Latest / The Learning Corner by Precursor / Episode #70: The VC Returns Suck Narrative, Venture's Nifty Fifty Moment, The Four Approaches to Early Stage Venture

Transcript

Episode #70: The VC Returns Suck Narrative, Venture's Nifty Fifty Moment, The Four Approaches to Early Stage Venture · The Learning Corner by Precursor

• 0:00
  If the AI trade blows up and people lose a lot of money, I could name at least five or six
• 0:04
  firms that I'm sure will still be in business despite losing tons of money.
• 0:08
  And if that's the case, maybe the incentive is to just do the hot thing.
• 0:12
  Maybe that's actually, unfortunately, the system incentive, which makes me really
• 0:16
  uncomfortable to say it.
• 0:17
  But it is possible that there is not this corrective force that says if you underperform
• 0:22
  or if you make a blunder, that you'll have to pay.
• 0:25
  It's the week of March 9.
• 0:27
  Welcome to the Corner, a Precursor Ventures podcast.
• 0:34
  Welcome to the Learning Corner by Precursor.
• 0:36
  I'm Mia Farnam, and I'm joined by RGP and managing partner, Charles Hudson, where we will
• 0:41
  be talking through the articles and Goodreads that caught our team's attention over the past
• 0:45
  week.
• 0:46
  For our first topic of the week, Micah Rosenblum, managing partner at Founder
• 0:50
  Collective, shared a piece on LinkedIn titled the VC returns suck narrative and how we all
• 0:56
  spin the data.
• 0:57
  The piece was sparked by two links that were dominating VC Slack last week.
• 1:00
  Dan Primak's newsletter declaring that VC returns kinda suck, and a viral dashboard
• 1:05
  analyzing 49 mega funds concluding that scale compresses returns.
• 1:10
  While Micah's first instinct was to celebrate what he always knew to be true, he pushed
• 1:13
  himself to be a little bit more intellectually honest, and what follows is five questions the
• 1:17
  partnership debated internally.
• 1:19
  He walks through everything from the structural flaw of comparing fully baked older vintages to
• 1:24
  half baked recent mega funds to the unsustainable one to one ratio of VC firms to
• 1:29
  fundable startups at the 2021 peak to the reality that many of the sacred tenants
• 1:34
  venture, like ownership targets, pro rata rights, concentrated portfolios, or heuristics
• 1:39
  built off a handful of lucky nineties and February vintages.
• 1:44
  He closes with a point that TVPI is heavily influenced by Marx, and Marx are in many cases
• 1:50
  managers grading their own homework.
• 1:54
  In this content of media floods, it is really simple to see a piece that fits your belief
• 2:00
  system and then use it as armor to double down into why your model wins.
• 2:04
  So I do really admire the self awareness here.
• 2:06
  When you, Charles, see narratives like this catch fire on LinkedIn or X, how do you
• 2:12
  personally pressure test whether the data is telling the whole story or just the convenient
• 2:17
  one for yourself?
• 2:19
  A lot of times people send me, see, this article says that portfolio constructions like
• 2:22
  yours work.
• 2:23
  I'm like, well, I already believe that.
• 2:24
  Otherwise Yeah.
• 2:25
  I think this is kind of a theme for a lot of the articles this week.
• 2:29
  The GP perspective and the LP objective are not always the same.
• 2:33
  And, you know, LPs are building portfolios of their own too.
• 2:37
  They have their own set of objectives, which are largely you know, if you're an LP and
• 2:41
  you're like, hey.
• 2:42
  I wanna get build a portfolio of 300 names.
• 2:44
  I want 10 managers.
• 2:46
  They each have 10% ownership in those three names, and I want no overlap.
• 2:50
  Mhmm.
• 2:50
  You could have two very good fund managers who love to collaborate together, and an LP could
• 2:53
  reasonably say, if I invest in one of you, I'll get most of the good stuff.
• 2:57
  Because hopefully, the best stuff is the stuff you do together.
• 3:00
  I remember when I at Uncor, there was another fund we had 40% overlap with.
• 3:03
  We went to an LP and like, we already think we have exposure to the stuff that you have that's
• 3:06
  gonna work because we think it's gonna be predominantly the stuff.
• 3:09
  And it turned out that they were wrong.
• 3:10
  It actually was not the intersection of the stuff we did together, but I can understand the
• 3:14
  logic behind that.
• 3:15
  That makes sense to me.
• 3:17
  And so I always ask myself, what would make this uncomfortably difficult to process?
• 3:22
  So everyone's like, oh, like, big fun size compresses returns.
• 3:25
  I'm like, but what if it doesn't?
• 3:26
  What if it doesn't?
• 3:27
  So every time someone provides me with something that feels intellectually soothing,
• 3:30
  I'm always like, but what if the converse is true?
• 3:32
  And so I think the big unanswered question well, there's two.
• 3:35
  One, we have these massive, massive private companies, and we don't know if the public
• 3:41
  markets will value them the same way.
• 3:43
  And to the TVPI point, if the manager's marks and the market's marks turn out to be the same,
• 3:49
  then grading your own homework isn't so bad.
• 3:50
  It's more like, you're not grading your homework.
• 3:52
  You're putting your own homework sheet into the Scantron machine, or you're, like, pushing
• 3:56
  submit on the form.
• 3:57
  If it turns out that VCs are wildly overly optimistic about companies relative to public
• 4:02
  markets, then we do have a problem.
• 4:03
  We have a problem with Marx.
• 4:05
  Also, there's just so little liquidity.
• 4:08
  Now, like, you could say for the big companies, you'd like, it doesn't matter what the public
• 4:11
  markets think.
• 4:11
  Stripe is basically a public company.
• 4:13
  If you want to trade out of it and lock in your DPI right now, you could.
• 4:16
  Plenty of firms have done that.
• 4:17
  But I think at the industry level, we've never had more private value creation without getting
• 4:23
  public market checks.
• 4:24
  So if it turns out that Databricks is as valuable as we think it is and OpenAI is as
• 4:29
  valuable as we think it is, then there will be mega funds that have small VC fund like cash on
• 4:35
  cash multiples on much, much, much larger piles of cash.
• 4:38
  The VENCAT thing from months ago, which is I think in a world where outcomes were all sub
• 4:43
  1,000,000,000, highly concentrated funds with high ownership were the only way to make the
• 4:48
  math work because the ceiling on exits for private companies was small.
• 4:53
  We have now lifted the ceiling by at least a 100 x.
• 4:57
  Maybe we lifted it by a thousand x.
• 4:59
  And so why would you believe the strategies that work at a billion dollar terminal outcome
• 5:04
  are the only strategies that work at $100,000,000,000 or trillion dollar outcome
• 5:08
  scale.
• 5:08
  The issue is we just don't have enough data or experience to say but I think LPs are smart
• 5:14
  people, and I think they will realize if the terminal outcomes can really be as large as we
• 5:18
  think, and we think that this manager can get into those companies, I think the next ten
• 5:23
  years of venture, I think LPs are gonna go back to you know, they always say venture capitals
• 5:27
  is not an asset class.
• 5:28
  It's an access class, and, like, you either in the right fund managers.
• 5:31
  I think we're gonna go back to a world where LPs are gonna say, I don't care what your
• 5:34
  strategy is.
• 5:35
  If you have access to the consensus hot companies that we think will become a
• 5:39
  $100,000,000,000, I don't care if you own 1%, point 1%, 10%.
• 5:43
  All that really matters is can you get into those companies, whether by hook or by crook,
• 5:47
  in large scale or big scale at pre seat, I think we might end up in a world where, like,
• 5:52
  all of this stuff gets scrambled.
• 5:53
  But would that be counter to then the hang up here with LPs that, like, I need you to get to
• 5:58
  one x DPI as quickly as possible?
• 6:00
  Because if you had but I guess to your point, if the outcomes are so outrageously large, if
• 6:05
  you had even just point 1% in that company, maybe it could bring you to one x or at least
• 6:11
  point five.
• 6:12
  I think, like, the more dystopian thing that can happen is if you're not in these 10
• 6:16
  companies, I'm not taking the meeting.
• 6:17
  You have all these really interesting companies.
• 6:19
  Good for you.
• 6:20
  You're not in the hot names that I care about, and your hottest name isn't that hot.
• 6:24
  You're not one of the chosen few who has access to these.
• 6:27
  But this is all predicated on the fact the belief, rather, that the companies that we
• 6:30
  think are hot today are gonna be the companies that matter in the end.
• 6:33
  We could go back two years and say, like, what companies were we really excited about in the
• 6:37
  world of AI?
• 6:38
  I think the list would look very different than it looks today, and it'll probably look
• 6:41
  different again in two years.
• 6:42
  One quote that I liked here is that Micah opened the piece saying that venture's starting
• 6:48
  to feel a little bit less like finance and more like politics.
• 6:51
  Just everyone kind of, like, advocating and pushing the place that their capital strategy
• 6:56
  sits in.
• 6:57
  I guess for you, at what point do you know when you're having a conversation with somebody
• 7:00
  where you know your strategies are completely different?
• 7:02
  You're like, oh, we're not gonna meet in the middle.
• 7:04
  Is it worth me chatting through why I sit where I sit?
• 7:08
  How do you even pick which, not fights, but like which conversations to double down in?
• 7:12
  I have a general rule, which is I'll talk to anybody who I think is trying to get to a yes.
• 7:18
  If they're really trying to get to a yes, there's a conversation to be had.
• 7:24
  Sometimes they meet LPs and they're just like, I don't believe in your portfolio construction.
• 7:27
  Well, I don't know where we go from here.
• 7:29
  I've not found that, like, trying to educate people about why it works works.
• 7:34
  So usually if I meet someone like that, I'm like, I'm curious as to why you don't believe
• 7:36
  in it.
• 7:37
  I'm I'm just obviously not gonna change what I do.
• 7:39
  You're probably not gonna change your belief.
• 7:41
  And that's most of the interesting learning I've gotten has been from people, for example,
• 7:44
  that one LP was like, I'm trying to build a portfolio of 300 names where every manager has
• 7:49
  consistent ownership and where I have consistent ownership in that fund as a
• 7:52
  percentage of the fund So that, like, I have kind of consistent exposure to those names, and
• 7:57
  my belief is that those 10 managers will get into the companies that matter.
• 8:01
  And because they're not overlapping, I'm increasing my pro I'm like, oh, you're doing
• 8:04
  what I'm doing, but just through the lens of a portfolio.
• 8:06
  And he's like, look, a fund that gives me a 100 companies in a single shot with less ownership
• 8:10
  than I want does not help me achieve my portfolio objective that I'm trying to create.
• 8:15
  He's like, but I actually understand why you're doing what you're doing because you're doing
• 8:18
  the same thing I'm doing but for yourself.
• 8:20
  And we in the end, we're kind of like, we can't work together, but we understand each other.
• 8:23
  Do you have the same thought process for peer GPs who just sit in different categories when
• 8:28
  you're kind of like sparring over strategies?
• 8:30
  Or it's just different, I imagine.
• 8:31
  I have people who I really respect who make like two and four investments a year.
• 8:35
  And I tell them, I don't think I could do what you do.
• 8:38
  My brain doesn't work that way.
• 8:40
  And they tell me, I don't think I could do what you do.
• 8:42
  It seems, like, terrifying.
• 8:43
  I'm like, well, what I do feels natural to me.
• 8:45
  And the weirdest compliment I ever got, I think it's a compliment, was from an LP who passed on
• 8:49
  our funds, and I showed them returns.
• 8:51
  And he goes, well, this I still think this is a bad idea, but it seems to be working for you.
• 8:55
  Thank you.
• 8:57
  Thank you.
• 8:57
  Think You want the follow-up?
• 8:59
  Or But I know what that person was trying to say, which is, like, I refuse to change my mind
• 9:03
  about the thesis and theory, but I'm willing to acknowledge that it might be working for you.
• 9:08
  Well, our next piece is from Jeff Weinstein, partner at FJ Labs on LinkedIn, where he opens
• 9:14
  with a history lesson that reframes how you look at what is happening in venture right now.
• 9:18
  He goes back to the sixties and early seventies when Wall Street became obsessed with a group
• 9:22
  of 50 blue chip growth stocks, companies like Xerox, Polaroid, Disney, and Coca Cola that
• 9:28
  were seen as so obviously transformational that valuation almost didn't factor into the
• 9:33
  conversation.
• 9:34
  Investors piled in with total conviction, and for a while, it worked until it didn't.
• 9:39
  By the mid seventies, many of those stocks had lost more than half their value, and anyone who
• 9:43
  bought at the peak consensus in 1972 spent the next twenty six years underperforming the
• 9:49
  broader market.
• 9:50
  The companies were great.
• 9:51
  That wasn't disputed.
• 9:53
  The prices were the problem there.
• 9:55
  And Jeff's point is that venture is at a serious risk of repeating the same mistake now
• 9:58
  where the power law logic backing only the biggest perceived AI winners has shifted from
• 10:03
  being a genuine investment insight into something closer to a crowded consensus trade.
• 10:07
  The concentration of capital into a small handful of mega rounds has only accelerated,
• 10:11
  and the size of these raises has started to make 2021 look restrained.
• 10:16
  He did also write a follow-up piece that gets into where he thinks the real opportunity
• 10:19
  actually sits at the result, and namely places that have largely been abandoned in the rush
• 10:24
  towards consensus.
• 10:25
  Jeff's whole argument is that everyone being right about the same companies at the same time
• 10:31
  is exactly what makes it dangerous.
• 10:33
  And if you zoom out, Charles, across cycles that you've seen up close in the past, is there
• 10:38
  a pattern to when the consensus trade actually breaks and what tends to be trigger?
• 10:42
  I know we've talked about the opportunity arbitrage window before, but anything that
• 10:46
  you've gleaned from the past.
• 10:47
  I just think life is easier if you go with the flow.
• 10:50
  Mhmm.
• 10:50
  And you're like, everybody else is doing this.
• 10:52
  I think easy and making money are oftentimes at odds.
• 10:55
  And so I read this, and I was like, if you look at the firms that are in Quince, one of the
• 11:00
  companies you mentioned, quite surprising.
• 11:02
  They are not the firms you would expect to be in that investment.
• 11:05
  And whenever I tell folks how well our CPG portfolio is doing, they're usually just like,
• 11:10
  first of all, why are you even in those companies?
• 11:12
  And second, why haven't I heard how well they're doing?
• 11:13
  I'm like, because nobody cares right now.
• 11:15
  And so I've always felt like one of the best things about being a generalist is you can,
• 11:19
  like, shop in different aisles.
• 11:21
  You can build a portfolio in different categories.
• 11:23
  The upside is when things go out of favor, you might have the in favor thing.
• 11:27
  The downside is when everybody's consensus in on one category and you're a generalist, it can
• 11:31
  feel hard because you're working with founders and you're working in sectors that people just
• 11:35
  literally don't care about.
• 11:37
  And it can be hard to, like, show up to a party and tell people about the last investment you
• 11:41
  made, and they're like, it's not an AI harness company or it's not a fund like, you're like,
• 11:45
  no.
• 11:46
  It's not.
• 11:46
  And they're just like, do you understand your job?
• 11:49
  And so I do think it's a good reminder that oftentimes, the companies that matter in the
• 11:54
  end are not the companies we think matter in the moment.
• 11:56
  And I won't pick on specific companies, but I can think of even three or four years ago,
• 12:00
  there were brand new AI companies that people just assured me were gonna be consequential,
• 12:04
  and they just haven't been.
• 12:06
  It also made me think a little bit more macro, but about the generation of investors coming up
• 12:12
  right now who are a little bit more junior who've really only known one playbook, which is
• 12:17
  right now to just find an AI angle, back consensus winners, compete for similar logos or
• 12:22
  things that look like them.
• 12:23
  And I think it's just because that's ultimately what their partnership is telling them to do.
• 12:27
  They haven't lived through a cycle where the consensus trade completely unwound, and the
• 12:32
  people who looked contrarian ended up right yet.
• 12:35
  What are, I guess, your perspective, the real downstream consequences of raising a generation
• 12:39
  of investors inside that kind of monoculture?
• 12:43
  I think it depends on and this is something I've been thinking about a lot for, like, the
• 12:46
  last two years.
• 12:47
  You know, in a in a theoretical world, if you if this were public equities and you were bad
• 12:51
  for two years, you would just get fired.
• 12:53
  Your LPs would redeem and, like, would move on and be out of business if you're a hedge fund.
• 12:58
  I think in venture, it's actually not clear to me that there's much in the way of a corrective
• 13:02
  function.
• 13:03
  And I I don't say this to, like, blame any actor in the ecosystem.
• 13:07
  We kinda blew everything.
• 13:08
  I don't know.
• 13:09
  I watched Margin Call on the way back from a birthday party with a bunch of finance people.
• 13:14
  There's, like, this part at the end where Jeremy Irons is having dinner by himself.
• 13:17
  And he's just like, we just do this all the time.
• 13:19
  And he, like, goes through a litany of, like, finance generated bubbles that kinda blew up
• 13:23
  the economy in the end.
• 13:24
  He's just like, well, we're gonna do it again.
• 13:26
  As always happens, we blew it up, and we're gonna make a ton of money on the way back up.
• 13:30
  He says it, like, very matter of factly.
• 13:31
  And I think in venture, there aren't that many people who've had to pay a steep price for the
• 13:36
  excesses of 2021.
• 13:38
  We've kind of, like, memory hold that whole era.
• 13:41
  And I'm actually not entirely sure that if the AI trade blows up and people lose a lot of
• 13:46
  money, I I could name at least five or six firms that I'm sure will still be in business
• 13:51
  despite losing tons of money.
• 13:52
  And if that's the case, maybe the incentive is to just do the hot thing.
• 13:57
  Maybe that's actually, unfortunately, the system incentive, which makes me really
• 14:00
  uncomfortable to say it.
• 14:02
  But it is possible that there is not this corrective force that says, like, if you
• 14:06
  underperform or if you make a blunder, that you'll have to pay.
• 14:09
  I mean, I've heard LPs literally say, I just write off any vintage from 2021 in my
• 14:13
  portfolio.
• 14:14
  So in fact, no one gets no one gets dinged for that.
• 14:17
  I just have written it
• 14:18
  off.
• 14:18
  Which I think changes the incentive system for you as a fund manager.
• 14:22
  The incentive system is a long, long time ago, someone I've known since the earlier Internet
• 14:26
  days was like, look.
• 14:27
  If you went all in on bubble '96 to '99 Internet and you rode that thing and got out,
• 14:33
  it doesn't really matter what else you did since then.
• 14:35
  You would have done so well.
• 14:37
  And same thing was true of optical networking, probably early fintech.
• 14:40
  For some of these themes, you catch them early and you get them right, venture is a weird
• 14:45
  business where you can be right one or two times in your career and be fabulously wealthy
• 14:48
  for the rest of your life.
• 14:49
  I think for more junior folks who are coming up in this time, I actually think like this is the
• 14:53
  potential to be, I don't know, probably one of the biggest learning moments.
• 14:56
  Then questions to yourself of do I wanna be in venture?
• 14:59
  If either one happens, if it's a massive success and everything is valued at the way
• 15:03
  we've marked them, then you're sort of like, wow.
• 15:05
  I've I've participated in one of the most lucrative times with InVenture.
• 15:09
  And then if it doesn't do well and you're in an analyst at like a large fund or a mid sized
• 15:12
  fund who is participating in that, you will be able to take that learning to the next thing
• 15:16
  and decide, do I wanna participate in this and take these learnings with me, or am I out
• 15:20
  completely, and I'd rather be on the operating side or be completely outside of the asset
• 15:23
  class?
• 15:23
  I think there's a lot to be gained.
• 15:25
  Our last piece is from Rob Goh at Next View titled The Four Approaches early stage venture,
• 15:30
  and it's a clean framework for making sense of how different seed investors actually make
• 15:34
  decisions even when they all claim to be doing the same thing.
• 15:37
  Rob breaks early stage investing into four distinct models.
• 15:40
  He's got classic institutional seed, specialist seed funds, multistage mega funds that stretch
• 15:45
  down to seed, and then accelerators.
• 15:47
  And he's super candid about the real strengths and failure modes of each.
• 15:51
  What makes this piece interesting is that he does not declare a winner.
• 15:53
  His view is that the pie is big enough for all four to coexist and that the failure mode is
• 15:58
  not picking the wrong model, but not knowing what game you're actually playing.
• 16:01
  Yep.
• 16:02
  Closes with the line, don't let LPs be the tail that wags the dog.
• 16:06
  What's right for an LP in the context of their portfolio is rarely exactly right for the
• 16:11
  manager.
• 16:11
  Rob is pretty direct that the failure mode is constantly chasing your own tail and not
• 16:16
  knowing what game you are actually trying to play.
• 16:18
  How much more common is that type of failure now versus years ago when the venture industry
• 16:22
  was more of a cottage industry?
• 16:24
  I think it goes back to the thing we talked about at the beginning.
• 16:27
  I think the reason there's so much narrative competition in venture is because people feel
• 16:32
  the need to justify what they're doing in an environment that's become far more complicated.
• 16:37
  And my experience has been the big challenge with narrative competition is that everybody
• 16:41
  doesn't have the same size megaphone.
• 16:43
  The multistage funds have a disproportionate impact on the point of view of LPs, in my
• 16:49
  experience, because LPs have more money with them.
• 16:52
  They have more access to the senior decision makers, and I think the smart ones use that
• 16:56
  access to shape LP perspective on how the world works.
• 16:59
  And so I think if you're a seed fund manager and you haven't been asked how does the
• 17:05
  incursion of multistage funds into seed impact your business, I simply don't believe you.
• 17:09
  Everybody's being asked that question.
• 17:12
  And the answer you give is a function of, like, your strategy and and your belief system.
• 17:16
  But I think that question's being put to seed managers because multistage funds have said
• 17:21
  we're gonna play in seed too.
• 17:22
  Those folks have lots of access, and they they can tell a piece of story.
• 17:26
  All of these brand assets and reputation we have are transferable from series a to seed.
• 17:30
  And to some extent, it's true.
• 17:32
  You know?
• 17:33
  I don't know many people who wouldn't take money if a senior partner at a big multistage
• 17:37
  firm called them up and said, hey.
• 17:38
  I wanna invest in your company.
• 17:39
  That's actually not what's happening in most cases.
• 17:41
  And I would flip it around.
• 17:42
  The question is, I found it's very difficult to discredit somebody else's strategy, and it's
• 17:47
  probably not worth your time.
• 17:48
  It's probably more important to focus on why what you're doing works for you specifically as
• 17:53
  a manager and how your strategy can still outperform.
• 17:56
  And look.
• 17:56
  I think right now, we're in this, like, consensus squared environment.
• 17:59
  And so the less consensus to your idea, the harder you have to work to compete in this
• 18:03
  narrative environment.
• 18:04
  And so I find that the next best thing to being a multistage venture fund is being a seed fund
• 18:09
  that invest in the same deals as all the multistage funds because you can tell somebody,
• 18:12
  like, we're investing alongside these folks.
• 18:15
  But the tricky thing is seed fulfills different objectives for different firms.
• 18:19
  For a dedicated seed manager, seed is where you make your money.
• 18:21
  For a multistage firm, seed is where you do a lot of things.
• 18:25
  It's where you stay current.
• 18:26
  It's where you build new relationships.
• 18:28
  It's where you get intel on what's happening on the ground floor.
• 18:30
  You also can make some money, but making money is one of maybe three to four objectives you
• 18:35
  have.
• 18:35
  And I think that to me is the biggest difference now is that before everybody who was
• 18:39
  in seed, it was their primary business.
• 18:41
  It was the way that they made money, and everybody kind of had similar objectives.
• 18:45
  When you inject in a bunch of new players who have a variety of objectives where making money
• 18:50
  is one of them, it does change things in a pretty material way.
• 18:52
  My last question here is a little bit more existential about how much time we're spending
• 18:58
  within AI, but Rob flags that the best generalists often outperform specialists over
• 19:03
  multiple vintages, and I have a feeling we obviously agree with that one here.
• 19:06
  Yeah.
• 19:06
  But if the majority of capital and fund activity right now is concentrated in AI and
• 19:12
  nearly every deal has some AI angle to it, does it make the so called generalists of this cycle
• 19:16
  accidental specialists?
• 19:18
  Where do you fall on that?
• 19:19
  If, like, every single deal that I've done across health care, construction, against it's
• 19:24
  all has an AI piece.
• 19:26
  I think it's like mobile.
• 19:27
  There was a period of time when, like, mobile happened, mobile was differentiator.
• 19:31
  We're like this thing but mobile, or we're the mobile enabled version of x.
• 19:34
  And then it just became well, of course, you have a mobile version because you're you have a
• 19:38
  brain.
• 19:39
  And I think AI went from a specialty domain where you had firms, Basiset and Zeddot.
• 19:44
  You you had firms who were like, we're specialists in AI.
• 19:46
  And I've always felt like this is the the risk of being a specialist is that the thing that
• 19:50
  you do becomes so big.
• 19:51
  We've talked about this with SAS before with the merchants and other firms.
• 19:54
  Like, it becomes so big that everybody does it, then it ceases to be a specialty practice
• 19:59
  anymore.
• 19:59
  And I think AI has clearly crossed over into that.
• 20:02
  And I think now to be a journalist, you have to actually work much harder to find stuff that
• 20:05
  doesn't fit those consensus narratives.
• 20:07
  Well, we are all out of time today.
• 20:08
  Thank you, Charles, for hopping in.
• 20:10
  Thank you to our subscribers for jumping in as well.
• 20:13
  Please like and subscribe, and we will talk to you all next week.
• 20:15
  Thank you.
• 20:16
  Bye.