Latest / Newcomer Pod / Shaun Maguire's Growth Hack
Transcript
- 0:00Hey everybody, I'm back. We're back.
- 0:02Eric, Tom, Madeline, we're all back on the Newcomer Podcast.
- 0:07I've been here, I've been here. Yeah, So the show that you may
- 0:11have liked or may not have liked is now returned.
- 0:13Listen, I think we got a lot to go through.
- 0:16It's been a lot happening in AI. And I understand this is not the
- 0:19Cerebral Valley podcast specifically, but money is
- 0:22flying. Money is flying here in tech.
- 0:25If you are an AI researcher, you should be making at least 8
- 0:28figures. You should probably start
- 0:30claiming that you're an AI researcher to just see like what
- 0:33the offers look like. Where do we even start?
- 0:35Yeah, Madeline, you want to give us a rundown?
- 0:37I mean, you've been digging into, I feel like there are
- 0:40tables on tables of different Meta employees making a bunch of
- 0:44money. We've tried to make the table to
- 0:46end all tables, just pulling them all together with What's
- 0:49your read right now on Mark Zuckerberg's, uh, recruiting
- 0:53frenzy spending bonanza here? Right.
- 0:56Well, 8 figure comp packages from Zuckerberg are the entry
- 1:00level some people are getting, you know, 9 figure comp packages
- 1:02out of this hiring spree. But it really kicked off in
- 1:05earnest last month with Meta's Aqua hire of Alexander Wang from
- 1:10Scale AI to lead the new Meta Super Intelligence Lab team
- 1:14within the company. Meta paid over 14 billion for a
- 1:1714% stake in Scale just to get Alex over on the team.
- 1:22But since then, he's also poached Nat Friedman and Daniel
- 1:25Gross there and is, you know, buying out a stake in their
- 1:28investment fund to have them come work for.
- 1:30So really just unprecedented amounts of capital wrangling to
- 1:34get the biggest names in AI to lead the team.
- 1:37And they've got, you know, obviously they've been picking
- 1:40open a is pocket. They've got more than a half
- 1:43dozen people from Open AI. They've got people from
- 1:45DeepMind, Anthropic, Apple. As of this week, as of Thursday
- 1:49recording, they've poached at least 10 people from Open AI to
- 1:52join this team, in addition to those other big names.
- 1:55As a side point to all of this, I mean, you know, a lot of the
- 1:58narrative really kicked off, you know, when, like you said, the
- 2:01scale deal happened. But then Sam, I thought really
- 2:04unstrategically goes on I think his brothers podcast.
- 2:08And he's just like, yeah, they're coming at us, but they
- 2:10haven't gotten anyone good yet, which is a stupid thing to say.
- 2:14He's like getting them on. On all on all fronts, it's we
- 2:17have one clearly Zach, one of the most psychotically
- 2:20competitive people in the history of, you know,
- 2:22civilization saying like you can't get our people is like
- 2:25just red meat for him. But also like you're basically
- 2:28saying everyone that they did get was shitty and they did end
- 2:31up getting shitty people. And like I understand like Sam
- 2:34is still figuring out how to like measure his.
- 2:36Like I'm a deep thinking AI guy and I'm very calm with the fact
- 2:40that he also is psychotically competitive and doesn't like the
- 2:43fact that Zuck is coming at. The the Philosopher king posture
- 2:46with the cutthroat business leader of our AI future. 11
- 2:50could also say missionary versus mercenary in that dichotomy
- 2:53there. It's not just open AI, to be
- 2:55clear, they've poached a lot of deep mind researchers.
- 2:59And then roaming Peng from Apple Bloomberg is now reporting that
- 3:03they have poached with a 200,000,000 / 4 years total comp
- 3:07package. So these are serious offers just
- 3:11from across the table of AI research.
- 3:13People should be cheering for this.
- 3:14This is my my first take I think.
- 3:16You know gone are the old days when Apple and Google had anti
- 3:20competitive alliances to artificially repress salaries.
- 3:24I mean, these, this is the working class these these are
- 3:28workers getting paid a lot of money for producing or Tom, I'm
- 3:33curious what you take that like it's like.
- 3:35Yeah, this is ultimately the proletariat winning.
- 3:39Yeah, sure. The white collar proletariat,
- 3:41that they're all basically creating technology that will
- 3:43one day make white collar jobs no longer viable.
- 3:46Yeah, you're right. There's a certain logic.
- 3:47If you're the last man to ever work, it's like, yeah, you take
- 3:50the money, man. A lot.
- 3:51Like, right, like this is generational wealth because your
- 3:53children, if they're also white collar workers, will never have
- 3:56jobs. Yeah, I don't know.
- 3:58I, I guess like we've been following the kind of turmoil at
- 4:02meta on their AI team for as long as I've been with, with you
- 4:04guys. I think it was like the first
- 4:06story I wrote at newcomer was like, man, Llama is, is sucking
- 4:10right now. Like they just have not done a
- 4:12good job keeping up and DeepSeek, I think really fuck
- 4:15things up for them. And so at some point, and there
- 4:18were some good stories in the, you know, New York Times and
- 4:20elsewhere about kind of like internally at meta what's been
- 4:22going on, but suck. Just realized, like, either if
- 4:25we're going to be playing at this, we need to use the only
- 4:27asset that we really have here right now, which is like, you
- 4:30know, endless, deep, deep, deep pockets.
- 4:32It's just money. Yeah.
- 4:34And that's always kind of Ben Zuck's thing, right?
- 4:36You know, like the ideas side of him is not always sharp.
- 4:39He he has money and he has network effects.
- 4:42There's a lot of users, yeah. Yeah, yeah.
- 4:45So like Facebook is, you know, not growing at all almost.
- 4:48And so you buy Instagram and you and you get WhatsApp, you
- 4:50basically just buy your way into like continued relevance as a
- 4:53company. And like there's no specific
- 4:55company to buy really an AI, an AI that's going to make you, you
- 4:59know, top, you know, top tier. So just get the research because
- 5:01it's really just talent. Well.
- 5:04There's some, there's, there's a very, very limited pool of
- 5:07researchers that can be building these models at this scale.
- 5:09So it is really like a small pool of people that can be bid
- 5:14on at the Super high comp packages.
- 5:16Like there is just a limited pool of worker resources to go
- 5:19around. You guys are a little deeper in
- 5:21it than I am, but I have to imagine open AI's weird
- 5:25shareholder structure is not helping it retain these
- 5:28employees. Like, first of all, you have
- 5:30this question mark around what these shares really mean and how
- 5:34how the company sort of sorts out.
- 5:36Where is it meta? You know, you're getting public,
- 5:38public shares, you know what they're worth.
- 5:41And then there have been all these secondaries that allowed
- 5:43open AI people to get liquidity. And so it's like, whoa, I've
- 5:48already gotten paid out from you, I can go on to the next
- 5:50thing. Or what's your read on how
- 5:52opening I set this thing up? Yeah, I think well the the this
- 5:55was brings back to like Sam's, you know like hubristic comment
- 6:01like you can't poach from U.S. Open AI has been losing people
- 6:04constantly over it's life as a company.
- 6:06This is not like a a company full of lifers.
- 6:08There are. Some, I mean Anthropic exists as
- 6:10a defection from open AI. The story of Open AI is a story
- 6:14of people leaving open AI. Basically right?
- 6:16Like like Elon, Dario, you know, Mira Ilia, like it's, it's, it's
- 6:22an endless list of alumni that, that come there.
- 6:24So this idea that these guys are like all missionaries who want
- 6:27to stay there for life, it's just not true.
- 6:29And it's a complete, I think they're the sympathetic argument
- 6:32is that they're missionaries in that they want to be there when
- 6:35AGI is bored. I think a lot of them do want to
- 6:37be be in the place where it's born.
- 6:39And if you can convince them like, hey, you know, our our
- 6:43superstar team is moving from this place to this place, you'll
- 6:46have another shot at it. Or you think this is about money
- 6:48and not not about well, it is AGI.
- 6:51Yeah. Well, I also like in reporting
- 6:52out that piece that I did a couple weeks ago about basically
- 6:56all of this and and Meta's recruitment, I was really struck
- 6:58by 1 researcher I talked to who was like basically creating an
- 7:02AI model as a recipe. And there are certain people who
- 7:04have the recipe in their brain and the what you're paying for
- 7:07when you poach someone from a labs to go to the new place so
- 7:09they can recreate the recipe. And what's going to end up
- 7:12happening is that like you have all the recipe knowers at the
- 7:15different labs around the, the industry and they're probably
- 7:18all going to reach AGI at the same time or very close to each
- 7:21other. And so I think what my guess is
- 7:24like the realization a lot of researchers had is like, there
- 7:26isn't the place that it's going to happen.
- 7:28It's all leafy. Now, and everybody's going to
- 7:29have it, might as well get paid the most along the way, yeah?
- 7:32So why not, why not get yours? Like, if that money's out there,
- 7:34what are you going to be like a missionary and take less money
- 7:37with like weirdly illiquid stock so that you can like maybe get
- 7:40AGI two weeks before the next lab does?
- 7:43Like, it doesn't really compute. And so I think that more than
- 7:46anything, which is man. We need to build a business just
- 7:49for these 100 million employee like whatever they want to read
- 7:52and then just the advertising budget on that, like get in
- 7:54front of like the ludantic. Just like just people pays.
- 7:59Like a ten person e-mail, you know, they're like hedge fund
- 8:01newsletters we'll have just like the AI researcher, you know,
- 8:04like 20 person. How much would Zuck pay us so
- 8:08that we could have like, you know, ads in front of like the
- 8:1010 most important research? Exactly.
- 8:12Just like on their minds. Yeah, so like that, that's,
- 8:15that's all been very funny to watch.
- 8:17Yeah, play out. And then you know, like we'll
- 8:19see how it all happens with meta, right?
- 8:21Like they have the teams in place and I'm interested to see
- 8:24how long it takes for them to like redo the llama training to
- 8:28get to a point where this was all worth.
- 8:30At least they'll get a model that's competitive because
- 8:32that's the problem. I mean, making money off of it
- 8:33is a whole different story with with that company.
- 8:36They're open source. They've never really made a dime
- 8:38off of it. Like this is going to be a long
- 8:40term play for for for meta to make it work.
- 8:43But in the meantime, all these guys get rich.
- 8:46So what does this all mean for Llama?
- 8:48Obviously Llama 4 was sort of a disaster.
- 8:50Do you think Meta sticks with its open source strategy of just
- 8:55making no money and foiling the competition?
- 8:58Or you think, now spend all this money and better come up with a
- 9:01a proprietary product for itself?
- 9:03I hope they do. I hope they keep with the open
- 9:05source thing because to me if Zuck's whole play is about
- 9:08fucking over Sam just. Burn the forest down.
- 9:11That would be the best way to do it, right?
- 9:13And I always thought that was the smartest aspect that he had
- 9:15there in the 1st places, like looking, because he's very
- 9:17jealous, I'm assuming, of Sam. And as he was coming up to being
- 9:20like the voice of Silicon Valley in this moment that he's like,
- 9:23oh, well, what if we just take the same technology but release
- 9:25it for free? That'll destroy these people.
- 9:27They'll have no business, which I thought was cool.
- 9:30I actually thought that was like an actually badass play.
- 9:32And if he decides to get rid of it because he just wants to make
- 9:35money in some consumer product, I will be really disappointed.
- 9:38There is a very subtle monetization mechanic on this,
- 9:41which I saw YouTube said that people who dump like AI stuff
- 9:47onto YouTube will not be allowed to monetize it.
- 9:49And there's certain argument that like Google and Facebook,
- 9:53they could be the content providers.
- 9:56You know, it's like if, if there's this AI appealing
- 9:59content, it's like you block everyone else from it and you
- 10:01are sort of the exclusive content provider.
- 10:04I don't, I don't know that's. I haven't heard that.
- 10:06That's interesting. I mean, The funny thing with
- 10:09Facebook and AI or Meta and AI is that, you know, like when
- 10:13stories would come out that like Llama 4 Behemoth is delayed,
- 10:16their stock drops. And I didn't ever really
- 10:18understood that because they don't make any money off of it.
- 10:20But there's just this general acceptance on the part of
- 10:23investors that like, well, AI has to matter for everyone.
- 10:26So it's important that you like have a, you know, a.
- 10:28And there is like, you know, open source projects if you're,
- 10:31you know, if you're around them, usually you know, people have
- 10:34figured out how to make money. Sure, it's a distribution play
- 10:37like that's what it is with Google and Chrome and and and
- 10:39Android. Did I say Chrome?
- 10:41I mean, Android with Google and Android, so yeah, sure, I don't
- 10:44worry too much about like their ability to make money off of it.
- 10:47But for this in the short term, having this, you know, open
- 10:50source high level model, like like we saw DeepSeek really
- 10:54through the industry in a turmoil for at least a little
- 10:56bit. You know, you think Meta could
- 10:58do that at a much larger scale. So it's kind of exciting.
- 11:01I'm betting on closed source to be clear.
- 11:03I mean, that's the smarter bet, but also the two biggest
- 11:06companies that are succeeding on it right now are closed source,
- 11:09right, and Tropic and Open AI and then I guess Google as well.
- 11:12So yeah, that's a. That's a safe bet.
- 11:14Well, moving on, outside of open source or closed source, you
- 11:16probably don't want your model of becoming Mecca Hitler.
- 11:19So that was the other big story, political.
- 11:22Orientation. These days, that's.
- 11:23True, I don't know. If it happens, it happens, yeah.
- 11:28Unfortunately, we probably don't want the API to be Mecca Hitler
- 11:30on this podcast. But yes, we're talking of course
- 11:33about Grok 4. XA is new model that was debuted
- 11:38by Elon Musk earlier this week, which was overshadowed by its
- 11:41Grok chatbot interface on X. Just ramping up the racism and
- 11:45politically incorrect speech in all of its replies to 11.
- 11:48So you can't. Really be like, oh, we know what
- 11:50that last name means, you know, just like it's it's so.
- 11:55Bad. I couldn't tell if it was real,
- 11:58right? Did you see the tweet or it was
- 12:01like a screenshot of one that was like Grok 4 was getting
- 12:04really explicit about Linda Yak or Reno.
- 12:07Did you see that? One I I saw something about Will
- 12:09Stansel. Yeah.
- 12:11How did Will Stancil get inserted?
- 12:13That was true. Oh, no.
- 12:14Will Stancil is like the lead anecdote to the journal story
- 12:17about this, which I think is a little ridiculous.
- 12:19Like, you know, come on, It's like I.
- 12:20Mean people are like walking. They're like convincing the
- 12:22model to do this right. It's like they're doing like
- 12:25thought experiments sort of stuff.
- 12:27Yeah, it's like, don't, I don't know, I there's clearly an issue
- 12:31in terms of it's like, you know, guidelines and and and how it's
- 12:35allowed to interact with people. I don't honestly think Rock was
- 12:38giving people specific ideas on how to like attack and rape Will
- 12:41Stancil. And this was like a, you know,
- 12:43he. Was a random Lib Twitter account
- 12:46that's too deeply ingrained in our brains.
- 12:48He also ran. Thanks to thanks to you and I
- 12:51guess. Can I speak out in defense of
- 12:53Mecca Hitler for one second? Sure.
- 12:56Not a take I expected. I can't believe I'm going to say
- 12:59this. I think X AI deserves a lot of
- 13:00credit for what they've built within the model game.
- 13:04They are a far smaller lab than almost all the other companies.
- 13:07They basically were made-up by, you know, Igor Babushkin, who is
- 13:11this ex DeepMind guy who came in there definitely on the, you
- 13:14know, premise of wanting to build the anti woke open AI, the
- 13:17anti woke, you know, woke chap GBT and they fucking did it.
- 13:21Like Grok 3 is a really good model.
- 13:23It it like it's absolutely competitive with all the other
- 13:26big models. They built this giant database
- 13:28in Nashville Colossus. I don't know yet how good Grok 4
- 13:32is, but like according to the benchmarks, which are can be
- 13:34which can be games, so you don't want to give too much credit to
- 13:37it. And as we saw with with Llama 4,
- 13:40you can do well on the benchmarks and actually have a
- 13:41completely useless model. But like I look at Microsoft and
- 13:45what they tried to do with Mustafa in hiring a top tier
- 13:47researcher. They had tons of resources.
- 13:49They have not put out a large model.
- 13:51You look at Meta, they have completely fucked up their whole
- 13:53thing. Grok is out there putting out
- 13:55models that seem to be fairly competitive, if not better than
- 13:57the rest that's out there. So like, I can't believe I'm
- 14:00defending Elon in all of this, but like maybe you get a little
- 14:04Mecca Hitler on the way to doing it.
- 14:06It's like smart enough, who cares what it's?
- 14:10I mean, maybe the smart, yeah, it's like, you know, we all have
- 14:14friends like that, that sometimes they're really smart,
- 14:16but sometimes they, you know, advocate Mecca Hitler ideology.
- 14:19I just think about. Your your tanks are all over the
- 14:23place, no? No, no, I mean, I I'm I'm
- 14:25obviously joking about Mecca Hitler, but genuinely speaking,
- 14:29like I think you know what, like Maddie, you said like it
- 14:31overshadowed like the release of Grok 4, which is maybe true
- 14:34because Mecca Hitler was very funny, maybe scary, if you're
- 14:36will Stancil. And also inadvertently led to
- 14:39Linda Yakarino stepping down, I think in some big parts.
- 14:42So we. We don't know.
- 14:44We don't know. I haven't read a good story
- 14:46about it yet. Yeah, it could have.
- 14:48It could have, I just, I don't know.
- 14:50I, I, I don't even want to come off as contrarian here.
- 14:53I just think like what Xai is doing is impressive.
- 14:55Like in the world where like most labs do not put out very
- 14:59good models, like they consistently or at least twice
- 15:01in a row put out something that seemed to matter.
- 15:03I mean, somebody was commenting that, you know, xai like their
- 15:06their presentation was like the stuff of engineers.
- 15:09It just feels like a company that's all engineers, no
- 15:12philosopher kings, and I would prefer one of the philosopher
- 15:16king companies win or xai get some scruples.
- 15:20I mean, what is freedom in the context of a model?
- 15:24You know, I mean, the model has like a point of view, you know
- 15:25what I mean? What there isn't this sort of no
- 15:29matter they're going to direct it towards one thing or away
- 15:32from another. I'm interested in seeing how it
- 15:34actually is implemented and used by people because if it's like
- 15:38actually impressive, you know, people shouldn't just discount
- 15:42all of that because, you know, it seemed to have some post
- 15:44training issues. Right.
- 15:45I mean, I don't think they will. I mean, first of all, Grok has
- 15:47this huge advantage of being in front of everybody on X.
- 15:50So if they have something good, it's it's got the users in a way
- 15:54that people like Anthropic don't.
- 15:56Yeah, they also have a great base of text based training data
- 16:00with all of this, you know, real human text data.
- 16:03Well, that's clearly where the Nazi shit came from.
- 16:05I, I will do the classic reporter thing, which is I, I
- 16:08wrote in a story. I forget if it was my bear case
- 16:12in opening an eye or what, but I've been pretty bullish on Grok
- 16:15for a while, despite, you know, being somewhat of an Elon.
- 16:18Testing. I didn't know that.
- 16:19What's your? What's your argument for that?
- 16:21What? Was we're just like talking to
- 16:22investors that they really thought like Grok was going to
- 16:25catch up. And that like the internal signs
- 16:27were that they had really good people and that just investors
- 16:30were taking it very seriously in a way that, you know, they still
- 16:35had to prove to the public. I mean, it's raised a ton of
- 16:37money, you know, and I, I don't think the people making those
- 16:40investments are idiots. You know, I think they thought
- 16:44Elon will catch up, and it seems like he has.
- 16:47It's always unclear with the Elon investments, right?
- 16:49Like to a degree, it always feels like it's a little bit of
- 16:51like, well, if we give Elon money for this maybe idiotic
- 16:54project, we'll get access to space.
- 16:55But Rock was so much the X AI stuff and Twitter, there's a lot
- 16:59of money to do it. Also, I think those started to
- 17:03need to pay off. Those were the ones I think, you
- 17:05know, like Boring Company. Some of those you're like, oh,
- 17:07OK, well, I'll throw them a bone.
- 17:09Hope some of the other ones workout.
- 17:10But like, I think X AI was one of the ones that needed to
- 17:13workout. You know, it'll be interesting.
- 17:15What financial machinations he pulls between Tesla and XAI?
- 17:19Think he already said that Tesla would soon start having Grok in
- 17:25the cars, so I'm sure there's going to be all sorts of self
- 17:28dealing between these companies. Well, yeah, that's that's kind
- 17:32of his thing. Which the investors want.
- 17:34It's like, how do we make sure they all make money like?
- 17:37Basically it's like, it's like, yeah, yeah, if you could end up,
- 17:40if you were an ex, you know, Holder and you ended up with
- 17:43XAI, that's a little bit better. But yeah, if you can somehow
- 17:46spin this off into Tesla stock or or even better, SpaceX stock,
- 17:50that would be nice. Well, that's one of the core
- 17:52reporting questions. I, I think we didn't really ever
- 17:55answer was like did X shareholders or XAI shareholders
- 18:00get a better deal in that merger.
- 18:02And I think the answer was like on metrics at the time, you
- 18:06would have said obviously X shareholders, but based on what
- 18:10you just said and what a lot of people think, like XAI has this
- 18:12unlimited valuation potential. So you're getting a good deal
- 18:16getting access to XAI. That said, they don't really
- 18:20make all that much money, right? Now, right, it's all, I mean,
- 18:23it's like, but yeah, with Elon, what is making money, you know,
- 18:25it's like about it's, it's a valuation thing.
- 18:28So while we're talking about unlimited valuation for for XAI,
- 18:32guess you guys have been spending a lot of time thinking
- 18:34about, and I was talking to our colleague Jonathan Weber about
- 18:37this. Are we in a bubble?
- 18:40Why now do you think we potentially are in a bubble?
- 18:42Stocks have been going up for for some time.
- 18:45Like what's what's particularly bubble like about what's
- 18:47happening now? I think the biggest thing really
- 18:51was just the RAMP report that the RAMPS research data put out,
- 18:55which is of course based, it's a smaller sample size of people,
- 18:58you know, using RAMP and what their customers are spending
- 19:01money on. But I think there was a
- 19:03noticeable kind of peak hit in the last month around tool
- 19:08adoption. And while it's not a bubble, I
- 19:10would say on AI, it could be a bubble on how much people are
- 19:14willing to pay for AI tools, especially at the enterprise
- 19:16level. So while that may not hurt AI
- 19:19overall in the long run, all of these companies that are
- 19:21building the next big tool for enterprise AI may have been
- 19:25charging a bit too much. And I mean, we saw that, you
- 19:27know, with cursor prices going way up and people complaining
- 19:30and the getting subsequently reversed.
- 19:32Like it could be just a point where not everyone is willing to
- 19:36pay for infinite AI anymore if they think it'll get them ahead.
- 19:39They may want to focus on cost at this point.
- 19:42Let me take a stab at this. I mean, first of all, newcomer
- 19:45is not declaring it's a bubble. I think we're pointing out that
- 19:48there are a couple, an impressionistic take that there
- 19:51are some things that might raise bubble questions. 1, is this
- 19:55survey from Ramp showing that sort of spending on new AI apps
- 20:00is going down, right? The number of apps I'm also
- 20:03interested in sort of core Weave and the infrastructure companies
- 20:07and the heavy valuations. That those companies have
- 20:11received you know they've seen a lot of valuation growth and now
- 20:14people are starting to ask questions about how they account
- 20:18for things so so yeah I think they're they're cut there's the
- 20:21sort of highly valued infrastructure players racing
- 20:24into AI. There's this sort of ramp data
- 20:28turn. So the, I think there are
- 20:31questions about things. I think the, the key, the key
- 20:33point here though, is like you can have like sort of a
- 20:36correction in one sector, but not necessarily in AI.
- 20:40Like I, I think the mood could turn against infrastructure
- 20:43plays like.com era what there was this huge telecom build out
- 20:48that was sort of correct, but they, they did it way too fast
- 20:52and it wasn't that valuable. And so I think they're going to
- 20:55be definitely nobody, nobody can test that there will be bubbles
- 20:58within AI where some pieces of the stack are not that valuable,
- 21:03but need to exist. People spend a lot of money to
- 21:06build them and then it turns out they don't get to retain a lot
- 21:09of the actual profits from AI. So it could be foundation model
- 21:13companies, it could be infrastructure companies.
- 21:16But certainly I think there are pieces of this where a lot of
- 21:19money will be spent and not a lot of money will be sort of
- 21:22retained by that category. And, and we don't know how
- 21:25that's going to play out, but it's it's been to the benefit of
- 21:27everybody who wants to build an AI.
- 21:29I mean, say, we've also kind of seen this with sort of the
- 21:31initial chat bot wave, right? Like there was a period of time
- 21:34where there were all these start-ups coming out with chat
- 21:36bots, which ultimately kind of just absorbed into a few players
- 21:39and that wasn't actually monetizable.
- 21:41And that wasn't the end of chat bots.
- 21:43That just meant that that frame of how to make money off of AI
- 21:47with chat bots wasn't going to be the way that everything
- 21:50became as an individual company or application and had to be
- 21:53blended into something like an open AI which could provide all
- 21:55these other services. So I think there certainly will
- 21:58be winners. It just may not be so stratified
- 22:00across all these different products.
- 22:02It's also, if you buy the argument that a lot of the bulls
- 22:04have on AI that this is like a multi year, you know,
- 22:08transformational cycle, we're gonna have dips in it, right?
- 22:12It's just, it's just the nature of it.
- 22:13Very bullish on AI while saying there's seems like maybe a
- 22:16bubble right now. Yeah, I'm not even sure where I
- 22:19fall in it necessarily, but I just don't think we're going to
- 22:21be see like the peak bubble now. Like I think there's going to
- 22:24be, it could be a couple years down the line where we see like,
- 22:27oh, in fact, this did not turn out to be as lucrative as a lot
- 22:30of the people thought it was going to be.
- 22:32And the investments were well out of their skis.
- 22:34I'm just not ready to call it now.
- 22:36You know, I just haven't seen enough indication.
- 22:37Catalyst beside we're over extrapolating from one ramp
- 22:42survey otherwise, I mean, yeah. Yeah.
- 22:45Well, there were things that were, well, there were things
- 22:47that I think caused people to take notice, like, you know,
- 22:50Microsoft pulling out and kind of rejiggering some of its
- 22:53leases on data centres, which people for a time were deciding
- 22:56that meant that they no longer wanted to build data centers,
- 22:58which just isn't true, May not seem very specific to Microsoft
- 23:02and its open AI relationship. But people look at those sort of
- 23:05things and private equity companies that own a lot of the
- 23:08data center companies may be changing their their leases.
- 23:10So this has been going on for some time, but I just don't
- 23:13think like, like you're saying, I just don't think we would have
- 23:15seen it now. It felt very out of the blue.
- 23:17Here's Here's what's sort of confusing about this situation.
- 23:21OK. If NVIDIA produces high quality
- 23:26AI chips at a faster pace like they say they're going to, if
- 23:29they're going to refresh these chips, will that be catalyst to
- 23:35a bubble bursting or will that be good for AI, right.
- 23:38In some ways, if they update their chips more often, the
- 23:41infrastructure companies which have spent a lot of money on
- 23:44chips may be miscalculating how valuable that investment is and
- 23:49may need to spend a lot of money to move on to the next one.
- 23:51So, but obviously improving technology will be good for what
- 23:57consumers experience in terms of AI, but it could disrupt some of
- 24:03the infrastructure providers. So these are really sort of
- 24:06complicated questions in terms of which companies get hit and
- 24:09which companies do really well. Yeah, totally.
- 24:11And I think you could see AI being as transformative as a lot
- 24:14of the bulls expected to be. And NVIDIA takes a hit in in a
- 24:18different way because there's more competition or because
- 24:21there doesn't need to be, you know, the next generation of
- 24:24chips being as expensive as they are, you have real breakthroughs
- 24:27in terms of efficiency that makes it a better product for
- 24:30people, more affordable for businesses.
- 24:32But someone like it, NVIDIA gets its valuation just kind of
- 24:35decimated. Not predicting that, but I'm
- 24:37saying like, yeah. Both can be true.
- 24:39I mean, I'm, I'm very bullish still on what's to be built in
- 24:42AI in a lot of these start-ups. What we're hearing from people
- 24:45like to me, I think there's, yeah, it remains my point of
- 24:50view that there's a lot of capability available from the
- 24:55existing models that people have not figured out how to turn into
- 24:58products and people are going to turn those into products and
- 25:01those products will make money. What is not clear to me is if
- 25:04the foundation models get to keep return all the money they
- 25:08spent to build up that capability, and whether the
- 25:11infrastructure companies are properly valued based on
- 25:14everything they've contributed to creating all that value.
- 25:17Right. And then, you know, on the
- 25:19enterprise side of things, like I still think we're not at a
- 25:24point where businesses are seeing a great amount of benefit
- 25:27from this technology. They fair indicator they just
- 25:30haven't figured it out yet. Like I did this panel a month or
- 25:33so ago with like all these chief data officers at fairly, you
- 25:36know, old line companies, you know, like a Ford Credit Union
- 25:40or Danone, like the yogurt company and they have the.
- 25:47Earth business guy now. Well, I mean, I was invited to
- 25:50do it. You know, I don't, I don't, I
- 25:52didn't prior to this have a deep relationship with the yogurt
- 25:55industry. But but, but like, you know,
- 25:58they all have chief data officers.
- 25:59They all want to figure this thing out.
- 26:00But none of them, as I could tell, had really gotten like,
- 26:03oh, we've implemented it in this way and now we've transformed
- 26:06our business across the company to use this.
- 26:08So they all want to do it. They don't really know how to.
- 26:11And I think the agent buzzword, as exciting of an idea as it is,
- 26:17I think we are far away from regular enterprises really
- 26:21implementing agents as sort of employee equivalents that are
- 26:25running around spending budget making decisions.
- 26:28And so that conversation, to me is sort of a signal of hype
- 26:31getting out of control. Yeah.
- 26:33As a daughter of as a daughter of a lawyer who works in this
- 26:37kind of space, I can just report back that a lot of these
- 26:39companies are a little bit risk averse in letting agents run
- 26:43rampant, even if they are in theory very excited about this.
- 26:46So I do think it's going to be kind of a multi year adoption
- 26:49curve on this one, but it could mean that the hype right now is
- 26:52a little bit too much. So far as I can tell, the
- 26:55biggest adopters this stuff is like the makers of the
- 26:57technology themselves who seem to be all about laying people
- 27:00off in order to like take advantage of the efficiency.
- 27:03Like Microsoft just laid off like another several 1000 people
- 27:07and it really seems like were. They middle managers or which?
- 27:09Which tier were they? Do you know?
- 27:12They had they had a middle manager layoff a couple months
- 27:14ago, but this one was mostly in the sales org.
- 27:16Do you know what percentage? Because I, I, one of my, you
- 27:18know, the other version of the newcomer podcast is the through
- 27:21Bull Valley podcast, which we've been recording with Max and
- 27:24James. And in that we have these
- 27:26predictions and I think one of mine is that people, a Fortune
- 27:30500 will attribute layoffs to AI.
- 27:34So I I'm I'm unfortunately for workers rooting I think for a
- 27:38mass AI attributed layoff in terms of my prediction becoming
- 27:42true. Just for just for the the Cal C
- 27:44bet. Just so I could say I was right,
- 27:47you know? Yeah, Microsoft was not explicit
- 27:50about it at all, but there were some interesting pieces that
- 27:52came out afterwards that like in the meeting after they laid off
- 27:55all their, you know, 50% of the sales team in certain
- 27:59departments, you know, the head of the unit, the CVP was like,
- 28:02by the way, you guys should really start using AI more.
- 28:05It can make your life a lot better.
- 28:06And so there, you know, I don't we're not quite like the quiet
- 28:09part out loud, but it seems like we're at least with those
- 28:12companies like leaning towards that they want to set an
- 28:14example. Here's how you lay people off
- 28:16when you when you use our products.
- 28:18Just mood check before we move on to the next subject.
- 28:21Are you bullish or bearish on AI over the next six months?
- 28:25Six months. I, I think there's wait.
- 28:28Oh, it's, it's, I mean, it's gonna continue to grow up to go
- 28:31up. There's no indication that
- 28:32things are going off a Cliff. All right, same.
- 28:35We're all bullish in six months, but maybe that means we're all
- 28:37blind. We're all bought into the hype.
- 28:39You know the the air comes out at some point.
- 28:42You got to do better than the ramp data for me to just decide
- 28:44to be bearish though. I agree.
- 28:46We're that's why this is an impressionistic certain signals
- 28:48to pay attention to. We're not saying this this thing
- 28:51is a bubble. All right, the last topic, an
- 28:54exciting one. Spicy.
- 28:56Complaining about tweets, Tom, you're all fired up about this.
- 29:00Somehow. You are OK with Mecca Hitler,
- 29:02but you're can't stomach Sequoia's Sean McGuire.
- 29:06I can't, I can't deal with Sequoia, his earnest.
- 29:08Political has political views. Sean McGuire is the guy I don't
- 29:14know personally. I have, you know, I've met many
- 29:16Sequoia partners. I've gone to the Sequoia spring
- 29:19mixers and fall mixers and there's quite a lot of people
- 29:21there that I like. I don't think Sean McGuire is
- 29:27oh, I'm sorry, let me, let me before we get into that.
- 29:29He has been going off on Twitter for about two years almost,
- 29:33mostly with his hardline right wing, pro Israel politics.
- 29:37He likes to make people mad. This is very exciting to him.
- 29:42He really got into hot water, so to speak, over the 4th of July
- 29:46weekend when he, after the story in the New York Times about
- 29:51Zaran Mamdani maybe being misleading on his Columbia
- 29:55application, Sean goes out there and says, you know, Mamdani is a
- 29:59liar because he's an Islamist. And this is culturally baked
- 30:02into what Islamists do. And then people got mad because
- 30:06it really sounded like he was saying all Muslims are liars,
- 30:09culturally speaking. And then he goes out and says
- 30:11that's not true. I wasn't talking about all
- 30:13Islamists. They're all, all Muslims talking
- 30:14about Islamists. And then that became a whole
- 30:16conversation about what it means to be and whether or not Sean is
- 30:19a racist. And I don't really know.
- 30:20And Bill Ackman and people have rallied around.
- 30:24Sean was offense. I saw Josh Wolf also rallying to
- 30:27his defense. Too Josh Wolf.
- 30:30There was a petition of people condemning Sean.
- 30:32There was a counter petition of people supporting Sean with
- 30:35signatures from all walks of this world.
- 30:38We we, we're not cancelling Sean McGuire that that is that is
- 30:43we're being very clear. And I think you you do a good
- 30:46job of pointing this out, Tom. But like, yeah, the one thing
- 30:50everybody in Silicon Valley can agree on is no more
- 30:52cancellations. And we're we're we're sticking
- 30:55to that Armistice. Like we're not trying.
- 30:57Nobody on this podcast is trying to cancel Sean McGuire.
- 31:02Yeah, the idea of someone losing their job because of their
- 31:04opinions in general, I'm, I'm generally against it.
- 31:07I think that's not a great thing for societies to have happen.
- 31:09I wouldn't advocate for it, I don't think, Sean.
- 31:14So OK, before I get into my personal opinion, I listen to a
- 31:17podcast interview Sean McGuire did with Jack Altman.
- 31:19Second time he's come up on this episode when which he talked
- 31:21about his decision to kind of come out as an opinion haver and
- 31:26talk about all of his right wing politics.
- 31:28And he was saying after October 7th, you know, and he Sean was
- 31:32pulling from his history. I think he worked for the DoD.
- 31:34He understands information warfare and the way that
- 31:37misinformation can be spread online.
- 31:39It was extremely important for him to start talking about his
- 31:43politics and he even said on this podcast on that, on that
- 31:46episode, I'd be willing to die for, you know, for the things
- 31:49that I believe and I don't advocate for that at all.
- 31:52That would be awful. I don't really think the things
- 31:54that Sean is saying are necessary to say.
- 31:58I don't really think. And he has a pretty large
- 32:00platform. He has 250,000 followers.
- 32:03He really thinks that he is putting important stuff into the
- 32:05discourse based on that podcast and the fact that he keeps going
- 32:08on about this stuff. And this is very common now
- 32:11within Silicon Valley. Oh, it's very common.
- 32:13There are certain VCs that have decided that their platform and
- 32:16their takes are so important that they want to build a whole
- 32:18media empire around it. And we've got, you know, the
- 32:21best version of that is the all in guys who have built an
- 32:23actually successful show with their takes on things.
- 32:26Mike Solana, who I also don't know, he's not really an
- 32:29investor, but he's in, you know, he's obviously in the VC world.
- 32:32He's got. CMO.
- 32:34Yeah, yeah. He's not AVC, but but you know,
- 32:36whatever he's he's in that world.
- 32:37He built Pirate wires, which is like actually an interesting
- 32:40publication. And I and I totally give him
- 32:42credit for all of that. You're basically saying, Sean,
- 32:45if you're gonna make us listen to you constantly, these, you
- 32:48need to be a little more creative.
- 32:49It can't just be like have. Something interesting to say.
- 32:52Like yes, have something. More interesting to say than
- 32:55just like what I. Find racially tinged things that
- 32:58get everybody into fury. It's sort of like the lowest
- 33:00common denominator way to get people.
- 33:02To right, It's a hack. It's a hack to get people to to
- 33:05to think that you're interesting because you're saying stuff
- 33:07that's kind of pushing the envelope in a racial way and,
- 33:10or, or whatever way and culture of a bigoted way.
- 33:12And I just think if it's that important to you, Sean, that you
- 33:15really want to like be flooding those of us who want to maybe
- 33:19see your takes on other things, you know, within the the tech
- 33:22world, talk about your politics, start up a sub stack, like let's
- 33:24just see how interesting your views really are.
- 33:27Like build up a little media empire around you and see if
- 33:29people actually care. Because he believes in all sorts
- 33:31of conspiracies too. He's like.
- 33:34Yeah, he does have his election denier stuff.
- 33:37I mean, I guess I could just be furthering your take, Tom, with
- 33:41this. I first of all, clearly it
- 33:43hasn't affected Sequoia's brand standing at all.
- 33:47So anyone who's concerned about cancellation, the attendant, you
- 33:50don't think so? I.
- 33:51Think it hurts the brand a little bit or like it's looms so
- 33:54large in terms of what Sequoia's brand is like.
- 33:56I think of Sequoia again, I don't wanna go after Sean or
- 34:01have them tell him to stop, but I do think just from a brand
- 34:04perspective, it's one of the loudest parts of their brand
- 34:07right now. But I think they're still really
- 34:09regarded as, yeah, I think they're still, I think the
- 34:12results is overwhelmingly supersedes this.
- 34:16Like I really don't think it's hurt at all.
- 34:18And I think if you're kind of in the ecosystem, especially where
- 34:21maybe you hold some of these views too, or you're like a
- 34:24little bit enjoying the kind of vocal trolley posture, which I
- 34:28will say is kind of a case for a lot of people in certain sectors
- 34:32and different industries. Like it could be a value add for
- 34:35you. You could be like signaling
- 34:36like, hey, we don't care. People think we just want the
- 34:38returns. And so I don't think it hurts
- 34:40them really at all. And so, yeah, let me.
- 34:44Make a couple quick points. I mean, 1, you know, being close
- 34:48to Elon and being in having somebody in this sort of
- 34:51conservative troll world gets you a certain contingent of
- 34:54people. And so if Sequoia can get away
- 34:56with, OK, we're the normie guys, but we also are friendly with
- 35:01Elon world, you know, that's, that's there's a financial play
- 35:04there. And I think that's part of
- 35:06what's going on. The other thing I just want to
- 35:08say on this conversation is that, like, kudos to Amjad, the
- 35:11CEO of Replit, I'll put it that way.
- 35:13I mean, he's been very bold in, you know, voicing unpopular
- 35:20opinions. And I think Sean represents
- 35:22himself as like articulating this sort of like, I don't know,
- 35:25out of consensus point of view. Well, but Amjad is an
- 35:30interesting case here because, you know, he went on Joe Rogan a
- 35:33couple of weeks ago, talked about all, all kinds of things.
- 35:36But, you know, within that he talked about, you know, his
- 35:38being of Palestinian descent and he's Jordanian.
- 35:42And he was like, yeah, it's not easy being, you know, one of the
- 35:45few people in Silicon Valley willing to speak up, you know,
- 35:47speak out against Israel and and for the Palestinian.
- 35:51Yeah, right. Paul Graham is an interesting
- 35:52case and all of that, but it feels more personal with Amjad.
- 35:57And he's gotten into it a little bit with Sean on X.
- 36:01And Sean completely took out of context something that Amjad had
- 36:04said and cut off the date and like a Washington Post article
- 36:07so that he could claim that what he had said on Rogan was
- 36:09inaccurate. And you know, that's standard
- 36:11trollish behavior on on she's part like he was trying to get a
- 36:14rise out of people. I don't even think I'm John.
- 36:16Is that honestly, like, you know, the politics that he was
- 36:18describing on Rogan doesn't even really conform with like the
- 36:21most leftist stuff. He was talking about two state
- 36:23solutions. He was, you know, like I I care
- 36:27a lot about what happens to the Israeli people.
- 36:29I mean, that kind of stuff would not actually, you know, get
- 36:33endear you too much to the left. Sure.
- 36:35But the stuff you're hearing on tech Twitter right now is like,
- 36:39just like, let's send every weapon we have over to Israel
- 36:42and like, let let the chips fall.
- 36:44They may. Sure, Joe Lonsdale is probably
- 36:46the extreme version of that. And and like, I don't know, I
- 36:48should be interested in your take, Eric, but like Sequoia,
- 36:51you know, with Doug Leone as the head of the firm also like
- 36:54fairly well known conservative like as the firm, they certainly
- 36:58if you want to look at the personality.
- 37:00They had Doug Leone and they had Mike Moritz, both really strong
- 37:03investors, opposite sides of the political spectrum.
- 37:07Clearly Doug Leone won. I mean, Moritz isn't even part
- 37:10of the firm anymore. And it seems like that ideology
- 37:14is 1. I mean, I, I Rudolf doesn't talk
- 37:16about his politics a ton, but my sense is he's on the more
- 37:19conservative side of things. And so yeah, I don't, I don't
- 37:23see who the beacon of liberal or left wing thought of, at least
- 37:28publicly as Sequoia is right now.
- 37:30And does do you think Doug being there gives Sean a lot more
- 37:33cover to be able to to say the stuff that he does?
- 37:36I mean, my first piece for the newsletter was about, you know,
- 37:39him being a Trump watching conservative.
- 37:41It was about China and about how this what the circle the center
- 37:46could not hold with Sequoia having China while having anti
- 37:50China sort of anyway, I was I was right that China spun out.
- 37:53But anyway, that was my first piece and it, it talked about
- 37:57how Doug Leone was a big Trumper.
- 38:00And so, yeah, I, I mean, Doug's not in charge anymore, but he's
- 38:03still around the firm. And I think that the fact that
- 38:07if if Sean was so out of step with the rest of Sequoia, it
- 38:11wouldn't last. Yeah, but he also kind of wants
- 38:14to be maybe fired a little bit, a little bit like, I mean, look,
- 38:18he I don't know, I, I, I mean, on that podcast he talked about
- 38:22the fact that, you know, he's independently wealthy now
- 38:24because he had sold the company earlier.
- 38:26So he has a lot of money. So he's able to, you know, take
- 38:30these risks, these risks with his, with his opinions.
- 38:33And then, you know, he changed his Twitter bio or his ex bio
- 38:36for a bit to say like Sequoia partner in parentheses for now.
- 38:39Like he's he's certainly baiting people to to and, you know, he
- 38:43talked about how the the petition, you know, was was
- 38:47another example of cancel culture.
- 38:49And these people love to be, you know, victims before the fact.
- 38:54So yeah, I don't know if like I don't, I don't even talk to
- 38:57Sean. I don't know.
- 38:58Why? It would it would play into his
- 39:00martyr. This Islam is narrative Muslim
- 39:03distinction like why? Why, why why say something
- 39:08that's so easily misinterpreted? I guess is my my issue here.
- 39:12Because he's a troll, right? And I hate that kind of
- 39:15communication. I just like I, at my core, want
- 39:18to speak sort of earnestly and openly and say be heard as I'm
- 39:24what I'm intending to say, you know?
- 39:26And I feel like he's trying to hack engagement.
- 39:29And that is what I object to. Right, right.
- 39:33No. And that was sort of the point
- 39:34of my column is, is I don't even really deal with the specifics
- 39:37of what he's saying and interrogating the logic.
- 39:39I mean, that's a whole other thing.
- 39:40But like, he is kind of claiming that he is on this holy war to
- 39:45put his opinions out there because it's extremely
- 39:47necessary. And in fact, I don't think these
- 39:49opinions are that surprising. I don't think they're all that
- 39:51necessary. And he's hacking the system
- 39:53anyway to get them out there. Like he's not really saying.
- 39:56And he's friends with the guy who runs the platform, the guy
- 39:58who's shifting his own AI to bring this full circle to sound.
- 40:02More like him, you know? It's like, oh, I could be one of
- 40:04the voices that says the things that Elon wants me to say, you
- 40:08know? Yeah, there, there are certain
- 40:09people there, there are certain people who I think have been
- 40:12elevated by the Elon, you know, Elon era Twitter and the 4U page
- 40:17and the algorithm and it's like Sean McGuire is one of them.
- 40:20So also is the die work wear guy but but you know different
- 40:24different sides of the. I mean, one of the most iconic
- 40:26things that has been said on this podcast is Ben Smith saying
- 40:30we're all in the courts of billionaires.
- 40:32And I think that theme has continued to play out.
- 40:35And yeah, Sean Mcguire's in the court of Elon Musk.
- 40:39Right. And, and, and benefited from it
- 40:41in lots of ways financially and also like in this little mini
- 40:45unimpressive media thing that he's put together.
- 40:48But we are not cancelling you, Sean.
- 40:50Keep tweeting. Godspeed.
- 40:51Just make him less hate, more more thought.
- 40:55Well, he's not hateful only towards Islamofascists or
- 40:58Islamists those who can hate which.
- 41:00Which it is offensive. It is offensive to say that
- 41:04Mamdani is an Islamist. Like there, there's nothing
- 41:07there's. It's like he's a normal mayoral
- 41:10candidate. Like I feel like we can't let
- 41:12that part of it slip by. Like it's just like.
- 41:14That point like it is just blatantly false.
- 41:16Also, it's very it's, it's just, I don't want to let that slide.
- 41:20Have you done the research though, Madeleine?
- 41:22Have you done done the research? I'm familiar with who his father
- 41:26is, is a professor at Columbia. I'm, I will just say I don't
- 41:29think that Zoran Mamdani is an Islamist and there's no evidence
- 41:34to that to be supported and to say that is inherently
- 41:37inflammatory. Correct me if I'm wrong, by the
- 41:39way. As the non New Yorker over as
- 41:42the as the non New Yorker on this pod, as the non New Yorker
- 41:45on this podcast, He didn't even really run on like he didn't run
- 41:48like an identitarian campaign. Right now that's all the right
- 41:53wing effort now is to basically say you were misled like this
- 41:56happy go lucky campaign that you saw.
- 41:59That's not who he really is. It's all bullshit.
- 42:02I'm sorry he ran a campaign on affordability.
- 42:04But anyway. I feel like the substance of it
- 42:07matters, you know, not just the optics.
- 42:08Anyway, that's our episode. We'll see you next week.
- 42:12Thanks for listening. See you next week with Sean
- 42:13McGuire Co hosting and I'm. Sure.
- 42:16There you put that on to the world.
- 42:18There's some percentage. Is Brock still too woke next
- 42:21week? All right, see ya.
- 42:24Bye.