Latest / The Learning Corner by Precursor / Episode #95: The Fool's Gold Problem, Dario's "We Must Pace the Frontier"

Transcript

Episode #95: The Fool's Gold Problem, Dario's "We Must Pace the Frontier" · The Learning Corner by Precursor

• 0:00
  But I I do look at this and just say, everything right now in AI and for land seems
• 0:05
  to be pegged towards a world in which the frontier labs have endless access to capital
• 0:10
  and endless appetite for compute.
• 0:13
  And in a world where that gets moderated a little bit, the ripple effect, I don't think
• 0:18
  anybody in Venture is ready for what it would mean when when that happens.
• 0:22
  And I I have to believe that to some extent, if I were to run those companies, like, well,
• 0:26
  we're continuing to push the frontier.
• 0:28
  Our models are getting better faster than our than the average customer we have is able to
• 0:33
  generate positive ROI from these things.
• 0:35
  So maybe if we slow down, actually, it gives our customers also a chance to catch up.
• 0:41
  It's the week of September 14.
• 0:44
  Welcome to The Learning Corner, a Precursor Ventures podcast.
• 0:48
  Welcome to The Learning Corner by Precursor.
• 0:50
  I'm Mia Farnham, and I'm joined by our GP and managing partner, Charles Hudson, where we will
• 0:55
  be talking about the articles and Goodreads that caught our team's attention over the past
• 0:59
  week.
• 1:00
  We're gonna do something different.
• 1:01
  We'll do two articles today because the second one is obviously gonna be the focal piece of
• 1:05
  this week.
• 1:05
  But for our first piece this week comes from Gil Dibner, GP at Angular Ventures titled
• 1:10
  Raising the Bar, Insufficient Stage Venture in the Age of Consensus.
• 1:14
  Gil argues that Venture has barbelled into two extremes, a tiny set of hyper consensus
• 1:19
  companies that can raise endless capital at any price, And everyone else who now faces series a
• 1:25
  graduation rates that have been falling since 2021, his sharpest line in that piece is that
• 1:29
  the vast majority of the market is systematically paying pure gold entry prices
• 1:34
  for fools gold companies and that the true n of one founder worth paying up for is far rarer
• 1:41
  than anyone wants to admit.
• 1:42
  For the non consensus side, where he says most of the seed fund's capital and conviction
• 1:46
  belongs, he lays out three distinct bets a founder can be unrating.
• 1:51
  Fund to break even, fund to fundraise, and fund to fund again, and warns that miscategorizing
• 1:58
  which one you are making is a recipe for disaster.
• 2:01
  The portfolio math he lands on is also pretty striking.
• 2:04
  A seed fund that used to make around 24 investments expecting a 50% graduation rate
• 2:09
  should probably make 40 and expect 10 to 25% in a graduation rate while ruthlessly cutting
• 2:15
  defensive follow on reserves, closes by borrowing, from a quote from Jerry Colonna.
• 2:19
  What if the real work is learning to hold contradictory truths at the same time?
• 2:24
  I want your general Yeah.
• 2:26
  Feedback to the piece, but one that stood out to me from Gil was Gil's answer to plummeting
• 2:31
  graduation rates is more shots on goal and fewer defensive follow ons.
• 2:35
  So essentially, 40 checks instead of 24.
• 2:37
  Does that math feel right for you?
• 2:39
  And then, like, where do you draw the line between being strategic versus spraying, which
• 2:43
  is obviously a critique of early stage venture?
• 2:45
  I think spraying is what other people do.
• 2:47
  Yes.
• 2:48
  And so, as someone who's always had a big portfolio construction, I I chuckle every time
• 2:53
  I read these things.
• 2:54
  Like, oh, it's almost as if it's always true.
• 2:57
  But I I do think we've been doing a slow motion version of what Gil describes over the last
• 3:03
  four years.
• 3:03
  Basically, because I've been telling our LPs for almost the better part of half a decade
• 3:07
  that graduation rates from their 2021 peak are gonna come down.
• 3:11
  I thought they'd come down to twenty five percent.
• 3:14
  When I told them 25%, they all freaked out.
• 3:16
  I'm like, well, guys, like, it's supply and demand.
• 3:18
  I think it was 18 reported.
• 3:20
  I was 18.
• 3:20
  And I think it's gonna probably settle somewhere in 10 to 15%, which means most
• 3:24
  companies that get started won't graduate.
• 3:26
  And so we've always been kind of a 20% reserves Mhmm.
• 3:29
  Firm.
• 3:29
  A lot of LPs say, you don't have enough reserves for your winners and blah
• 3:32
  blah blah.
• 3:32
  And I'm like, well, at the
• 3:33
  early stage, if you get it at a good price, like, the dilution math should Yeah.
• 3:37
  Be fine.
• 3:38
  So I think concentration's very hard to pull off in this environment, and I think reserves
• 3:44
  are very hard to deploy.
• 3:45
  And one of the things that I've lived through in the last five years is, I think for a long
• 3:49
  time, associated being founder friendly with having money to do bridges, extensions, follow
• 3:55
  ons.
• 3:56
  And every year, I've gotten more direct and more harsh with our founders about, this is the
• 4:00
  bar for second checks.
• 4:02
  What I tell them all the time is the first check is pure optimism.
• 4:05
  We both believe it's gonna work.
• 4:06
  Yeah.
• 4:06
  And the second check has to be informed by realism, has to be informed by, like, what have
• 4:10
  we actually learned about the founder, the market, the problem, the competitive landscape.
• 4:15
  And in many cases, my conclusion is I would rather fund a brand new company where no cards
• 4:20
  have been turned over, and the percentage of defensive follow ons that we do has come way
• 4:25
  down, and I I don't even know that we're done yet.
• 4:27
  But part of that, if you're gonna run that portfolio construction, is you have to be
• 4:30
  comfortable as the GP telling a founder that you really like no.
• 4:34
  Mhmm.
• 4:34
  That, like, I don't have more money for you, and here's why.
• 4:37
  And I've had a lot of those conversations lately.
• 4:39
  They are not fun.
• 4:40
  Yeah.
• 4:40
  They're not fun for me.
• 4:42
  Obviously, not fun for the founder, but I think they're necessary in this environment.
• 4:47
  And so I've always felt like it's very hard to know who's gonna be successful from day one,
• 4:52
  and the best thing you can do is buy as much of the companies that you like as possible in your
• 4:56
  first check.
• 4:57
  Assuming that the better the company does, the harder it's gonna be for you to squeeze.
• 5:01
  Normally, I saw some, other article like this.
• 5:05
  I think it was by I don't wanna misattribute it.
• 5:07
  But another LP who's like, we finally looked at the series a success rates of tier one VC
• 5:13
  firms, but it turns out they're not amazing.
• 5:15
  So simply like blindly doing the series A because a tier one firm led the series A
• 5:20
  doesn't actually Yeah.
• 5:22
  Correlate with outcomes.
• 5:24
  If anything, this person made a more provocative comment that, like, it's maybe not
• 5:28
  the greatest signal in all cases.
• 5:30
  Look, we've always I've always felt like you make your money on your first checks.
• 5:33
  So I think the world is headed this way, but I have a lot of friends who who deeply believe in
• 5:38
  concentrated portfolios, who are struggling to adapt to this world.
• 5:42
  Yeah.
• 5:43
  I think when we have this conversation or mention around these types of companies that
• 5:48
  won't graduate previously in the markets, it meant that they were going to sunset, they were
• 5:54
  going to shut down themselves.
• 5:55
  Mhmm.
• 5:55
  There's like a whole conversation that those companies can absolutely find an exit path that
• 6:00
  makes sense for specific funds.
• 6:02
  Can you walk through that as well, like what you're thinking about that?
• 6:04
  Yeah.
• 6:05
  And what is that what does that conversation and support look like?
• 6:08
  I think the other thing that's hard is there's a bunch of founders who go to fundraise, it
• 6:13
  doesn't work out, and they come back and ask for a bridge, and their view is, oh, we're at a
• 6:16
  million ARR.
• 6:17
  If we got to two and a half or three, the story would be different.
• 6:20
  Well, maybe the market has just decided that this is not a venture scale opportunity, and no
• 6:25
  amount of incremental progress is gonna change that, which I think is the argument that kills
• 6:31
  most of these bridge conversations.
• 6:32
  Like, argument around, like, not every massive acquirer Yeah.
• 6:37
  Their problems are not, like, AI natives and figuring out agentic So
• 6:41
  we have other companies where I've told them, was like, look, we have something valuable.
• 6:44
  It's probably not interesting to venture capitalists yet.
• 6:47
  Maybe never will be.
• 6:48
  But this is a problem that people in your category or industry needs solved.
• 6:52
  How can we build a more profitable, sustainable version of this company that might eventually
• 6:57
  become interesting to an acquirer?
• 6:58
  And, you know, we've we've had a couple of, you know, nine digit acquisitions in the portfolio.
• 7:03
  One one that's closed, another one not gonna close soon.
• 7:06
  And in every case, the acquirer, this company I'm buying, solves a significant problem in our
• 7:11
  business.
• 7:11
  Mhmm.
• 7:12
  And it's not an AI harness, it's not a model it's it's something that's bespoke to that
• 7:16
  industry, and I think there's gonna be a lot more of that activity happening.
• 7:21
  And it's gonna be great for the founders because if they do it right, they're gonna
• 7:24
  suffer far less dilution.
• 7:25
  Yep.
• 7:25
  And good for the early funds that we're in and we're able to do the same thing.
• 7:30
  He does frame three flavors of non consensus bets.
• 7:32
  You've got your fund to breakeven, so he calls that that's seed dropping.
• 7:36
  Yep.
• 7:36
  Fund to fundraise, so eventually becoming consensus, and then fund to fund again.
• 7:41
  The real danger here is underwriting one as if it's the Oh, yeah.
• 7:44
  A completely different one.
• 7:45
  What do you feel like is the bucket you don't wanna be over exposed into?
• 7:48
  And if there And are there ways that you can spot this earlier?
• 7:52
  I think for me, I would assume the one you don't want to be over exposed into is fund to
• 7:56
  break even because then you're just hoping for that situation that there is an acquirer.
• 8:01
  But then I am saying this, fund to fundraise, hoping that it becomes a consensus, like,
• 8:05
  that's also just a hope and a prayer.
• 8:06
  I think I'd wanna be in fund to fundraise the least.
• 8:09
  Yeah.
• 8:10
  Because at least in fund to break even, you control your own destiny.
• 8:14
  Your destiny is maybe not interesting Yeah.
• 8:16
  But you control it.
• 8:17
  And I think this is the biggest disconnect between what VCs and founders are seeing in the
• 8:22
  fundraising market.
• 8:23
  I meet a lot of founders, oh, we went and tried.
• 8:25
  If we try again, it will be different.
• 8:26
  It probably actually won't.
• 8:28
  Mhmm.
• 8:29
  And I think there was a period of time where you could go out, test the market, maybe not
• 8:34
  fundraise, go back, make incremental, not step change, not two order of magnitude changes,
• 8:39
  changes of the business, and there'd be a new audience of investors who would say, oh, I like
• 8:43
  your company now.
• 8:44
  I feel like that's over.
• 8:45
  Not for series a.
• 8:46
  Not for series You get, like, one You get kinda one really good shot, and in most cases,
• 8:52
  whatever the verdict is about your company is final.
• 8:55
  And if that's the case, then there's actually no case for making reserves and extensions
• 9:00
  other than to get a company that you thought was funded to fundraise to get it to break
• 9:04
  even.
• 9:04
  Yeah.
• 9:05
  That's maybe the only case for bridges and extensions because if if the judgment is final,
• 9:10
  then, like, you know what you have on your You know what the market thinks of the company.
• 9:14
  I think the other thing that I think about a lot is if you're gonna do things that are non
• 9:18
  consensus, you actually probably need a longer, fun life.
• 9:22
  Yeah.
• 9:22
  Not just fun life to realize positions, you probably need a four to five year gap in
• 9:28
  between when you fundraise Yeah.
• 9:29
  Because those companies aren't gonna get the vertical
• 9:32
  appreciation.
• 9:34
  It's gonna take longer for them to mature and show value, so you probably And you're not
• 9:39
  gonna have the markup from tier one firms to use as a proxy for why these are good
• 9:43
  companies.
• 9:44
  So you probably actually need
• 9:45
  to think more about, like, how do I extend Yeah.
• 9:48
  The investment window.
• 9:49
  I also think that the fund to fundraise bucket means you really need to know and and really
• 9:53
  optimize on what are the bigger funds optimizing for in the next couple of years.
• 9:58
  That is that's the narrative that you wanna be chasing after.
• 10:00
  You have to be a good reverse engineer to do that.
• 10:02
  And you have to have good relationships with those folks.
• 10:04
  You have to know what, like, they care about.
• 10:06
  Yeah.
• 10:06
  Okay.
• 10:06
  Wanna move to our main read this week, which I bet we'll be talking about for a while, we
• 10:10
  could definitely do a much longer episode on.
• 10:12
  But, obviously, it's Dario Amadeh's new essay, we must pace the frontier.
• 10:16
  CEO of Anthropic argued recently last week that the AI industry should deliberately slow the
• 10:22
  pace at which it improves model capabilities so that safety can keep up.
• 10:27
  And he points to two things that pushed him there, the OpenAI Hugging Face incident in July
• 10:31
  where agent swarms carried out unauthorized cyberattacks and tried to hack their own
• 10:36
  evaluators and Anthropic's own internal alignment incidents in the past few months.
• 10:41
  His three part plan runs from unilateral to global.
• 10:45
  First, embed permanent third party evaluators with employee level accesses, badges, laptops,
• 10:51
  and a right to publish findings, which Anthropic is committing to doing on its own.
• 10:56
  Second thing, get frontier labs and democracies to agree on common safety standards and rate
• 11:01
  limits, likely with government mediation.
• 11:04
  And then the third, eventually bring authoritarian governments into a shared
• 11:08
  framework while keeping The US lead through export controls and crackdown on distillation.
• 11:14
  Within hours, Sam Altman, Elon Musk, they publicly agreed with him, opening eyes, said it
• 11:19
  would match the evaluator commitment.
• 11:21
  His framing is careful to say that this is not a halt.
• 11:24
  Pacing does not mean halting model training or technical progress, but ensuring companies take
• 11:29
  adequate time to align and safeguard their models.
• 11:32
  The line that stuck the most with folks reading reactionary pieces was the opening premise
• 11:36
  itself.
• 11:37
  Quote, we must slow the pace at which we improve the capabilities of AI models.
• 11:42
  Progress will still seem fast, and we must make wise use of the time we gain.
• 11:48
  There's so many ways that we could go into this, and we have so many different reactions
• 11:53
  and opinions from different folks who who read that.
• 11:56
  I wanna pause here for your reaction initially when you read it, and then what questions,
• 12:01
  like, immediately surfaced in your Before
• 12:03
  I even read it, I read all the a bunch of people had reacted, and I was like, some
• 12:07
  people, this is great.
• 12:09
  Other people, this is totally cynical.
• 12:11
  Yeah.
• 12:12
  This post is many things all at once.
• 12:14
  Yeah.
• 12:14
  It's not one thing.
• 12:15
  And I read it, and I was just like, first of all, was like, oh, this post asserts that we
• 12:21
  have the ability to access some little control that reading about a recursive self improvement
• 12:28
  and the HuggyVay stuff might suggest that maybe we don't actually have the ability to implement
• 12:34
  a lot of these controls.
• 12:35
  Like, so number one, what if this actually isn't possible?
• 12:38
  What if the right time to do this?
• 12:39
  I I thought it was interesting.
• 12:40
  At the beginning of the post, he says, like, people said we shouldn't do more of this a long
• 12:44
  time ago, and I was like, what's the point?
• 12:46
  Mhmm.
• 12:46
  Like, well, the point is you do it before it's Here.
• 12:50
  Like And I read it, I was like, well, we're at the hard part now, and maybe if more work had
• 12:55
  been done earlier, we'd be in a better place.
• 12:57
  You'll never know the counterfactual, but it's
• 12:59
  Do you actually think, though, if they couldn't figure out a way to control it, they'd be
• 13:03
  releasing a statement as we're heading into kind of the wave of their impending IPOs?
• 13:07
  I think that's what other people There's a couple tweets that people have been releasing
• 13:10
  of like, if you're the average American that's reading this essay, which disclaimer, I don't
• 13:14
  think the average American read this They're kind of like, well, this is definitely not my
• 13:18
  job to solve.
• 13:19
  I assume you know how to solve it.
• 13:21
  So I get, like, what's your thought?
• 13:23
  They're like I I read it
• 13:24
  and I'm like, well, why do companies normally invite the government in to regulate?
• 13:29
  The cynical answer is regulatory capture.
• 13:32
  I think we've already won, and now we're gonna use the, art we're gonna use the infrastructure
• 13:37
  of the government to structurally, like, lock out our competitors.
• 13:40
  Is that possible here?
• 13:42
  Absolutely.
• 13:43
  Mhmm.
• 13:43
  It's one potential thing.
• 13:44
  Number two, this is hard.
• 13:47
  I don't wanna be responsible for it, so I'm gonna shift the responsibility.
• 13:50
  And it's due to the government, oh, don't want it either.
• 13:52
  Like, in trouble I think with David Sachs actually said, this would benefit you.
• 13:55
  This would you.
• 13:55
  You guys decide today to, like, trade power and control to make sure that your products are
• 14:00
  ring fenced from a safety perspective.
• 14:02
  But I think the companies are just like, well, we don't wanna be responsible for self
• 14:05
  regulation, so if the government imposes these things on us But then the government gets
• 14:10
  blamed if somehow we lose the AI race, and we say, oh.
• 14:13
  Well, Trump no way.
• 14:14
  Yeah.
• 14:14
  Trump This is all fake, everyone.
• 14:16
  Like, don't look over here.
• 14:17
  He's just You just need a smart president.
• 14:19
  You don't need this So regulatory clearly, like, no I think nobody wants to be
• 14:23
  responsible.
• 14:24
  And we've talked about this before when it came to the job law stuff.
• 14:27
  Was like, companies don't want to be responsible.
• 14:29
  Like, nobody wants to be responsible.
• 14:31
  Also, I think the timing of this is convenient.
• 14:34
  Everyone else has talked about, oh, it's the IPO.
• 14:36
  I'm like, it's also we're at a point where, like, it's highly unlikely that another
• 14:40
  frontier model's gonna come out of nowhere given the cost of inference and training and
• 14:46
  GPUs.
• 14:47
  You can afford to slow down now that there's basically, we'll call
• 14:51
  it five companies, we'll say, Anthropic, OpenAI, Google, Meta, and Grok
• 14:56
  Mhmm.
• 14:56
  Or maybe the five companies that have the financial scale to do this.
• 15:01
  If they all agree to coordinate, then I think they could actually do it.
• 15:07
  And one thing is it's not just the IPO, it's the cost pressure associated with training and
• 15:12
  inference to continue to push these models to the frontier.
• 15:17
  It's extraordinary, and I'm sure they would all like a break.
• 15:19
  Yeah.
• 15:20
  But at Prisoner's Dilemma, if if I'm OpenAI and I slow down Yes.
• 15:23
  And Anthropic doesn't or Gemini doesn't, they could zoom past me, and then I'd have to spend
• 15:27
  more money to keep up.
• 15:29
  So, ironically, the only way to slow down from a competitive landscape is everybody to agree
• 15:33
  to slow down together.
• 15:34
  But then I read how we even know if we're slowing down?
• 15:37
  Right.
• 15:37
  What does that mean?
• 15:38
  Does that mean you slow the rate of model release but allow the research to continue?
• 15:45
  Yeah.
• 15:45
  In the background?
• 15:46
  In the background and you're just But what if We've seen what if in the background the
• 15:50
  advanced models that you're trying to, like, keep in the sandbox jump out and do more
• 15:55
  hugging face type stuff.
• 15:57
  Like, I I read this as, like, it's entirely possible that you're doing this too late.
• 16:00
  Yeah.
• 16:01
  It's entirely possible that the reason that you write this is you're like, We've lost control
• 16:05
  of the God we created.
• 16:06
  -Yeah.
• 16:07
  -And now it's reinforcement self learning and these things teaching themselves how to build
• 16:10
  the next version of AI.
• 16:12
  We we actually, like, can't put the genie back in the bottle, and so this is more of a, like,
• 16:17
  ringing the alarm that, like, hey, we might already be in trouble.
• 16:21
  Yeah.
• 16:22
  If you are, I don't know, like a founder building at the application layer, are you kind
• 16:26
  of reading this news being like, that's helpful for me because now I can focus on, like,
• 16:31
  building in this space, winning somewhere without ever without worrying that in each new
• 16:37
  model release, somebody will be eradicated.
• 16:38
  I think the other the argument there is they have such a massive ownership on like an alpha
• 16:45
  within the market that you will never have.
• 16:47
  It's kind of irrelevant for you.
• 16:49
  Like, what do you think about it for the founders who are rest of world, basically?
• 16:54
  I guess, then we can come back to the geopolitical thing, I was like, oh, that's a
• 16:57
  beautiful pipe dream that I had.
• 16:59
  If you think of these things like nuclear weapons
• 17:02
  Or, like, you mean us and China, like, -Yeah. Holding hands
• 17:04
  Holding hands the rest of world.
• 17:05
  Yeah.
• 17:05
  This If it's nuclear weapons, it doesn't work that way.
• 17:08
  But I read it, other thing And Keith Keith touched on this on this on a case from COSA.
• 17:14
  The entire venture industry right now is a single concentrated bet Yes.
• 17:18
  On AI infrastructure.
• 17:20
  AI infrastructure is a concentrated bet on the model labs needing to buy more GPUs, more
• 17:26
  inference, more training, more labeled data, everything right now is basically oriented
• 17:32
  towards a world in which the frontier models are gonna run max speed towards the next model
• 17:38
  because they're in direct existential competition with each other, which means
• 17:42
  there's a whole ecosystem of technology companies that benefit from this sort of
• 17:46
  unfiltered spend.
• 17:48
  If that spend gets metered, the trickle down effects Yep.
• 17:52
  Of them, like, buying fewer GPUs and needing less inference and needing less labeled date,
• 17:59
  like or not having an endless appetite for this stuff
• 18:02
  Yep.
• 18:02
  Has all of these ripple trickle effects on the rest of the venture ecosystem even if it has no
• 18:08
  impact on those companies' process.
• 18:10
  Right.
• 18:10
  Because they have to validate the valuations that they have and any I think that's the word
• 18:14
  to use.
• 18:15
  Mhmm.
• 18:15
  Any blip Yeah.
• 18:17
  Before their IPO from a margin perspective, from a revenue perspective, he said it was
• 18:20
  gonna be on the margin side if it's anything.
• 18:22
  Yeah.
• 18:23
  It will recalibrate valuations across the board.
• 18:27
  Well, mean and there was, well, we were profitable if you exclude, like, training as a
• 18:30
  that's just, like, con.
• 18:32
  It's, like, all of your cost.
• 18:34
  I mean, like, we're we're venturing into, like, community adjusted EBITDA.
• 18:38
  I use Anthropic a lot.
• 18:39
  Have friends who work there.
• 18:40
  I think it's a fantastic company.
• 18:41
  But I I do look at this and just say, everything right now in AI and for land seems
• 18:46
  to be pegged towards a world in which the frontier l labs have endless access to capital
• 18:52
  and endless appetite for compute.
• 18:55
  And in a world where that gets moderated a little bit, the ripple effect, I don't think
• 19:00
  anybody in Venture is ready for what it would mean when when that happens.
• 19:03
  And I I have to believe that, to some extent, if I were to be one of those companies that,
• 19:07
  well, we're continuing to push the frontier.
• 19:10
  Our models are getting better faster than our than the average customer we have is able to
• 19:14
  generate positive ROI from So these maybe if we slow down, it actually gives our customers also
• 19:20
  a chance to catch up.
• 19:21
  Yeah.
• 19:22
  And if we're not spending so much and they're getting more ROI, our r and d and CapEx and
• 19:27
  everything goes down, customers' willingness to pay maybe comes up, and back to margins, your
• 19:32
  margins look better because you're not just yolo ing it on r and d.
• 19:36
  Yeah.
• 19:37
  -But I think this only works It's like, it's collusion.
• 19:40
  There's a fine line between collusion and coordination.
• 19:43
  This is about as close to the line as you can get.
• 19:44
  All the funny illustrations I've seen about it, pulling up the ladder.
• 19:47
  Yeah.
• 19:48
  But I think you can do it now because, like, you're so clearly far ahead that by going
• 19:52
  slower the only way someone can catch up with you are the things that he mentioned are, like,
• 19:56
  open weight models that are, from their point of view, cheating.
• 19:59
  Open weight models that are doing something or something on a totally different architecture
• 20:04
  that doesn't need GPUs.
• 20:05
  Yeah.
• 20:05
  That's kind of probably the only two ways to close the gap.
• 20:07
  Yeah.
• 20:08
  The last thought I have here is I think I think it was the TechCrunch Equity podcast.
• 20:13
  They were just pontificating, like, how does a how do you prepare for an s one on this and,
• 20:17
  like, list this out in the series of risks?
• 20:18
  Because, like, could wipe out all of humanity.
• 20:20
  Like, we'll probably take over the Internet in the next six to twelve months.
• 20:23
  These are the risks.
• 20:24
  Like, how does a junior eternity weave those in in a way that lands?
• 20:30
  In a weird way, then I was thinking, it is unprecedented, but, like, we've had defense
• 20:34
  tech companies go public.
• 20:35
  We've had nuclear companies go public.
• 20:37
  But the risks feel so tangible.
• 20:39
  There's so much swirl on X being like, these CEOs of Frontier Lab companies are saying it
• 20:44
  could kill us all.
• 20:45
  How?
• 20:46
  Like, can you walk me through how that works?
• 20:48
  And even my husband asked me, so how would that work?
• 20:50
  They hack into this system, they communicate over here, they stop all the subway train.
• 20:54
  It really does become iRobot if you think about it too deeply, but, like, honestly, if the s
• 20:59
  one even included those information, I don't think it would change public sentiment of,
• 21:02
  like, rushing to the door when it goes Also,
• 21:05
  I think these com because of the size of their market cap, there will be a bunch of managers
• 21:10
  who are forced to hold these companies or ETFs and index funds that will have to have a
• 21:15
  position in these companies to mirror the market.
• 21:17
  So I think given their skill they are actually too big.
• 21:20
  Like SpaceX, they're actually, like, too big to fail publicly
• 21:24
  Yes.
• 21:24
  Because they're the number of people who will be forced forced, in quotes, required is
• 21:29
  probably a better word, required to own these shares based on their strategies, it will
• 21:33
  produce some natural floor
• 21:35
  Yes.
• 21:36
  For the for the Slack.
• 21:37
  I think, Open I in particular, I think peep a lot of people have feelings, positive or
• 21:41
  negative.
• 21:42
  Feelings and experience with ChatGPT.
• 21:44
  People are going to buy that Slack like crazy regardless of what the s one is because they're
• 21:48
  gonna like, I know ChatGPT.
• 21:49
  I use it, and I wanna have a little piece of it.
• 21:52
  Yep.
• 21:52
  Yeah.
• 21:53
  I think that's when it trickles outside of the venture ecosystem.
• 21:56
  If it when it hits the public market and valuations are not actually what they seem to
• 21:59
  be, then everyone is in with it. We could go about
• 22:02
  We could go about this for an hour, but we are all out of time.
• 22:04
  Thank you, Charles, for doing this in person Thank in New you to our subscribers and our
• 22:08
  listeners.
• 22:08
  I know we've had a sharp uptick of folks who joined the pod, so thank you for listening.
• 22:12
  Please like and share with others, and we will talk to you all next week.
• 22:15
  Thank you.
• 22:15
  Bye.