21:55 So let's start there,
21:56 'cause I know you guys describe yourself
21:57 as a research centric something thing,
21:59 you guys valued, whatever the thing is.
22:01 Talk about that and talk about how you got to Compound.
22:04 - Yeah, so yeah, we describe ourselves
22:06 as a research centric thesis driven investment firm.
22:08 What that means to us is doing the thing
22:10 that everyone experienced in venture tells you not to do,
22:12 which is we build very prescriptive views of the worlds
22:15 and like the futures we believe in.
22:16 And then we find founders to partner with
22:19 that fit into those theses or shatter them.
22:21 And usually the thing that we are pretty good at
22:25 is saying what do the next 10 to 15 years
22:28 of technological change look like?
22:30 What are areas that have very high asymmetric upside
22:32 and what is like value accrual going to look like
22:34 within those areas?
22:35 And so for us, that usually falls
22:38 within a few core categories of robotics,
22:40 machine learning, healthcare and bio and crypto.
22:43 And then there's probably 20% of adjacencies
22:45 that look earlier on the commercialization curve,
22:48 carry significant science risk
22:49 and those might look like specialized chips,
22:51 those might look like things like AR and VR,
22:53 those might look like quantum computing, stuff like that.
22:55 But typically it's science or engineering risk early on.
22:59 And then we get there early and learn
23:02 and back a bunch of companies and partner with them
23:04 in a fairly concentrated way, join the boards,
23:06 work with them over the next decade
23:08 and exploit kind of the learnings we have through cycles.
23:12 And I think the thing that we are learning
23:14 as the firm evolves is understanding how to study things
23:17 when they're coming apart to kind of continue
23:19 to extract value.
23:20 Crypto is a great example of that.
23:21 Funny enough, AI would have been an example of that
23:23 a few years ago if you had asked us
23:25 and now obviously is a new thing.
23:27 And my background, I kind of was always interested
23:30 in investing, started off my first job at a hedge fund
23:33 doing long, short and volatility trading around derivatives
23:36 as well as some cross-border private equity
23:38 and eventually let it carve out to do early stage investing.
23:41 Fell in love with that and knew that's what I wanted to do
23:43 but should understand how startups work.
23:45 So joined a company called CB Insights really early on,
23:47 helped build out a lot of the data and research team.
23:50 And in that time really found these areas
23:54 that were really kind of like technical and unique
23:55 and said, okay, here's areas that I think
23:58 are hard to pattern match to and have built my career
24:00 around investing in very technical early categories.
24:03 - I would say you guys are a very shareholder-friendly firm
24:06 meaning you guys have returned capital,
24:08 you've taken the opportunity to monetize some of the stuff.
24:12 You've also set up SPVs, so to speak, on one-off
24:15 follow-on investments outside of it,
24:17 which is a great thing that investors
24:19 get to kind of partake in.
24:20 So let's dive in kind of to where we sit today,
24:23 kind of greatest hits in the portfolio.
24:25 And I know that Dan's gonna wanna go here
24:27 and on this podcast they talk about often is,
24:30 whether it's electric vehicles,
24:32 artificial intelligence and crypto,
24:33 these are themes that come up nonstop
24:35 and you guys have been early and right in various names.
24:37 So we can start any way you want,
24:39 but maybe just kind of the highlight reel right now
24:41 of what you think the most exciting stuff
24:43 is in your portfolio.
24:44 - Yeah, I think it's across a bunch of different areas.
24:47 I would say like in the AI space,
24:48 we have a few investments we made quite early.
24:51 Runway ML here in New York is probably an example of that
24:53 of a company that has done really well.
24:55 We backed them in 2017 as their first investor.
24:58 Our view at the time was AI was gonna revolutionize
25:02 kind of creative industries.
25:03 Everyone thought that would be the last thing
25:05 AI was gonna hit.
25:06 Turns out it might actually be the first thing.
25:08 And they have since grown to build a model called Gen2,
25:11 which is a video generation model.
25:13 And you can think of them simplistically as open AI,
25:16 but for video instead of text.
25:17 And they build a bunch of amazing products around it.
25:20 Wave is another example.
25:21 - Whoa, whoa, whoa, hold on.
25:22 Let's not skip over what just happened.
25:24 Someone just took an investment in this thing, in Runway.
25:28 All right, so that's out in the news.
25:29 You can confirm.
25:30 So who was it?
25:31 'Cause I think people wanna understand this area,
25:34 what it means, what is the market cap?
25:35 We are seeing crazy market caps start to happen, right?
25:37 And 2 trillion, 3 trillion.
25:39 Here you are actually witnessing an application
25:41 that's actually already in motion
25:42 and who invested in this thing.
25:44 - Yeah, so Google invested--
25:46 - Oh, Google, oh, sorry.
25:47 Yeah, Google, yeah.
25:48 - In a recent round.
25:50 And a lot of that enables the company
25:52 to continue to spend money on compute, scaling up resources,
25:54 and also really continuing to push forward
25:57 a bunch of different enterprise use cases.
25:58 - Hold on, I'm sorry, I gotta stop you again.
26:00 What was the valuation that Google invested in?
26:04 That's public, isn't it?
26:05 - The valuation has been reported,
26:06 but I can't confirm any valuation.
26:07 - I saw $1.5 billion, something like that.
26:10 I just wanna take a step back there.
26:11 Were you invested in it?
26:12 - Yeah.
26:13 - Roughly, what was the valuation then?
26:15 I'm trying to get to a point of not a mania,
26:17 where we sit, but just in general.
26:19 - Yeah, it was under $10 million.
26:21 - Congratulations by that on the way.
26:22 All right, next one.
26:23 And I know I wanna get to the women's healthcare platform,
26:25 which I know is incredible,
26:27 which Melinda Gates is an investor in.
26:28 But I jump ahead, Dan.
26:29 I'm sorry that I've come onto OK Computer and I have--
26:33 - This is why we're doing it here.
26:34 And we really actually wanna expand away
26:37 from the runway thing too.
26:38 But it seems like there's other verticals
26:40 that you guys were early on.
26:42 And they're starting to really come to fruition here.
26:45 And I know that your life cycle is usually, what,
26:47 five to 10 years on some of these sorts of things.
26:49 But to have an investment in something that,
26:51 were you invested, I'm assuming at pre-seed stage,
26:54 at 10 million, and to have a company like Alphabet
26:58 confirm that thesis four or five years later.
27:00 - See, that's why I'm not on OK Computer.
27:02 I still call it Google.
27:03 See, this is my problem.
27:04 - No, I call it Google too.
27:05 - Yeah, we all do.
27:06 And then the other thing I definitely wanna talk about
27:08 in a second also is that, look at what Nvidia,
27:10 look at how they're investing, look at all these things.
27:12 It's pretty fascinating that, I wonder if smaller VCs
27:16 are just getting crowded out a little bit now
27:18 because all these massive incumbents
27:19 were actually benefiting in the public markets.
27:22 Now, by all this stuff, it behooves them
27:24 to make these sorts of investments,
27:26 sign them up to cloud contracts,
27:29 and using their tools and that sort of thing.
27:31 But give us a couple other examples in the portfolio
27:33 of things that you saw a few years ago
27:35 and now are being appreciated in the markets.
27:37 - Yeah, I'd say Wave is probably the other big example.
27:40 Wave is a self-driving car software platform.
27:43 - W-A-Y-V-E, right? - Y-V-E.
27:45 They're building what I would call
27:46 a foundation model for robotics,
27:48 but the kind of easily understand thing
27:51 is self-driving car software.
27:52 And they started at a time in which
27:54 post-GM acquisition of Cruise, post-Waymo scale-up,
27:58 post-Zooks, all these companies, everyone said,
28:01 "Self-driving's done."
28:02 And our view was machine learning was gonna play
28:05 a much, or artificial intelligence now,
28:06 was gonna play a much larger role within self-driving,
28:08 and you would actually need to build a model
28:10 that truly understands how to drive
28:12 versus teaching these hand-coded rules
28:14 to not get too technical.
28:16 Team out of London, we backed them.
28:18 I met the founder when he was getting his PhD.
28:19 I read his thesis.
28:21 Few years later, he started a company.
28:22 We led their seed round, been on the board ever since.
28:24 And that company has trials going in London,
28:28 has cars on the road there.
28:30 And our view is they have one of the best,
28:33 if not the best shots, of building scalable autonomy.
28:36 So being able to go city to city quite quickly
28:38 instead of spending months gathering data,
28:40 spending tens to hundreds of millions of dollars.
28:42 And they take a similar approach to Tesla
28:44 in the sense that they don't believe in LIDAR.
28:46 They believe the same thing that we believe,
28:47 which is humans can drive with two cameras, our eyes,
28:50 and they believe they can get a car
28:52 to drive itself with two cameras.
28:53 - So how would you compare their technology to Tesla?
28:56 Because as far as being ready,
28:58 whatever level autonomy we're talking about here.
29:00 - I think the Tesla is very performant on highways,
29:04 as we all know.
29:05 It's a great lane follow and things like that.
29:07 Wave is very centered on urban driving.
29:09 So their cars drive around London,
29:11 which is one of the most chaotic environments ever.
29:13 And from an urban driving perspective,
29:15 I think Wave has one of,
29:17 if not the best technology in the world.
29:19 - So is it commercially appliable at this point?
29:21 Are they still, like, where are we in the stages?
29:23 - They're in commercial trials now on grocery delivery.
29:26 - Because if it was Musk,
29:27 he'd already have it out on the road, right?
29:29 No. - Maybe, maybe.
29:30 - Not maybe, 'cause he has.
29:32 So his product's inferior,
29:33 and he has it out in beta testing all over the--
29:35 - Very different products, but yeah, I think it's--
29:37 - Why are they so different?
29:38 You said because one's built for--
29:39 - Urban, yeah.
29:41 - But Tesla owners don't know that.
29:43 They're probably driving in cities.
29:45 - I think some of them are, yeah.
29:47 I think there's varying degrees of comfort
29:50 that people have in putting things in the world.
29:52 And to be democratic about it,
29:55 I think that each of these founders
29:56 have different decisions they make on risk.
29:57 And I think we're seeing that across everything in AI.
30:00 We see it with Zuckerberg putting out LLAMA
30:02 as an open source model and giving it to everyone,
30:04 where a few years ago,
30:05 OpenAI was worried GPT-2 might be AGI
30:08 and threaten the world.
30:09 - So is there a level for this also?
30:11 Is it considered, is there a standard level for--
30:14 - The goal is to build true L5 self-driving,
30:17 but I think we'll see these things progress in a gradient
30:21 more than we'll see kind of like one day you flip a switch
30:23 and it's--
30:24 - Is Bill Gates an investor or is Microsoft an investor?
30:26 - Microsoft is an investor.
30:27 - Okay, got it.
30:28 - I think what Danny's also trying to get at
30:30 is like Elon Musk has been talking about L4, L5,
30:33 you know what I mean?
30:34 Like we should be there right now,
30:36 if you were listening to what he was saying
30:37 five or seven years ago.
30:39 And so I wonder, it's interesting,
30:42 how many standards do you think there will be?
30:43 And I think what just happened with the charging network,
30:46 it could be a little bit of a precursor to what we see,
30:49 but it has totally like other massive implications,
30:53 regulatory and like all that sort of stuff.
30:55 So it's interesting because the EU
30:57 is gonna treat this very differently than we are here, right?
31:01 And so there might develop more standards
31:03 than we would probably want as a technologist.
31:07 So talk to us a little bit about that
31:09 because I think that, you know,
31:11 it seems like your guys are a bit more sober
31:13 about when they're gonna get to L4 or L5.
31:18 So just give us a sense for that
31:19 because that's something that a lot of the Tesla bulls,
31:21 and I don't know if you know this,
31:22 but we're Tesla bearers on the podcast.
31:24 - Michael Lewis.
31:25 - Tesla bull, you know, we would tell you
31:27 that there's hundreds of billions of dollars
31:29 embedded into the current market capitalization
31:32 of value of Tesla based on all of these promise
31:35 of autonomous driving.
31:37 - Yeah, I think that I believe in two things at once,
31:40 which is one, I think technology moves faster
31:43 than governments can kind of shoehorn it or control it.
31:47 And so in some ways, these things that are inevitable,
31:50 especially when it's like driving,
31:51 where it's one of the largest killers in the entire world,
31:54 will move at a pace that forces the hand of governments
31:58 and will likely be deployed in ways that,
32:01 while there will need to be sign off,
32:04 the burden of data will go down over time
32:06 as people can feel the experience of being in a car
32:08 that truly can reason its way through urban environments.
32:11 That said, there is meaningful amounts of data
32:13 that needs to be gathered when you go into a new city,
32:15 a new environment, whatever.
32:17 And I think that all of the work
32:18 that all these companies are doing
32:19 from a lobbying perspective and a regulatory perspective,
32:22 I think will work fine on any time horizon
32:25 for full L4, L5 autonomy.
32:27 Like it is not something I worry about literally at all.
32:30 And I think a lot of people look at the shift
32:34 of how scary it could be to have self-driving cars
32:36 on the road.
32:37 And I think it's, once you realize
32:39 just how terrible human performance is as driving,
32:42 and this is something we see in basically every industry
32:44 we look at with AI versus healthcare,
32:46 AI versus driving, AI versus whatever.
32:47 Like humans are generally bad at most things
32:50 relative to AI.
32:52 I think it's gonna just get past.
32:54 - So Wave is a software developer, right?
32:56 They don't make cars.
32:57 So could Wave's product be on any type of car?
33:02 Could it be put-
33:02 - Well, a lot of the thesis of why we love what they do
33:05 is if you look at the CapEx for like a Waymo car,
33:08 they have hundreds of thousands of dollars of sensors.
33:11 And Wave is a camera array,
33:14 and then one or two cheaper sensors,
33:17 depending on how they kind of continue
33:19 to progress with that.
33:20 And so the bomb cost is like many, many orders
33:22 of magnitude cheaper.
33:23 And the thinking is there will be in this space
33:27 similar to how Mobileye kind of held everyone captive
33:29 for a while, and then eventually Tesla migrated away,
33:31 there will be other companies that develop autonomy
33:33 that get with OEMs or with other types of service providers.
33:37 - Is the idea for that technology
33:39 that it shouldn't matter if it's a Toyota Civic
33:42 at a $27,000 price point or a Mercedes S-Class at 85,000?
33:47 Like that autonomous technology
33:52 will be the same in both cars?
33:54 Or that's-
33:55 - I think generally, yes.
33:57 There will be stages of this
34:01 because of compute required or sensor
34:03 and how you generally think about building
34:05 from an OEM perspective,
34:06 a car with a certain amount of margin.
34:08 Obviously that cost will come down.
34:10 But the eventual future is,
34:12 I cannot imagine a world in which because you pay more,
34:14 you get a better autonomous driver.
34:16 I think there will be other adjacent things,
34:18 perhaps around driver behavior
34:19 or perhaps around the same things we compete now
34:23 for interiors of cars that are built around autonomy though.
34:26 But I don't think that it'll be like,
34:27 you pay more for a better driver.
34:29 I just, I think that won't fly.
34:31 - All right, so let's shift gears to healthcare.
34:33 Tia, which the really cool thing about Compound
34:36 is that they have an annual meeting
34:38 and they do it in New York City
34:39 and all their companies, their portfolio companies come in
34:42 and you can sit in the audience and it's Q&A,
34:43 but it's very intimate, like I said,
34:44 because it's not a huge fund.
34:46 It doesn't have hundreds of investors.
34:47 It's got a group.
34:48 And a lot of your investors are strategic in nature too.
34:51 They may bring you ideas and so,
34:53 which I think is always great to have both ways.
34:54 But this was one of the presentations
34:56 that I think was most exciting in the last few years
34:58 was Tia, what's the state there?
35:01 And I know that Melinda Gates is an investor in that one.
35:03 - Yeah, so Tia, I think one of the things that we do
35:06 when we're building a thesis is we say,
35:07 okay, what on the science side that it's like emerging
35:11 is interesting and what does that tell us
35:12 about eventual futures we should believe in?
35:14 And maybe seven or eight years ago,
35:17 it just became more and more clear
35:18 that like a few things were happening.
35:19 One, obviously healthcare in the US is broken,
35:21 specifically women's healthcare in the US
35:22 is incredibly broken.
35:24 And more and more science is coming out
35:25 around how you even treat women's healthcare.
35:28 Like, and I think we still struggle with this today
35:29 in a bunch of areas.
35:30 - Talk about the founder, sorry,
35:31 because she was so impressive.
35:32 I think that's key.
35:33 - So Tia, I wrote a post about this years ago
35:36 and I ended up getting connected to Carolyn Witt,
35:38 who's the CEO and Felicity Yost, who's the other co-founder.
35:41 And they were thinking about building kind of a new platform
35:45 to enable millennial women to find new types of healthcare.
35:48 And what they quickly realized early on,
35:50 and Carolyn's background, she was at Google
35:52 and helped build a bunch of the kind of early marketing
35:54 around a rebrand and a new launch for the search engine
35:57 years, a few years ago.
35:58 And one of the things that they kind of realized
36:01 early on in the company was that you can't actually impact
36:03 care without giving care.
36:06 And I think that's something that a lot of
36:07 digital health companies saw.
36:09 You don't impact like outcomes.
36:11 And so they have now built a full kind of healthcare clinic
36:13 with a very science driven,
36:15 what they call cycle connected care model.
36:18 And it's in this bigger trend of how do you think about
36:21 holistic care, but it's targeting a just
36:24 grossly underserved customer in women,
36:27 especially women in the United States where you often
36:29 don't build a relationship with your doctor until you go
36:31 to have a child or something.
36:32 And so there are a lot of kind of like doctor orphaned
36:34 women in this country.
36:36 Slave Scaled, they raised a hundred million dollars
36:39 a few years ago and now have multiple clinics
36:42 throughout the country,
36:43 partnerships with large hospital systems.
36:44 And our hope is they will kind of become an effective
36:48 one medical for women because we don't think that
36:51 it's a one size fits all care model for that class.
36:55 - What do you think are some of the common themes
36:57 as I think about these?
36:58 These are obviously very disruptive companies
37:00 that you're talking about taking on big incumbents, right?
37:02 That are at like the, you know,
37:03 the precipice of huge technological shifts.
37:06 And Tia, is this something that you see the way
37:10 in some of these other investments that you've described,
37:13 you know, technology playing a huge role in this,
37:15 or is it filling a gap in something that you,
37:18 you know what I mean?
37:18 Like you think that that has to happen and then technology
37:21 is really the next phase of that.
37:23 - I think in Tia's case, technology is the thing
37:27 that enables it to be a venture business
37:28 and not like a private equity business.
37:29 It enables better margin, better scale,
37:32 and that creates better outcomes
37:33 'cause you have better care as a customer.
37:36 And there's also a digital health component,
37:37 telehealth component, all these types of things.
37:39 I think like the main theme,
37:41 which is why we kind of always get so deep
37:44 on research centric, thesis driven is like,
37:46 if you look at what is happening in the earliest stages
37:49 of any industry, and that typically happens in academia
37:52 or research groups within large organizations,
37:54 you can just see enough nuggets that sit out there
37:58 that will tell you what is going to happen
37:59 over the next few decades.
38:00 And the example that now is helpful for anyone is,
38:03 and we wrote about this in our annual letter,
38:05 is like GPT-3 sat out in the open for two years
38:08 before chat GPT.
38:09 And like that paper was there, everyone saw it,
38:11 you could hit the API if you wanted to.
38:13 And it just took someone kind of like looking at it
38:15 and candidly a very small team at OpenAI being like,
38:17 we need to build a product around this thing
38:19 to show everyone what it could do.
38:21 And I think as more and more capital
38:23 has come into innovation at like by large,
38:27 the gap between academia and research and commercialization
38:31 and what we call like science projects
38:33 versus mass market venture scale businesses
38:35 is just collapsing in time faster and faster.
38:37 And I think the most important thing is that the people
38:40 in a lot of these areas are now seeing
38:42 that they can be founders.
38:43 It's not like this dark art,
38:44 they're seeing their friends do it.
38:45 And that makes them be like,
38:46 well, if this person can do it, I can do it.
38:48 And so you see more talent migrating
38:50 through these areas, which just enables people
38:53 who wanna do the work of really spending time reading
38:57 a lot of stuff and meeting with people
38:58 who maybe aren't ever gonna start companies
39:00 but have really bright ideas and research.
39:02 I think it just will be able to tell you a lot
39:04 about how the future unfolds.
39:05 - That's a good shift to crypto
39:08 because we talk about something that's been out there.
39:10 And when I tell people,
39:11 some of the smartest people in the world I know
39:13 believe in blockchain and crypto,
39:14 so therefore I'm just gonna put my brain in the closet
39:16 and let them go do it.
39:17 And you were early and often,
39:18 you're one of the people that I actually refer to.
39:20 So we don't really call it Bitcoin anymore.
39:21 We don't call it crypto.
39:22 It was just blockchain because are the applications
39:25 that we're seeing, this is why again,
39:26 I don't belong on this show,
39:28 applications, how do I think about the commercialization
39:31 of that where we are when you think about,
39:32 you just mentioned we had open code on chat,
39:34 GPT was sitting there.
39:35 Blockchain's been sitting there.
39:36 Crypto's been sitting there
39:37 and there's applications being built.
39:39 Help me understand where we are in that process
39:41 and what you guys are doing.
39:42 - Yeah, I mean, I think,
39:44 so crypto, when we started investing in it,
39:46 it was 2016 and one of the things that we saw
39:50 were a really interesting group of developers
39:52 building largely around Ethereum.
39:53 And we said, this is interesting.
39:55 This is a future that feels pretty far off,
39:57 but it feels very asymmetric.
39:59 And so we should allocate some amount of time
40:01 and resources to understanding that.
40:02 We'll make a variety of investments,
40:04 largely across infrastructure,
40:05 some application layer things,
40:07 and we'll see how this plays out.
40:08 And to be honest, in 2016, I was interested in crypto.
40:12 I was not a maxi who's like, this is it.
40:14 Like Bitcoin is real money.
40:15 Everything else sucks, whatever.
40:17 I thought it was really compelling
40:18 from a technological perspective.
40:19 I think each year that goes by, regardless of price,
40:23 price going up obviously helps,
40:25 but each year that goes by,
40:26 more and more things continue to happen around the world
40:29 that create like really massive volatility inducing events.
40:33 And it make me more and more bullish
40:34 on a very core concept, which is,
40:37 will the next generation of people
40:39 be more focused on privacy, on trustless environments,
40:43 on self-sovereignty, and will coordination
40:46 amongst these groups of people be more decentralized
40:49 in a sense of like remote work is like decentralized work,
40:51 right, like not in a blockchain sense,
40:53 but just more spread out.
40:55 And will you want more efficient systems
40:57 to deal with all of these changes?
40:59 And will eventually a lot of these changes
41:01 that have been unwilling to be more efficient
41:05 from technology be rewritten because the decision makers
41:08 that enter into those institutions say,
41:10 this needs to, we need to expand margin
41:13 using things like smart contracts or whatever it is.
41:15 - But isn't that kind of the knock on it too,
41:17 is that like these decentralized, let's just use as an
41:21 example, like a social network, were not more efficient.
41:25 Right, and if they were more efficient,
41:26 they would have had more uptake in the last few years
41:29 at a time where people, millions of people were obsessed
41:32 with PFPs and NFPs and buying shit coins
41:36 and all that sort of thing,
41:38 but the experience was not great.
41:40 You know what I mean?
41:41 Like, and so, you know, like, and so that's one of the things
41:43 that I find really fascinating because I've been really
41:45 curious in and around this space, but also fairly skeptical.
41:48 And I think Danny in the same way, I mean,
41:51 we started our podcast, our first one
41:53 in January 7th of 2021.
41:55 It was like right before all of the people stuff was in
41:59 and all the doge and all the, you know,
42:00 all the celebrity promoters and it like looked to us
42:03 like an all out media, you know,
42:06 and then on the other side in the,
42:08 in the regulated stock market, you had meme stocks,
42:10 you know, the GameStop and everything like that.
42:12 And everybody was obsessed with all this crap
42:14 and it was all like merging into one.
42:15 And then Clubhouse started,
42:16 then people were doing Clubhouses and Spaces
42:19 all day and night on these things.
42:20 You know what I mean?
42:21 And then it all went away.
42:22 You know what I mean?
42:22 And so it is kind of interesting, but it doesn't seem like,
42:25 I don't know too many normies like us who are engaging
42:29 with anything blockchain based right now.
42:31 - Yeah, I think right now, I mean,
42:32 I think if we were to look at like,
42:34 I don't love this analogy,
42:35 but the one that some people like to use is like,
42:37 the internet was invented in 1989,
42:39 by 1995 it had 15 million users.
42:42 It's like a pretty long time relative to,
42:44 you know, where Ethereum is today.
42:46 I think that in general,
42:48 I'm more so just looking at this idea that like,
42:51 the core concepts underlying crypto
42:52 and specifically areas like decentralized finance,
42:54 decentralized compute, more efficient markets
42:57 that actually create value,
42:58 like gambling is an incredibly valuable asset.
43:02 I think that a lot of the stock market has shown
43:03 that that is also what that has been for a while.
43:06 But I think that the principles underlying it
43:08 will eventually continue to rise,
43:10 especially as demographic shifts happen,
43:13 as well as like the world isn't as American centric.
43:17 And so I think that will matter quite a bit.
43:19 - Can you give us an example of something in your portfolio
43:22 that is working that, you know, application wise?
43:24 I'm sure several are,
43:25 but just give us an example of kind of your favorite.
43:27 - I think, so we have funny enough,
43:29 decentralized finance protocol in our portfolio
43:30 that we see it is called Compound.
43:32 And it's a pure simplistic lending market.
43:35 And so the idea is if right now it's over collateralized
43:38 because doing trustless things and under collateralized
43:40 is very, very difficult, as we've all learned.
43:43 I think the main thing that they are facilitating
43:46 is if you have a lot of wealth,
43:47 whether it's through Bitcoin, Ethereum, whatever,
43:49 and you want to get a loan against it,
43:51 the way in which you would do it today
43:52 takes a ton of time, a ton of people, all these things.
43:54 - Well, there's a particular man in Washington blocking that
43:56 but yeah, you feel good.
43:57 - And instead, all you could do is you can put up your ETH
44:01 and you can take out at a collateralization ratio of 150%.
44:05 The requisite US dollars, whatever it is.
44:08 And so I think those types of things,
44:10 which today sit in a decentralized way
44:13 in which there are lenders, there are suppliers,
44:15 and there are borrowers,
44:17 will continue to be more interesting as one,
44:20 like people understand that institutions are quite fragile
44:23 and they maybe lose because of institutions.
44:25 But I think even two, like the most crazy part
44:28 that you could get to is like the inevitable idea
44:31 that like every country wants to control
44:33 its own financial infrastructure,
44:35 like in the US, everyone cares about the dollar,
44:37 everyone in the world cares about the dollar.
44:38 In many countries, like why do I want to control
44:41 my monetary policy and my financial infrastructure
44:43 outside of ability to maybe control my citizens?
44:45 But if I have a military, I can do that anyways.
44:47 I think more and more of those things will work.
44:49 So Compound has hundreds of millions of dollars
44:52 locked into today, at its peak,
44:53 it has many, many billions of dollars.
44:55 And that is something that is trustless, is instant.
44:58 You can get a loan off of your capital
45:01 in $10 of ETH gas fees and 10 seconds, if you really want to.
45:05 - So perfect segue into Silvergate or into the banks, right?
45:09 'Cause Silvergate tried to do some stuff,
45:11 not like that exactly, but obviously lending,
45:13 using crypto as some type of collateral.
45:16 You're in the VC world.
45:17 You were, I think I was with you guys,
45:20 I was with David for sure,
45:21 the day the news was breaking on Silicon Valley,
45:23 watching in real time, your group,
45:25 not just your group, the group VCs,
45:27 and so was Dan, and we know a lot of people in the space.
45:30 Harrowing moment, realization,
45:32 because this last thing you guys think about,
45:34 you're out there trying to find great ideas and startups
45:36 and they're funding with cash,
45:37 and all of a sudden, forget about your cash, their cash.
45:40 Talk about that moment, and if you can,
45:42 you guys weren't really directly impacted.
45:44 I mean, you had, you know, whatever,
45:45 you don't have to talk about it, but it all ended up fine.
45:47 But your portfolio companies, the panic,
45:49 like where are we in that, and how did that play out?
45:52 - That was like the longest four days of my life.
45:55 And it was so funny because like that Sunday,
45:58 when at the end of the day,
45:59 they kind of said they were gonna rescue the bank,
46:01 like that was a day of like five different
46:02 emergency board meetings.
46:04 And then at the end, you're kind of like,
46:05 this was all pointless.
46:06 I think like that, to the point,
46:08 like that is the other thing that,
46:11 SCB was a unique example because there were so many people
46:13 on the public side that looked at that book for months
46:15 and was like, this is clearly, something is wrong here.
46:17 But the transparency around that is still quite difficult,
46:20 and so not real time.
46:22 And like these institutions don't fail
46:24 in a decentralized public blockchain way.
46:27 Like you can see it, code instantly liquidates people,
46:29 it goes, and we saw that with FTX.
46:31 We've seen it with everything that has happened,
46:33 all Black Swan events, DeFi protocols have kept going.
46:36 On the SBB side and these other banks,
46:38 I think a lot of it was just a like realization
46:41 for a lot of people, myself included,
46:43 that if, just how fragile so many of these things are,
46:47 and all the founders were kind of partially in this idea of,
46:51 I can't believe I'm dealing with like this,
46:53 I'm dealing with like a contraction of multiples,
46:55 dealing with the growth markets going away,
46:56 and now I have to like care about like,
46:58 am I putting my money in the right treasuries?
47:00 What is the bank account?
47:01 What's my FDIC coverage?
47:03 And it's just something that founders aren't used to,
47:04 especially our types of founders,
47:05 who are like highly technical people.
47:07 But I think these types of moments are,
47:11 like that was maybe one of the best outcomes
47:13 because it taught a lot of people a lesson
47:15 and no one theoretically got hurt yet, maybe.
47:18 - So that happened at the same time rates were already,
47:20 obviously rates going up is what caused it,
47:22 but at the same time, that type of liquidity
47:24 was becoming scarce in that world anyway.
47:27 It just exacerbated it.
47:29 So shift now to kind of where we are
47:31 as far as companies that are getting funded
47:33 and companies that aren't getting funded
47:35 or companies that didn't prepare for rainy day
47:37 and are now either forced to sell, shut down,
47:39 or just take a much lower valuation kind of feast or famine.
47:42 - Yeah, I would think of it as there are AI companies,
47:45 'cause everyone loves AI right now.
47:46 - Oh, we're well aware.
47:47 - And those get funding and we can talk about that,
47:50 and there's non AI companies.
47:52 And I think the biggest lesson over the past four years,
47:55 'cause it's funny, we've gotten to see the crypto cycle
47:58 and the AI cycle back to back.
47:59 - What about the metaverse cycle, remember that?
48:01 - Yeah, luckily we didn't have to deal with that.
48:02 - That was kind of the issue.
48:05 - Is that founders often look around
48:07 and because now these financings are so popular
48:09 and so covered in the media,
48:11 you have founders at every moment
48:13 wondering what's wrong with them
48:14 because something else is happening.
48:16 And so now I think you have a lot of founders
48:17 who are not building an AI, who've just been heads down,
48:20 and they're wondering like,
48:20 why does nobody care about what I'm doing,
48:22 even though I'm trying?
48:23 So the growth market for those types of founders
48:26 is still pretty dead.
48:28 And eventually it'll come back,
48:30 but I think there's a lot of people who are terrified
48:32 to deploy real capital right now.
48:34 And even with the NASDAQ ripping like it has this year,
48:37 they're still terrified
48:37 because they know the next thing that's happening,
48:40 which is a bunch of companies raised in 2020, '21, '22,
48:43 raised a ton of money, they hopefully responsibly cut,
48:47 but they're still so far out ahead
48:49 that they're not gonna be able to raise
48:50 'cause the restructuring of the cap table would be so bad,
48:53 no one wants to deal with it.
48:54 And they're probably gonna go out of business
48:55 in the next six to 18 months.
48:58 And it could take longer, it could take slower,
49:00 whatever those things are.
49:02 I generally am of the belief that this is like healthy,
49:05 investing is supposed to be hard.
49:07 I think everyone, especially in my cohort
49:09 of kind of like younger fund managers and investors
49:13 didn't realize that.
49:15 And they thought it was easy
49:16 and they thought it was all about deploying capital.
49:18 - Let's talk about that,
49:19 deploying capital within the AI space right now.
49:22 And we hear this again and again.
49:24 You're seeing a lot of companies pivot
49:25 and thinking about whatever they kind of had whiteboarded
49:29 how they're gonna use machine learning and AI,
49:32 like down the road they're doing right now
49:34 and they're changing their decks and that sort of thing.
49:36 And to me, what's interesting is that
49:39 we've seen this a couple of times now to your point,
49:41 over the last few years,
49:42 we saw people rushing to develop web three strategies,
49:46 rushing to develop AR, VR, metaverse strategies.
49:50 And what the fascinating thing as a public markets person
49:53 who's kind of adjacent to the privates is that,
49:56 from the moment that Mark Zuckerberg
49:58 changed the name of his company in late 2021
50:01 and really reoriented the company,
50:03 the stock went and dropped 75% over the next year.
50:06 And then we saw what happened in crypto.
50:09 It was a $2 trillion market cap as a whole, right?
50:12 And went down probably 80 some percent from its highs.
50:16 And when I look at what's going on with AI here,
50:18 all right, let's just talk about like this week, all right?
50:21 So yesterday or one day this week,
50:23 Microsoft announced the pricing for Copilot,
50:26 which is AI tools that they're gonna sell to enterprise
50:29 in a seat-based fashion.
50:31 And they announced it could be $30 a seat.
50:33 And Amy Hood, the CFO of the company,
50:35 was quoted maybe a week or so ago
50:37 that these could be $10 billion in revenues.
50:39 This is a company that's expected to do
50:41 $210 billion in revenue this year.
50:45 Now, we know how enterprises buy, okay?
50:48 So right now, you know, they might scurry around
50:51 and they might like do this because they want their people
50:53 using these sorts of things or that or whatever.
50:55 It might cannibalize some other things
50:58 that they're already buying from you.
51:00 But the stock rallied on that headline,
51:03 $130 billion in market cap.
51:06 It went up, this is the second largest company
51:09 on the S&P and the NASDAQ, and it rallied.
51:12 So think about that.
51:13 Let's just say we take that $10 billion number
51:15 and let's say it was just meant for these sorts of tools.
51:18 It's trading at 13 times sales before they've even sold
51:22 the tools and the licenses, okay?
51:24 So then the next day, there's a headline that comes out
51:27 that Apple, $3 trillion market cap company,
51:30 is rushing to develop a open AI type tools
51:34 to compete with, obviously, Microsoft and Google.
51:37 And that stock rallies 2%.
51:39 Well, 2% on 3 trillion, so there's a mania going on here.
51:42 Okay, so help us make some sense of this
51:45 because, you know, Danny and I think
51:47 we have a good sense for it.
51:49 There's a chance that this could go the way
51:52 of those last two manias that we saw.
51:54 I don't mean that Microsoft and Apple and Google
51:57 are gonna lose 75% of their value.
51:59 But if they've just gained a half a trillion based on this,
52:03 I don't know, man, it seems like it could get
52:06 a little dicey in the not so distant future.
52:08 - Yeah, I think one of the things that is most interesting
52:12 about public market investing right now
52:15 is that everyone centered around the idea
52:19 of hyperscalers and compounding.
52:21 And so you talk to any of the Tiger Cub managers,
52:24 any of these people, their belief in how they invest
52:27 is you find something and you let it compound.
52:30 And so if you believe AI is a winner-take-most market,
52:33 and if you believe this is the first technology
52:36 that incumbents have, as some people say,
52:38 first right of refusal to, you just might assume
52:41 that just compounds in a way that is a scale
52:44 that you didn't appreciate before.
52:46 And I think that that's what happened in private markets,
52:48 like in 2020, which was people saw businesses
52:51 that became 30, 50, $100 billion companies,
52:56 like Snowflake is an example of that.
52:57 And they assumed that that meant that,
53:00 they were so surprised by the upside of that,
53:02 they said, well, so many of these businesses
53:03 are gonna surprise so much larger to the upside,
53:05 'cause they're gonna compound at a rate
53:06 that we just can't appreciate.
53:07 Data dot. - Like Zoom, right?
53:08 - Zoom, data dot, all these things.
53:10 - And I say Zoom, right, like,
53:13 I'm being sarcastic in a way.
53:15 - But can I just add to that?
53:16 Because I think it's important.
53:17 So obviously, the favorite name in the space, NVIDIA,
53:21 certainly had a coming out party on AI, right?
53:24 And it's been nonstop ever since.
53:26 If you're that portfolio manager of the public markets,
53:29 and you had that theme, right, you're right.
53:32 Where's the discipline?
53:33 And I'm not saying that they're not selling,
53:35 you get caught up, I think what Dan's talking about
53:36 is the incrementalism of this AI craze and trade.
53:40 It's going into various pockets.
53:41 We're talking in the trillions of dollars now
53:43 that's been added, and eventually, right,
53:45 it'll catch up, I mean, at this rate,
53:47 it could be 20 years before you can catch up
53:48 to validate that.
53:49 Again, not my area, I don't know.
53:51 But the problem with that is that,
53:52 and this is when we shifted over to the public markets,
53:55 this is on the tape podcast, would be,
53:58 how do you tell people, oh, it's down 10%,
54:00 this is where you buy it, or it's down 20%
54:02 from the highest, because there wasn't a fundamental reason,
54:04 there was a thesis-driven reason, and there is fundamental.
54:07 So I think what Dan's getting at,
54:08 and you're sitting in a position
54:09 where you have private companies,
54:10 you're not mark-to-market each day,
54:11 you know what you, you're not missing out
54:13 because you already, so you have an objective point of view,
54:16 which is what I think Dan is trying to get at,
54:17 that you can see it.
54:18 And I'm not telling you that you should be shorting
54:20 in video or something, but that, I think,
54:22 is what Dan is saying.
54:23 It's like the people that are just buying it
54:25 because it's a theme versus what does it actually mean
54:28 to the numbers, how big is this thing?
54:30 - I think it's, I think what we've seen
54:33 is if you have built-in distribution,
54:34 it can be a meaningful revenue generator.
54:37 Like the notion, kind of story now,
54:40 is like they add AI, revenue, they hit the revenue number
54:43 that they expected in like a week, for the year.
54:46 - But the great example is what you just gave
54:48 of COVID in 2020. - For sure.
54:50 - Is that the pull forward, and the problem right now
54:52 is that these massive market caps
54:54 are discounting that future, you know what I mean, growth.
54:57 And at some point, you're gonna have the deceleration
54:59 that we started to see in 2021,
55:02 and that's when it was lights out.
55:04 And some of these companies in the public markets
55:06 haven't, Zoom, which I mentioned,
55:08 have not been able to get out of their own way,
55:10 and they're profitable, right?
55:11 And they're actually probably,
55:14 we're operating as a better company now
55:16 than they were two years ago.
55:18 And so part of the point is that, you know,
55:20 investors, you know, they have very short memories
55:23 when it comes to some of this stuff.
55:25 So it feels euphoric right now, you know what I mean?
55:28 But it could change.
55:29 And the last point I'll make,
55:29 and I wanna hear you dissect all our idiocy here,
55:33 is that Nvidia has also been a part
55:35 of every one of these crazes.
55:37 Crypto mining, metaverse, gaming, like every single one.
55:42 Like, you know, and maybe Jensen Wang,
55:44 it doesn't get the clout that Elon has
55:48 as far as being a salesman, but they're there.
55:50 And I know people are double, triple ordering
55:52 these things right now, but at some point,
55:55 this is gonna moderate quite substantially.
55:57 And if you just think of the last two or three months,
55:59 this company has gained $400 billion in market cap
56:02 based on a beat of $4 billion and a quarter
56:06 of graphic chips.
56:07 - Yeah.
56:08 Listen, I longed Nvidia last year.
56:10 I sold a couple of weeks ago.
56:12 I think that like, in some ways,
56:14 you guys are getting at the main thing,
56:16 which is like narratives are driving these markets
56:18 more than anything.
56:19 There is now what is effectively an infinite bid
56:22 into these markets because people deploy
56:24 into their ETIs, the S&P, ETI, whatever.
56:27 And there is now, this is all happening
56:31 in a terrible macro environment for equities.
56:34 And so like a more simplistic view could be like,
56:37 this is great.
56:38 We're now like looking through rate lowering.
56:41 And like, if that happens, like I have another 30, 40, 70,
56:45 100% to run on all of these hyperscaler things
56:48 from a just pure narrative capture perspective
56:50 on a market cap.
56:51 I think this is all to say, like,
56:54 I just don't believe a lot of these investments
56:57 are fundamentally driven.
56:58 And I think a lot of it is as thesis driven as we are
57:02 of let's look at things and extrapolate them.
57:03 And if we're on the margins off,
57:05 but we look at 10 year time horizons,
57:07 I think weirdly a lot of equity investors today
57:09 have that same lens because they've only been long only,
57:12 they've been largely tech and they've been in the same
57:16 80% portfolio overlap across all the different funds
57:19 and they get annihilated at the same time,
57:21 depending on how well they sell,
57:23 like CO2 sold very well, others sold horribly.
57:26 And then they all long right back again.
57:28 So I don't think that there's a narrative.
57:31 I don't think there's anything you can fundamentally say
57:33 outside of like something will happen
57:35 that will create significantly larger value
57:38 than anyone appreciates.
57:39 And people believe in asymmetric upside now.
57:42 And likely the lens at which they're looking
57:44 at these equities is in a similar lens
57:46 that we look at our portfolio,
57:48 which is we need to have two good ideas out of 30
57:51 and our investors make money.
57:52 And theirs is probably we need to have one good idea
57:54 out of five.
57:55 - All right, well, Danny and I are still searching
57:56 for one good long idea in 2023 in the stock market.
58:00 Michael Dempsey, this was fascinating
58:02 because Danny and I, we're kind of dummies
58:05 when it comes to some of this sort of stuff.
58:06 So it was really interesting to hear
58:08 about some of the things in your portfolio, your process,
58:11 and obviously shedding a little light on the public markets
58:14 because you did come from the public markets.
58:16 And so valuation and innovative tech,
58:18 these are things that are very much part of your DNA.
58:21 So it was really helpful to hear it from your perspective.
58:23 So we really appreciate you joining us on the pod here.
58:25 - Yeah, thanks so much.
58:26 - Michael, thanks for coming on.
58:27 It allowed me to come on this pod.
58:29 - Well, listen, next time he comes back,
58:31 maybe we'll invite you back too, Danny.
58:33 - I appreciate it.
58:34 - Thanks for being here, Danny.
58:35 - There we go.
58:36 - My pleasure.
58:37 - Thanks.
58:37 (upbeat music)
58:40 [BLANK_AUDIO]