Latest / The Learning Corner by Precursor / Episode #34: AI’s Labor Problem, Bridge Round Realities & Big Tech’s Acqui-License Deals

Transcript

Episode #34: AI’s Labor Problem, Bridge Round Realities & Big Tech’s Acqui-License Deals · The Learning Corner by Precursor

• 0:00
  Sometimes you go into these things with a set of beliefs and hypotheses, and if you're wrong,
• 0:05
  it's okay to just shut it down.
• 0:07
  You don't have to do some YOLO pivot.
• 0:09
  It's actually okay to just be like, look.
• 0:11
  The conditions that needed to be be true for this to be a good idea are no longer true.
• 0:14
  I recognize that and renew to something else.
• 0:16
  It's the week of May 26.
• 0:19
  Welcome to the Learning Corner of Precursor Ventures podcast.
• 0:26
  Welcome to the Learning Corner by Precursor.
• 0:29
  I'm Mia Farnam, and I'm joined by our GP and managing partner, Charles Hudson, where we will
• 0:33
  be talking through the articles and Goodreads that caught our team's attention over the past
• 0:38
  week.
• 0:38
  Our first topic is from a good friend at Alicore, Youssef, who reflects on the growing
• 0:43
  tension between AI advancement and labor displacement.
• 0:45
  Feels like since we mentioned this on a podcast a few weeks back, there have been a lot more
• 0:50
  open conversations about what AI threat to labor looks like.
• 0:53
  While AI promises productivity gains, it's increasingly eliminating entry level jobs and
• 0:58
  eroding traditional on ramps into professional life.
• 1:01
  Youssef calls for a new wave of investment focused on economic infrastructure.
• 1:05
  AI that empowers workers, improves public services, and expands educational access, not
• 1:10
  just automates headcount.
• 1:11
  Drawing on his own immigrant family background and experience at Uber, he warns that investors
• 1:16
  must grapple with the societal consequences of labor automation and argues that capital
• 1:20
  allocation is a patriotic act.
• 1:22
  Quote, where you invest shapes the future, he writes, calling on VCs to build a more
• 1:27
  inclusive AI powered economy.
• 1:29
  One great quote from the piece was, quote, in last month, such private conversations with
• 1:35
  friends in VC have intensified as more pitch decks sold cost cutting and fewer painted the
• 1:40
  inspiring vision of giving people superpowers.
• 1:43
  And it's really crazy how quickly the narratives can change.
• 1:46
  However, it honestly makes a lot of sense.
• 1:48
  Enterprise clients are finding that the value of these quote unquote, superpowers is much
• 1:52
  lower than marketed, and now companies almost need to pivot to the cost cutting story to
• 1:56
  attract and retain customers.
• 1:57
  So wanted any additional thoughts from your end on this one.
• 2:00
  Yeah.
• 2:00
  I'm glad Youssef wrote that post.
• 2:01
  I really enjoyed reading it.
• 2:02
  I think there's two things that feel different.
• 2:05
  One is I've been following the Duolingo social media response to the post.
• 2:10
  And, you know, it's interesting for a brand that was, for a long time, I think, held up as
• 2:14
  a paragon of how to do social and beloved.
• 2:17
  It's been interesting to just see the response to people.
• 2:20
  And I think we're in a world where most of the discourse around AI, we're only kind of hearing
• 2:26
  from the capital and management side of the equation.
• 2:29
  And so because we're only really hearing from, for the most part, capital and management, with
• 2:34
  one exception, which I'll get to in a second, I think what you're hearing a lot is a top down
• 2:38
  view of what this technology can do.
• 2:40
  And most of those folks are focused on improving efficiency and cutting costs.
• 2:44
  And so some people are more direct about that than others.
• 2:48
  But I think it's been interesting.
• 2:49
  I haven't heard or read the voices of customer service folks who've been negatively impacted.
• 2:54
  What I have been fascinated by is just listening to people who are professional
• 2:58
  software developers, who are using Windsurf Cursor or other copilots and the mixed feedback
• 3:04
  I get from them about the efficacy of these tools.
• 3:07
  And it seems to be also a function of where those folks are in their career and how much
• 3:11
  formal training they have in software development.
• 3:13
  And I would say it's not a universal, this is an amazing tool.
• 3:17
  And I hope that as we continue to roll out AI, I don't think we can put the toothpaste back in
• 3:21
  the tube, but I do hope that the discourse expands and we hear from people up and down the
• 3:26
  stack, not just the people in leadership, but the people who are being directly impacted by
• 3:30
  these technologies in terms of how they feel and, like, what their plans are, and people in
• 3:33
  the middle who I have real sympathy for who are trying to figure out, hey.
• 3:36
  I've got a team of five people, but I've got a top down directive to deploy AI.
• 3:39
  I wanna do my job and deploy AI, but I am on the front lines.
• 3:42
  I have to deal with the fact that we have four customer service people going to two, which
• 3:46
  means two of the people who report to me, I'm gonna have to really have a really tough
• 3:48
  conversation with them about what it means for them.
• 3:50
  And so I think the privilege of being a venture capitalist is the decisions you make, as Yousef
• 3:55
  mentioned, about what you fund impact the future.
• 3:57
  And we're literally in the business of funding the world that we wanna see exist.
• 4:01
  And I'm certainly not saying we stop AI or try to roll it back, but I do think we should be
• 4:07
  mindful of, like, what are the consequences?
• 4:09
  And it's been interesting.
• 4:10
  I have a lot of friends in venture, and this is becoming more silly because they have college
• 4:13
  aged children or high school aged children, and they're trying to figure out what advice should
• 4:17
  I give my child about what to study, what should I do if they can't find a job in their
• 4:22
  chosen field.
• 4:22
  And there is some early evidence that it's those entry level white collar jobs that are
• 4:27
  it's not that they're being eliminated.
• 4:28
  It's just the demand for those is going down, and firms might hire fewer of those people.
• 4:33
  I also think what we might realize is that sometimes when you graduate, it really matters.
• 4:37
  And we might learn that this optimism gives way to a, hey.
• 4:40
  We kinda overdid it.
• 4:42
  Yeah.
• 4:42
  We actually do need we do need these people.
• 4:44
  And one thing I'll never forget is I had a bunch of friends who work in white collar
• 4:49
  service professions.
• 4:50
  And in around 2012, '20 '13, they were all complaining.
• 4:53
  Was like, what's the problem?
• 4:54
  They're like, our partner classes are very small.
• 4:56
  We don't have the people we need to promote.
• 4:57
  I said, what happened?
• 4:58
  He goes, well, in o seven, o '8 during the GFC, we just stopped hiring.
• 5:02
  We knew things would get better, but, like, we just stopped hiring.
• 5:04
  So we have these donut hole classes that are just really small, and we don't have the people
• 5:08
  we need to promote into leadership.
• 5:10
  And I wonder if something similar will happen, which is, you know, you've already seen Klarna
• 5:14
  backtrack a bit.
• 5:15
  Will people say, wow.
• 5:16
  Like, for a few years, we underhired and underinvested in people.
• 5:19
  And now that we need those people, you can't go back in time and grow them.
• 5:22
  You would think that in so many cycles that this has happened to large companies.
• 5:26
  I've seen in COVID for every consulting company, they stopped hiring.
• 5:30
  And then they went to this massive hiring because they had too many projects to be able
• 5:34
  to service.
• 5:34
  And so they overhired, and they had to cut people back again.
• 5:37
  You would have imagined people to understand how a slinky moves.
• 5:39
  They don't, and I find that to be interesting.
• 5:41
  And over the weekend, all the news around Anthropic and OpenAI of, like, the models not
• 5:46
  actually following the rules of the people who hard code them and doing the things that you
• 5:49
  want them to do.
• 5:50
  It's this weird thing we're trying to bridge of, yes, you can add these capabilities in, but
• 5:54
  is now the right time to do them, and you change everything around the infrastructure of
• 5:59
  your business to fit that in today?
• 6:01
  Or is there a slow drip method that we can better run?
• 6:04
  One of my VC mentors, he's like, you know, consistency has value.
• 6:08
  If you just kind of invest at roughly the same pace all the time, you'll feel a little slow
• 6:13
  when things are busy, and you'll feel a little busy when things are slow, but you're not
• 6:17
  trying to guess or time the market.
• 6:19
  And I think there's something to be said in business for consistency and just saying, look.
• 6:22
  We're gonna hire these people at a consistent clip.
• 6:24
  Some years, everyone's gonna be stretched to the max, and there'll be other years where
• 6:28
  we'll probably be carrying a few people that we don't need.
• 6:30
  And on average, it'll be right.
• 6:31
  And trying to get it exactly right all the time seems like a fool's errand to me.
• 6:35
  Yeah.
• 6:35
  No different than the public markets of trying to time in and out.
• 6:38
  Last question on this one.
• 6:39
  Anything else that you're feeling folks are being more honest about in the venture ethos
• 6:44
  now?
• 6:44
  I've talked to a few people who have essentially said, I worry also about some of
• 6:49
  the people who are impacted because most people I've talked to have been like, we're not really
• 6:54
  worried about AI impacting our ability to get the most out of our top 25 or top third of our
• 7:01
  employee base, it's really the bottom.
• 7:03
  And I just feel like the stigma of being laid off due to AI and what that will do to the
• 7:09
  folks who are looking for work and they're like, hey.
• 7:11
  I was part of the such and such.
• 7:12
  People are like, oh, you were one of the ones that got laid off due to AI that therefore, you
• 7:16
  are an underperformer.
• 7:18
  And I think that's different from just pure cost cutting sometimes.
• 7:20
  So I think people are beginning to grapple with the fact that layoffs and staff redemptions are
• 7:24
  not this abstraction.
• 7:25
  They're like actual human beings who will have to go find something new to do.
• 7:29
  And some of them are in professions or verticals where, you know, the long term career
• 7:32
  prospects for entry level customer service done by humans are probably it's probably not great.
• 7:37
  Our next topic is from another friend, Peter Rocker, who dives into Carta data from around
• 7:42
  7,100 startups and finds that companies raising bridge rounds after their priced seed round are
• 7:48
  only half as likely to make it to series a.
• 7:51
  While founders understandably use bridges to survive harsh funding environments, the
• 7:55
  rationale from investors is less clear and the outcomes are statistically poor.
• 8:00
  He challenges the assumption that bridges help companies, quote, get over the hump, suggesting
• 8:04
  that investors might be better off backing new companies than doubling down into struggling
• 8:08
  ones.
• 8:09
  Charles, I think one thing that might be helpful is actually hearing from you when
• 8:12
  bridges became more popular and more Yeah.
• 8:14
  As they were emerging, was there more negative sentiment about the bridge as we see today?
• 8:19
  Well, I think I'm living through the fifteenth series a crunch in 10.
• 8:24
  I think it's amazing.
• 8:25
  We used to have bridges, then everyone's like, no.
• 8:27
  No.
• 8:27
  No.
• 8:27
  Bridge is a bad word.
• 8:28
  I'm sure everyone's heard the joke, like, we we only invest in bridges, not piers, the old
• 8:31
  bridge to nowhere.
• 8:32
  So I think people said bridge is a bad word, so we have to rebrand the bridge.
• 8:36
  And then we got post seed and seed two and seed extension, anything to avoid the word bridge.
• 8:42
  And I think there was a period of time where bridges worked really well, because it felt
• 8:46
  like, hey, we were just a quarter or two short of series a metrics.
• 8:50
  Let's go back and fund them, and we will go give it a try again in two more quarters.
• 8:55
  And, you know, if you didn't get it, you didn't get it.
• 8:58
  I think in the last few years, we've moved into a world where, I think Hunter was the one to
• 9:02
  coin it, you know, seed as a process.
• 9:04
  And thinking of seed as a discrete single funding round, to me, feels outdated.
• 9:08
  And I think what's happened is nobody really knows what good enough looks like at series a.
• 9:13
  People know what the progress looks like from the moment they invested in the company to
• 9:17
  where it is when the founder goes out to decide.
• 9:19
  In my experience, it's founders who decide when they wanna go raise.
• 9:22
  It's not VC's telling them when they should go raise.
• 9:24
  It's the founders making that choice, and they go out and raise.
• 9:27
  And what I've seen is sometimes you go out and you get very close to getting a series a, and
• 9:31
  you come back and you go, wow.
• 9:33
  There's one or two things we need to clean up about this company, but we could do that on a
• 9:36
  little bit of money.
• 9:36
  In some cases, you're like, wow.
• 9:37
  We're not even close, but we know kind of where we need to get.
• 9:40
  And with another million or two, we could get there and give it another try.
• 9:44
  And so I would say there are bridges of optimism and bridges of realism and bridges of
• 9:48
  necessity.
• 9:48
  Necessity.
• 9:49
  Sometimes, I think it's important to remember, it's really hard if you're a board member in a
• 9:52
  company.
• 9:53
  So I talked to RLPs about this.
• 9:54
  Most investors, I think, have a rational view of their portfolio companies and can say, this
• 9:59
  one is not a good one.
• 10:00
  And I don't think that many VCs will knowingly throw good money after bad.
• 10:05
  I think sometimes you're like, I'm on the board, and if I don't do this bridge or
• 10:08
  extension, it's gonna send a very strong signal and maybe a stronger signal.
• 10:11
  But honestly, Mia, there's lot of times indifferent.
• 10:14
  You're like, this company's okay.
• 10:15
  It's not so obvious that we shouldn't give them money, that we'll just decline it outright.
• 10:19
  And it's not so obvious that we wanna give them more money.
• 10:21
  Kind of in between.
• 10:23
  And more time will hopefully help that company figure it out.
• 10:25
  And everyone says, well, when are graduation rates gonna come back?
• 10:27
  And I wrote this to you all internally.
• 10:29
  I look at Peter's data, and what it tells me is if you're a startup, you get 7 to $10,000,000
• 10:34
  across two to three arbitrarily named rounds between pre seed and seed to prove that you are
• 10:38
  building a venture scale business.
• 10:40
  And to be kind, we'll say that's a business that people think could be worth
• 10:43
  $10,000,000,000 or more led by a very compelling founder.
• 10:46
  And if on 7 to $10,000,000, the series a community doesn't see it, they're probably
• 10:50
  never gonna see it.
• 10:51
  Yeah.
• 10:51
  And, like, that's what you have to work with.
• 10:53
  And you've gotta do everything in your power with that seven to ten at most to prove that
• 10:57
  you're building something special.
• 10:58
  And if you aren't, there's no amount of incremental improvement, metrics improvement.
• 11:02
  Like, none of that stuff's gonna matter.
• 11:03
  They've sort of voted on your idea and team.
• 11:05
  And I think it's important for founders to internalize that, and I've internalized that as
• 11:09
  an investor.
• 11:10
  And I think about that a lot when people come to me for bridges and extensions, and I'm just
• 11:13
  like, what if the market's already decided this just isn't a venture scale company?
• 11:16
  There's no reason to put more money in and push for that goal.
• 11:19
  I think a lot of the funds that specialized around series a for the last handful of years
• 11:23
  were really rocked by coming in at higher valuations.
• 11:25
  And so when you think about funds that aren't able to raise subsequent rounds, we're gonna
• 11:29
  see less folks who play around there.
• 11:31
  So that probably plays into that as well.
• 11:33
  It's, like, it's just gonna be harder to get seen by people because their benchmarks are
• 11:37
  just so much higher than they were prior.
• 11:39
  And if we had a healthy middle class venture of 400 to $750,000,000 funds, it wouldn't be so
• 11:47
  dire.
• 11:47
  But I think the challenge is both every GP I talk to at big funds is like, we're doing this
• 11:52
  barbell thing, really big funds and really small funds.
• 11:55
  And most LPs seem to believe that's where we're headed to.
• 11:58
  And without kind of this middle tier of funds, you're either a 10 to $25,000,000,000 company,
• 12:04
  or you've gotta find another plan because you're not gonna raise a series a from a big
• 12:08
  firm.
• 12:09
  And I wish we had a healthy middle class of firms.
• 12:11
  I'm not optimistic.
• 12:12
  We're not gonna have an explosion of firms filling that gap.
• 12:15
  And I think that's the tough thing to to swallow, which is if that's the opportunity
• 12:19
  where you could stand out as a fund, wouldn't you wanna concentrate your efforts there?
• 12:22
  But I know the math doesn't seem as exciting as it would be if you could become the biggest
• 12:27
  fund possible and then just come in at later rounds or own the entire life cycle.
• 12:32
  So this middle is not as attractive, I think, for fund managers because also it's this
• 12:36
  complete unknown.
• 12:37
  You're still dealing with imaginary metrics and traction, and you don't get to participate in
• 12:41
  the immediate upside that the earliest stage investors do.
• 12:43
  And you still need power law companies.
• 12:46
  So you need companies that are big enough to be interesting to you.
• 12:49
  So we'll say 1 to 2,000,000,000, but not so obviously huge that they would catch the attend
• 12:53
  because the other problem is you're not gonna win the bids for a lot of these because if
• 12:56
  they're obviously big, the firms with more capital will just pay up and they will outbid.
• 13:00
  So the reason that middle is under pressure, it's a hard place to occupy.
• 13:04
  One last question for this one.
• 13:05
  For founders who are thinking about bridges or raised them prior.
• 13:10
  How should they be thinking about maybe deciding to do another bridge, to not do
• 13:15
  another bridge, and what sort of counsel or soundboarding advice is the right one to guide
• 13:19
  them?
• 13:20
  I mean, I've, at some relationship risk, have decided in the last few years to just be more
• 13:25
  honest with people and direct about what I think their prospects are in two dimensions.
• 13:29
  One, what are their prospects for actually raising more money?
• 13:33
  And two, what do I think they have to do with that money to unlock the round?
• 13:36
  In some cases, I'm like, you can do this bridge extension, but you have to get the company
• 13:39
  profitable.
• 13:40
  It's not negotiable.
• 13:41
  In some cases, I've told them, you can raise this bridge round, but we have to get three
• 13:45
  times as much out of every dollar for this next set of dollars as we did for the last set of
• 13:51
  dollars, or else it's not gonna work.
• 13:53
  You just weren't efficient enough.
• 13:54
  And I think founders need to have some sense for what they're gonna do with the money other
• 13:58
  than simply remain in business and stay alive.
• 14:00
  That's not a reason to write a check to a company so that they can stay alive.
• 14:04
  And I feel for founders because a lot of times, the feedback they get about why people didn't
• 14:08
  invest, oftentimes, founders are like, we just gotta get from 50 k MRR to two fifty.
• 14:12
  I'm like, I don't think that's actually the root cause.
• 14:15
  I think the root cause is people think our market is small, or people think your team
• 14:19
  isn't that great, or people think there's a competitor in your category that's already won
• 14:23
  and you're and also these are not fun things to tell people, but I do think some things you can
• 14:29
  fix with time and money, and some things are really hard to fix with.
• 14:31
  If people just think, hey, such and such company has won the category or has the
• 14:35
  dominant model, I don't know how you'd change that in eighteen months with $2,000,000.
• 14:39
  And I think it's a difficult thing for investors to acknowledge because it means that
• 14:43
  everyone around the table was wrong.
• 14:45
  That's why investors don't wanna say it partly because I think they struggle specifically with
• 14:49
  saying really transparent things that would benefit founders, but also because then it puts
• 14:53
  them in the hot seat.
• 14:54
  I've told a few people we were wrong together.
• 14:57
  Yeah.
• 14:57
  We all like, you and I both thought the world was gonna go this way or this would be what
• 15:02
  mattered or, you know, we we have a portfolio company that had sort of bet on an Apple policy
• 15:08
  remaining in place for a long time as part of their business model.
• 15:12
  And I was like, it seems very unlikely that that policy is gonna get reversed because Apple
• 15:17
  just kept it for a long time.
• 15:18
  And we were both like, that's the big existent, and that policy is in the process of being
• 15:22
  unwound.
• 15:23
  And the founder's like, well, this business is not nearly as interesting in a world where that
• 15:27
  policy is unwound.
• 15:28
  And I'm like, I agree with you.
• 15:29
  Yeah.
• 15:29
  I think the other thing I will say is sometimes what I've told people is the right answer is
• 15:33
  you go into these things with a set of beliefs and hypotheses, and if you're wrong, it's okay
• 15:38
  to just shut it down and return the money or just shut it down if there's no money to
• 15:42
  return.
• 15:43
  You don't have to do some YOLO pivot.
• 15:45
  You don't have to try to scrape and find something.
• 15:48
  It's actually okay to just be like, look, the conditions that needed to be true for this to
• 15:51
  be a good idea are no longer true.
• 15:53
  I recognize that and renew to something else.
• 15:55
  Yeah.
• 15:55
  For our last topic, Bloomberg reported that the US justice department launched an antitrust
• 16:00
  investigation into Google's agreement with Character AI, the rising chatbot company known
• 16:05
  for its AI powered character emulation tools.
• 16:08
  The deal struck with Character AI happened in February and was valued at $2,700,000,000 to
• 16:15
  essentially buy out the existing shareholders, founders, and other investors.
• 16:18
  And in exchange, Character AI's original founders, who were ex Googlers, were rehired
• 16:24
  back to Google, and Google obtained license to use Character AI's tech.
• 16:29
  So regulators are probing into whether the deal, which is a license bundled with talent
• 16:34
  acquisition, was structured to bypass formal merger review while still giving Google an
• 16:38
  unfair edge in the AI ecosystem.
• 16:41
  The investigation is part of a broader effort by the DOJ to ensure big tech doesn't
• 16:45
  consolidate power in the fast moving AI sector, especially after recent court rulings against
• 16:49
  Google in search and ad tech.
• 16:51
  The ruling on these, acquihire plus license deals, signal a new playbook for incumbents and
• 16:56
  a precedent for rulings against similar activities in the future.
• 17:00
  Charles, thoughts on if you see these hybrid licensing and hiring deals as fair competition,
• 17:06
  and how should regulators be thinking about this?
• 17:09
  And honestly, also AI founders who are building in this space too, when there's, like, no
• 17:13
  equities changing hands, but significant IP and talent do.
• 17:17
  I just felt like I've been calling those things the aqua license.
• 17:20
  I won't impute.
• 17:21
  I don't work at Alphabet.
• 17:22
  I don't know what went into the decision.
• 17:24
  All disclaimers.
• 17:25
  Amazon has done a couple of these.
• 17:27
  These just feel on the nose.
• 17:29
  These just feel like a way to acquire talent without tripping or running afoul of the DOJ's
• 17:37
  desire to look at tech consolidation, and I was just like, these are gonna get investigated.
• 17:42
  I don't know whether that was the intent, but if you were looking for the least charitable
• 17:45
  interpretation, you would say these were designed specifically to avoid.
• 17:50
  And for all I know, that's not what happened.
• 17:52
  It's that they wanted the talent, and there was a set of people at that company who wanted to
• 17:55
  continue with the original vision.
• 17:56
  And this is just a very expensive hiring bonus, including some compensation for the original
• 18:01
  investors.
• 18:02
  But I just remember when those came out, was like, you do a couple of these, and someone's
• 18:05
  gonna ask questions.
• 18:06
  And so given the amount of legal and regulatory scrutiny that Google has been under and the
• 18:12
  purchase price or the license price, I have to believe that real work was done by Google to
• 18:17
  make sure that there's an underlying logic and that this should stand up to scrutiny, but
• 18:21
  we're gonna find out.
• 18:22
  I read the piece and then read our other related articles and was like, yeah.
• 18:26
  I assumed that this was going to happen.
• 18:27
  It's all in throes of the AI arms race.
• 18:30
  Everyone's trying to get an edge over the other.
• 18:32
  It's an it's an adjacent thread of thought.
• 18:34
  But over last week, when Google kinda released all the different AI products that we're
• 18:37
  launching, I felt like I probably watched or imagined a bunch of AI founders be like, my
• 18:43
  entire business was crushed in seven hours, basically.
• 18:47
  It's funny though for something like Google, the amount of money that you would need to just
• 18:51
  stand up your own versus just buying feels not that much money at all.
• 18:54
  So it's fascinating that they're going down this route at all.
• 18:56
  I also think it's interesting to me that so much of this is also narrative and vibes.
• 19:02
  I think people for a long time have been like, there is no way any of the incumbents can be
• 19:07
  real threats.
• 19:08
  It's anthropic.
• 19:09
  It's open AI.
• 19:09
  It's mistralop.
• 19:10
  And Google's products are getting better, and their products are starting to show up in
• 19:14
  interesting places.
• 19:15
  And they've got distribution, and I I wouldn't count them out.
• 19:18
  I really wouldn't count them out.
• 19:19
  We have more to come there.
• 19:20
  We are out of time today.
• 19:22
  Thank you, Charles, for joining.
• 19:23
  Thank you to our listeners for jumping in.
• 19:25
  Please like and subscribe, and we'll talk with you all next week.
• 19:28
  Thank you.