Pod Insights Summation with Auren Hoffman SignalFire CEO Chris Farmer on why venture funds now last 20 years and the hiring boom behind the layoffs About this episode Chris Farmer is the founder and CEO of SignalFire, an early stage venture firm managing over $3 billion. In this episode of Summation, Chris and Auren discuss: why the time to return 1x has stretched to 9+ years and funds are lasting 20 yearsthe engineering hiring boom hiding behind the layoff headlinesthe five functions of a VC firm and the one place data helps the leastwhy almost all conventional career advice is now badYou can find Auren Hoffman on X at @auren and Chris Farmer on X a...
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Auren Hoffman? Hello data nerds. My guest today is Chris Farmer. Chris is the founder and CEO of Signalfire, an early-stage VC firm managing over $3 billion. Chris, welcome to Summation.
0:10 Speaker 01 Thanks so much for having me. Great to see you, Warren.
0:12 Auren Hoffman? I'm really excited. Now, what the heck is happening in the venture capital industry? What is going on? I can't make sense of it.
0:20 Speaker 01 Yeah, it's been a very interesting time. You know, the one level you have an incredible excitement, if not euphoria, around everything going on in AI and a handful of companies that are raising unprecedented amounts of money, just unbelievable valuations. And, you know, I think there's a lot of groundswell of real truth and functionality of these technologies that are enabling this. But at the same time, you have lots of legacy companies also that are navigating, having to reinvent themselves. So you're sort of both at this period where the industry hasn't really fully exited, some of the mature companies, whereas like there's been an amazing new acceleration of technology as well that, that in some cases will unlock,
1:04 Speaker 01 you know, the historical companies and in other cases will disrupt. And so there's just a lot of simultaneous chaos going on.
1:13 Auren Hoffman? The industry itself is like, it seems like it's getting more competitive every year. Would you agree with that?
1:18 Speaker 01 Yeah, I mean, I think, you know, when I got into this industry, there was just dozens of firms and now there's something like 5,000 firms. And so, you know, the good news is, or the bad news is there's so many more people competing for opportunities. The good news is this is not just the tech industry anymore. It's basically infused into every industry.
1:37 Auren Hoffman? Yeah, the tech industry is much bigger. So the TAM is bigger. Yeah.
1:41 Speaker 01 Yeah. So, you know, relative to people managing money across all sectors, you know, that maybe hasn't changed overall, but certainly in the venture technology world, it's grown dramatically, but the industry has probably grown much more at the same time. So the IPOs 20 years ago looked like maybe a Series B financing. And today, you know, they're the large, you know, you've got a couple pending IPOs that would be the biggest in history. So it's just a completely different. An order of many orders of magnitude different.
2:12 Auren Hoffman? As a venture capitalist, okay, be my shrink for a second. So, you know, we I have these funds. They have some good DPI, but like, you know, my I still haven't seen like a lot of my carry checks yet. Like, like, do I have to wait like 12 years for me to get like my carry checks coming in? Or how does it work like?
2:30 Speaker 01 I hope you get to only have to wait 12 years. In the case of SpaceX, it's what 24 now. And I think the the whole period for venture funds to get to like DPI one was more like seven to nine years before to return the original cost of capital. So now it's looking way beyond that. I mean, and a lot of these these funds are now lasting 16 to 20 years. I I I have a race from my seat check into Stripe, whether my daughter will graduate high school before I get any liquidity from that investment. So it is definitely a long game.
3:07 Auren Hoffman? That's a relatively liquid position. So you could sell it if you chose to, right?
3:11 Speaker 01 Yeah, well, actually, it's my old firm, General Catalyst, that controls that decision. And they have not been not been willing to sell it.
3:18 Auren Hoffman? Yeah, they could choose to if they if they choose to, but like, you know, just want to compound, right?
3:22 Speaker 01 Yeah, I mean, I think the, you know, there's a relatively small number of companies that are driving these like unbelievable potential exits. And I think people are very reticent to to sell their their best companies for half the potential terminal value. And there's some disincentives because there's not a a time clock against like a hurdle rate that you've got to clear. Right? And so as these things continue to compound, even if the growth has slowed relative to its early years, it drives really large dollars. And so there's definitely some misalignment on the LPGP side. And then the founders who don't want to necessarily be having to think short term quarter to quarter
4:01 Speaker 01 and are staying private longer. And then so much capital has flown into the industry on the private side that you're now able to raise, you know, hundreds of billions of dollars as a private company if the performance is there. And so there's there's I think companies are just going to stay private longer. So Nature of venture has to change because I don't think any LP wants to invest in a product that takes eighteen, twenty years to sort of to terminate a fund.
4:31 Auren Hoffman? As a venture capitalist, I always just think, I always thought about just being a buyer. I never thought about being a seller and then, you know, I would just eventually a company would either sell or go public or maybe go under or whatever. There would be some sort of exit. And then I would just take the money from the exit. Now, really for the first time in the last maybe one or two years, we're starting to think about how, should we be a seller? Okay, this company just raised at a ten billion valuation. Should we be, should we be selling into that? Should we, it's not something I ever even, it wasn't even something I was good at, or maybe I'm still not good at that. Like, how do you think about those things?
5:08 Speaker 01 Yeah, and that's actually been one of the, the main things that we talk to our LPs about this year on our annual general meeting. Right? As we move to this active management posture. And now we want to do that in concert with the LPs in an LP and, and both founder friendly ways. So in many cases, we'd love our LP to be the buyer, right? And, and, and basically give some liquidity to the early fund and then continue to be net long the company. In fact, maybe even vest parata on top of getting liquidity. But just, it's really about the duration at the product level, not about conviction in the company. And I think the companies are having to think about this as well because the same is true, if not even more, in the more extreme sense for early employees.
5:51 Speaker 01 Right? The, the founders themselves are taking money off the table because they'd actually like to buy a house at some point and de-risk and diversify and early employees need to exercise options and, and also have life events they're going to need money for. And, and so I, I think if you're going to stay private for many years and, and companies like Stripe may choose to stay private for, for, for much longer than twenty years. And so i- in that scenario, we need to get better at creating mechanisms for people to reasonably diversify their holdings, get liquidity for, for, for buying a house or paying for schools or whatever it is. And, and I think that's just becoming part of a maturing market
6:32 Speaker 01 where, you know, there are some real advantages to being, remaining private and some founders are choosing to do that. Well, then concurrent with that becomes sort of an obligation to.
6:43 Speaker 01 To take the early believers and workers from your company and make this work for them as well.
6:49 Auren Hoffman? I saw a stat that said that the average exit is much higher, but the median exit is much lower. The tails really are driving everything. Do you agree with that? If so, like, how does that change the way we should be investing?
7:05 Speaker 01 It's really challenging because you have a lot of companies that have been private for a long time, and they're not growing at a rate that they're going to be able to go public.
7:14 Auren Hoffman? What do we do with those companies? Some of those companies may be worth less than the money raised.
7:18 Speaker 01 Yes, and so, and then you have, to your point, you have a lot of challenges on the preference stack and how much money these companies have raised and whether there's any residual value on the common equity side for employees and whatnot. And, you know, I think we've seen talent acquisitions where there's almost no value going to the cap table aside from employee retention. And especially with with model companies where there's a data model element to it. The acquiring company may have no interest in the underlying product or model that would become obsolete if you didn't maintain it. And so they just care about the team. And so I think you're seeing more private to private combinations. All right, so some of them are companies like Stripe or Databricks
8:02 Speaker 01 that are doing billion dollar plus acquisitions that would have had to be public to have the currency to do that, and that's no longer true. And and people from one company may be trading into equity of another company, and it may not have even a heavy cash component. Or, again, requiring some form of liquidity because that's just changing one currency to another. And if people eventually need to return capital of their LPs, they may not even, the company that they back may not exist anymore, and now they need to be able to sell. And then, secondly, we're seeing some companies that have disrupted on the AI side start to look at buying legacy companies from the prior generation that have a big customer base or systems of record or whatever it is.
8:44 Speaker 01 And you see consolidation of the new generation of the old generation, right? Because a lot of these are private equity backed companies that may not. Have the exit path that they thought they had either because of the level of disruption. And this is the SaaS apocalypse and the private credit withdrawals and.
9:00 Auren Hoffman? Yeah, some of these private equity PE buys were clearly overvalued.
9:04 Speaker 01 Yeah, so I think what's going to happen is you'll have some companies that just organically grow and get to a huge scale. And then you'll see a lot more private to private combinations of various kinds. In order to get to critical mass and something that's a profitable enduring business. And so I think, you know, the long tail of companies has to figure out. How to be a valuable component of one of those types of things, whether the secondary liquidity will be a component of it. Combinations of private to private will be another combination. Obviously, public acquisitions will continue to be a part of it, but that's maybe less so if you're not AI native. You know, that may be a difficult combination to do. And so it's going to be really interesting. And then there's just a sort of pure organic power law companies.
9:49 Speaker 01 And so, but I think obviously the industry has to figure out, they have to get very good at secondary liquidity, but also very good at getting these companies to scale and profitability in a configuration. That allows them to. Capture some of the value and dollars that were put into funding these companies. And they can't just leave them as sort of walking dead on the other end. And I think private equity has had that skill set, but venture has not historically had that skill set. So I think you see a fusion of these industries and playbooks from different sectors. Really changed the game. You're seeing like more financial engineering for larger private companies. Versus using equity for everything.
10:29 Auren Hoffman? I have a friend who's got a very good company. Profitable, doing fine, but you know, the company is worth less than the 150 million. You know, probably that went into the company. Maybe he's doing 30, let's say 30 million in revenue or something. I'm a, you know, maybe it's a 5X if he's lucky. But he's got this problem. He's got many venture capitalists in there. They're, they have different marks in the font, in their fund. Some of them are market low, but some of them still have the same 1X mark or even higher mark that they had before from future rounds. Like, he's, he's just like, has to go along with it until, I don't know, maybe they all market down or something. Like, I like, what would your advice be to this guy?
11:15 Speaker 01 Yeah, so I've run into this situation a number of times. I think at some point, you've got to get to a level of capitulation where somebody who's, that came in later, that has the perception of, hey, I'm first dollars out, I've got preference. But if, if the other stakeholders don't get anything out of it, whether it's the employees, most importantly, but also the early financial investors, if there's not some sharing of the economic pie, you know, you just end up being gridlocked. And so you've seen this with management carve-outs, right, that are incentivized on just finding the right exit. Obviously, the acquirer may have an employee retention component of this, but I think
11:58 Speaker 01 also negotiating with the other players on the cap table, whether it be debt or equity, where it's like, okay, well, we'll give you some slice of, of the, you know, of one of the proceeds of some sort of exit, and we won't take all of it at the top of the stack so that we get you, if you have blocking rights or whatever it is, to be incentivized to just find a, you know, a graceful way to sort of, you know, find a home for the company. And so I think it's, you know, but it's a complex negotiation with very different incentives, and it's hard.
12:34 Auren Hoffman? Super complex, because there's, and people are very, it's hard to even know sometimes their incentives. And sometimes the partner at the firm has different incentives than the fund. And so, yeah, it's like, so it's, even within their fund, they may have a negotiation that has to be done as well.
12:50 Speaker 01 Yeah, you know, and I think you're seeing some of these continuation vehicles, where a fund's natural life is 10 years plus two years per the contract. Many of them are getting extended long beyond that. But at some point, LPs say, you know, just liquidate the final component of the fund. Other LPs say, no, I want to, you know, ride this to exit and don't want to sell for 50 cents on the dollar. If a secondary buyer needs 2X plus. And so then you have these, what's called a continuation vehicle that people can either roll forward or take the cash out, but you've got to find a buyer on the other side that's willing to pay out the ones that want to sell. But they may be not willing to pay much of any value for some of those companies that you're talking about.
13:35 Speaker 01 And it's really one, two, three companies in the fund that they're, they really covet and they write everything else effectively down to zero. And so, you know, these are all complex negotiations. And as a result of this, relative to how these companies are being held in these funds, a lot of these continuation funds break, even with an investment banker trying to facilitate this process. And so, you know, I think we need to get to a point of where people don't have a choice anymore. Or they just, you know, acknowledge and accept that this is the reality of the market and they've got to reprice the way the companies are being held. Because the economic reality doesn't match the, the, the paper reality of, of, of what they're presenting of the value of that portfolio.
14:19 Auren Hoffman? People may not know, but you kind of come with like a talent lens. That's how you kind of originally really got started in this industry. What is going on in the engineering talent market right now? People are saying, okay, the people want to hire more senior engineers, but then, you know, there's still people who want to hire a lot of these people right out of college. Like what's, what's going on right now?
14:40 Speaker 01 So there's a lot of interesting things going on. It's very nuanced. So at one level, there's just an absolute explosion of startups and growth across the tech world and lots of capital going in. So in general, there's lots of companies being started and lots of hiring going around that. Some of the larger, more mature companies are, you know, hired a lot in recent years. And, you know, there's been arguments that they're, they over hired and have sort of bloated. Headcount and they're desperate to free up capital for all this capital expenditure. That's going out for infrastructure build out or whatever. And you've seen this with Oracle and Meta and a bunch of companies, Block, et cetera. That have done pretty substantial layoffs.
15:23 Auren Hoffman? And so everyone had like very high salaries too. So if you consider a stock as part of the compensation. Like a lot, some of these engineers were making like between 500,000 and $5 million a year or something that were laid off.
15:36 Speaker 01 So there's definitely a debate that we have not been great at having a dialogue for the rest of the industry is you're scaring everyone with layoffs. But at the same time, the number of engineers being hired is actually up. You know, that said, it depends on who and where also. So like front end engineering has gotten easier to do with these coding platforms. Like Claude or Cursor or whatnot, that that's gotten easier.
16:02 Auren Hoffman? But those were never the high priced engineers anyway, right?
16:06 Speaker 01 You know, and I think there's, but there's now more of an emphasis on people who can be full stack. And can be very good at the data side of things. So I think there's a premium on the data side. Maybe there's a even a further degradation of the front end. Engineers, because some of that has been automated and simplified. But at the flip side, going back to that point we made earlier, that the tech industry is now every industry. And whether you're in retail or banking or, you know, construction, there's just, you've seen automation across industries that were historically very different. And you're actually going to see a lot of hiring of engineering talent into that or forward deployed engineers or people to help them navigate.
16:46 Speaker 01 From being a legacy company to, to having a bridge into the AI world. So there's been a shift in who is hiring. There's been a shift in what the roles look like. There's obviously been a fair amount of reinvention of what people call themselves and how they describe their previous roles. So there's, there's a lot more experience in machine learning than used to be on people's resumes. Which is always an interesting thing as they try to reposition themselves for the future. And then one unfortunate component as well is that some of the bottom rungs. Hiring right out of undergrad, those types of things has gotten more difficult because those junior engineers, a lot of the AI performs some of the functions that you
17:27 Speaker 01 might have had there. So it'll be interesting to see how universities bridge to having graduates be able to hit the ground running on day one. And because the bar is getting higher and higher as these platforms get better and better at coding and on more different use cases. And so, you know, but that said, you know, there's programs like Neo and Zefello's that are also have like an explosion of young kids who are coming from an AI native perspective and looking to reinvent these industries. And it may be a very Darwinian sort of, you know, lots of people experimenting lots of different things, but they're going to come up with things that the
18:09 Speaker 01 experienced people from the industry would never invent, like, you know, no one from the hotel industry was going to build Airbnb, right? Or from the telecom industry was going to build Snapchat, right? And so you're gonna, I think you're gonna see lots of interesting companies there that get formed. And if you're a high quality engineer, you build experience and then you may get absorbed into one of those companies that's growing. And that's long been a thing at Y Combinator, for example, where the, the companies that don't work often end up joining the companies that do, and that's the path. Lots of people experiment. So I think it's, it's not one trend, it's a whole series of trend. But I think at the end of the day, technology is, is become a much bigger
18:49 Speaker 01 industry. It's infused into every industry and the surface area of people who know how to use these platforms can do much more with fewer resources. And, you know, I think, you know, if you can, if you can keep learning and stay ahead of it, there's a lot of opportunity. And if you don't and try to stick to the old ways of doing things, It's going to be a very rough road.
19:12 Auren Hoffman? You guys use a lot of data and one of the things you try to figure out is like, when is an engineer ripe to leave? You know, there's the obvious ones. They've been there for four years at the company and they're fairly invested. But what are some non-obvious things like, you know, in the tech labor market that you track?
19:31 Speaker 01 Yeah, so we track many, many things that are sort of intuitive. And then obviously, a lot of it is machine learned just from looking at all sorts of facets of what's going on. And so, you know, certainly a public company's stock price factors into those things, whether your options are underwater.
19:48 Auren Hoffman? You're going to be a loose in the saddle if you're at one of these SaaS companies that has gone down 75% or something like that.
19:56 Speaker 01 If, for example, the sales team starts to leave a company and go to new generations of technology or new competitors, et cetera, that's an early warning signal that maybe the product market fit is degrading or no longer there. And that will sort of lead to other parts of the organization that may not be as close to the front line of that customer conversation following suit. So different roles, you know, basically leave at different paces based on how much visibility they have into that and also frankly, how much agency they have. So certain people in certain roles on the engineering side have a very high opportunity cost right now and can get amazing packages if they have the right AI skill set or whatever
20:41 Speaker 01 it is. And so the inflation of the value of their skill set also factors into their agency to take new roles. And then also, there's a lot of like more personal elements. Leadership changes, like you may have followed somebody to the company and if they leave, it can cause departure of the folks that, you know, joined the company in order to work with them. And then obviously, there's a lot that's going on on the restructuring side, as we mentioned, with layoffs. And so as they feel like they may not be well situated because the company is starting to make cuts, that can be a forcing function for people to try to get Get ahead of it before there's a mass layoff. And so there's a lot of forces that go into it. It's a very complex set of decisions.
21:26 Speaker 01 There's certainly behavioral things that people do like start to polish up their online resumes, so to speak, that are an early signal. And if you see that at large volume across a company, that is obviously a very predictive signal, but but there's many permutations. If one company, one competitor is breaking out relative to to you, that that positioning of being one of the first one or two lead horses in the race is much more likely to drive a lot of economic value. If you start to fall back to the fourth, the fifth, the sixth player in that ecosystem, you start to see a lot more attrition at the companies as as as people want to get on the on the winning team. So, you know, there's lots of different dynamics.
22:09 Auren Hoffman? I mean, you you guys are a very data driven VC. How far could we take that? Like, could we could we have a bot basically make a decision at C in Series A just by looking at the signals and the LinkedIn profiles and reviewing the deck and a few other things, like not actually having to, you know, go through the presentation? Like, could could a bot do as well as humans?
22:34 Speaker 01 So from a picking standpoint, I think the answer is yes.
22:39 Auren Hoffman? Yeah, winning, no, obviously, you the winning like the bot wouldn't win as well. But the picking.
22:44 Speaker 01 At the end of the day, though, I don't think most people want to be backed by a bot.
22:49 Auren Hoffman? In some ways, you could have like a great relationship person who isn't good at evaluating, and then the bot in the backend is making the decision. And whereas you're usually today, we have like the opposite, like the person meeting with the companies, maybe not that great at relationships or, you know, et cetera. Like, you could see a new type of venture capitalist who's just like great at building those long term relationships with the founders.
23:12 Speaker 01 I think you'll see a lot of automation and other layers of the stack. So, for example, venture debt or those types of things could be underwritten much more quickly. I think you've seen things like the Consumer Value Fund, a general catalyst that's able to look at the cohort data and underwrite basically lending for customer acquisition based on those streams. And so, you know, I think the the earlier you are at inception, the more somebody wants a partner to be a builder. And I mean, there may be some founders that choose not to go that path and to to just like, hey, I just want to be left alone. Certainly, there are many funds like Founders Fund and whatnot. I've sort of taken more of
23:52 Speaker 01 that sort of worldview. There are other founders that want a human on the other side, where there's an emotional connection as well and a buy-in to the vision and not just it's going to bail on you because your metric this month didn't work and you may choose to pivot or retune the team or whatever it is, or your go-to-market approach. But it's, you know, most companies are a little bit of a roller coaster journey, even if they're up into the right. And so you want someone who is buying in at a higher level than just purely metrics. So we've always believed that data can do some things better than people.
24:33 Speaker 01 You can track a scale of people and companies that you couldn't do manually. But then there's also a very important human component on taking signal from where the world is going, not where the world has been, interpreting that in a way that so far humans are able to do better than machines. And then also that emotional, personal connection that are all really important components of it. But that said, the later and later stage you get, the more that you could probably just purely use data-driven approaches into underwriting a company or debt or other types of financial engineering. And I think you'll see more and more of that. So I do think parts of this can be automated. And you've already seen
25:18 Speaker 01 that in the world with companies like Square, Square Capital Block, using POS data, point-of-sale data, in order to understand the health and what people are buying down to the SKU level visibility. And if you keep relatively constant labor costs and input costs, you can understand whether you can continue to lend money to that bakery or whatever it is. And so I think you'll see more of that.
25:41 Auren Hoffman? The first meeting is this very interesting meeting when you have with a founder where they're getting you excited, you're getting them excited about your fund, you're kind of sharing some things, but maybe they're not ready to share everything with you because you have to build that trust. Maybe you're not willing to share everything with them. Them, could we see a first meeting that's agent to agent? And if that goes well, the next one is human to human?
26:04 Speaker 01 Whether it's agent to agent or better matching, right? Where there's some sort of matchmaking algorithm in between versus a full blown interview. But I do think there's just a tremendous amount of inefficiency and friction in the fundraising process, both on, I mean, I feel as a venture capitalist in talking to LPs.
26:23 Auren Hoffman? Right, LPs too, good point. You could do the first, the first meeting with LPs should be agent to agent too, right? Yeah, or something, yeah.
26:29 Speaker 01 Same is true for VCs. They may have a conflict in another company that's too close to you, or they may have already made several investments in a particular vertical or type of business, then they, they have enough or maybe even too much exposure already, and they wouldn't do something incrementally. A lot of times you'll see, oh, well, they've already made investments in this space. Well, that actually may mean they're less than likely to make additional ones as opposed to more. And so all these things, like, I think some of those types of matching could make sense where there's opportunities to be more efficient, you know? And certainly I get lots of things over the transom that are sent to me where it's out of geography or out of a sector that we invest in or a business model that we wouldn't do. And it's just poorly filtered or targeted to us.
27:15 Speaker 01 And so, you know, I think you'll see an agent respond, right? You know, on some of these types of things to be more founder friendly and responsive and more time efficient for them. And so there may be some back and forth there. So there, there may be elements of that, but I think this world is so competitive that it's going to be around how do you create the best founder experience? How do you be, be more efficient with each other's time? And, and then, but that it'll still come down to, at least to the very early phases of a company, that's human interaction and that belief in your vision and those types of things versus some analytic, you know, that, that is around some metric or whatever. Cause these things change so fast, both externally and internally to these companies
27:56 Speaker 01 that I think it, the, the earliest phases are less metrics driven and more, you know, more of a magnetic pull to a founder and an idea and shared philosophies and worldviews than they are, you know, negotiated in spreadsheets.
28:13 Auren Hoffman? How do you, how do you think of the LP as the customer? We've talked a lot about the founders and stuff like that. LPs are, are they looking to have a series of venture investments that are uncorrelated? Are they trying to reduce risk or are they trying to maximize risk? And they're okay if some, you know, some of these have a zero X or something like, how, what's, what is their perspective and how do you think about them?
28:41 Speaker 01 Well, there are many different LPs that function very differently. And so, you know, in the case of, say, a fund of funds that may be investing in dozens of venture firms, they're already going to get a tremendous amount of diversity. Right? There are some LPs that want to have just fewer relationships that go deeper, that may care more about concentration risk. And the way you've seen that is the core fund, maybe one product, and then a lot more is done through co-investment or opportunity funds or SPVs that allow people to continue to concentrate ownership or in other cases, maintain diversity. And, and so you've seen just his concentration that goes well beyond what funds used to do
29:27 Speaker 01 and very different portfolio constructions over time. And so I think it's much more of a partnership where certain LPs want to have fewer manager relationships, but a much more intimate and, and sort of almost like co-developed relationship. And then other people want something that's productized, that I just write you a check and you take care of everything.
29:50 Speaker 01 And it just depends on their level of sophistication, the maturity of their venture portfolio. But, and I think also people are looking at this very differently today. Venture historically was my high-risk category. And today it's also seen as a hedge against the legacy portfolio. So if I have a much larger software buyout portfolio, I need to be invested in AI just as a counterbalance, not just as my lottery ticket to, to drive returns. And then other LPs are looking for much more systematic, repeatable, compounding type of relationships. They don't need the grand slams and then the strikeouts. They want consistent, repeatable returns. And that's a lot of frankly, what we do is we work with some very large LPs
30:35 Speaker 01 that want an institutional relationship that's much more, that's systematic, repeatable, et cetera. Because They're a huge public pension and they need a relationship for decades. They can't have somebody who has a hot hand for a few funds because they're an operator that just started the seed fund and then they're going to retire. And now they need to go find another related. They just don't have the resources to constantly be trying to pick the managers. Whereas some family offices, that's the whole sport. It's like, oh, I want to make bets on, I want to go into this domain or I really like that person and whatever. And they're looking to just maximize returns, even if it's not repeatable. And that's part of how they want to allocate capital. So very different personas, depending on who the counterparty is on the LP side.
31:20 Auren Hoffman? At Flex, a very high percentage of our LPs, a very high percentage of our capital is from US taxpayers. Whereas for most funds, most of their LPs are endowments or pension funds or overseas. They're not paying taxes, so they don't have to think about that. Where for us, like, we're constantly trying to optimize the tax. You know, okay, can we get in, can we make sure it's QSBS? Can we make sure this? Can we make sure that? We're constantly trying to optimize the tax piece because we know that could add another X or so to our returns to our LPs. So you have to think about who your LPs are when you're making the investments.
31:56 Speaker 01 Yeah, that's obviously a real advantage of seed for a family office is there's some significant important tax advantages to investing in what is the innovation economy here. But as you start to get to later stage, that doesn't factor in. And some of these institutions, it's about, you know, that, you know, driving above sort of threshold returns that help endow their pension or the endowment that they're in the important causes that they're supporting and because they're a tax-free or low-tax entity. So like very different incentives, structures and tax regimes on there and that all impact how they behave and how they allocate and construct their portfolio.
32:41 Auren Hoffman? Now, I think you wrote down, there's like five functions of a VC firm. There's sourcing, picking, winning, support and construction. Which one does data help the least?
32:51 Speaker 01 I'd say probably the winning element. At the end of the day, it's connecting at like a heart level. At the earliest phases, you know, maybe on the secondary side or other areas, you're able to move faster or, you know, price better, more efficiently.
33:07 Auren Hoffman? Right. And they have to choose you and they may have many choices. It's not like you're the only one. Usually the deals these are, is there not non, complete non-consensus deals? Like there's often more capital than is available on the round, chasing the round.
33:22 Speaker 01 And they're ultimately making the decision of who is the best partner for the long run, who's going to help me maximize the value of this and who will I enjoy working with on a day-to-day basis and those types of things. Now that said, we use data as a component of winning because we, for example, have built an AI recruiting platform that's much more scalable. And so instead of randomly throwing a few resumes across the transom periodically, we have a very systematic approach of supporting founders at scale.
33:50 Auren Hoffman? So it's almost like, here's a suite of services we can add. I can add my, I can give you my personality, Chris' personality, which is great, but I can also give you these other six things that will come with my personality.
34:02 Speaker 01 Yeah, I can give you a platform that every recruiter at the company can be on this and use it for competitive intelligence, use it for go-to-market, use it for recruiting. Those are the types of things that complement the human component because they're much more scalable and we can do it no matter how large the check size is because our variable cost is near zero to support an additional company. And so we have a whole suite of things that are super scalable, no matter whether we wrote a hundred thousand dollar check or, or a much larger check downstream.
34:31 Auren Hoffman? Now, when I'm evaluating talent today, what I really care about is like, what did they build? What did they do? And it could easily just in their spare time, like they built something to manage their, you know, their lighting in their house or something or whatever. The problem is like evaluating what they built could take 10 to 15 minutes or something. Whereas if you just want to look at a resume, you see, okay, they went to this school, they have this degree, they worked at Google, it's, it's, it's super fast and you could even like automate a lot of that. Is there a way of like better automating, like to see what they built? Or is that just like so bespoke that you just have to put in the time to, to go do it?
35:13 Speaker 01 Some people are looking at, you know, their tokens that they've used, what their technical stack is. So how fluent are they on the most? So it's not just what they built, but how they built it.
35:24 Speaker 01 I think there's also things like open source.
35:25 Auren Hoffman? Certainly, you can like evaluate their GitHub stuff pretty quickly sometimes.
35:29 Speaker 01 Yeah, we've built a lot of automation to do those types of things. And I do think there's a lot of companies in the talent AI space that are, you know, finding ways to do, you know, effect equivalent of like paired programming that you might have done to see how they actually put that sort of background and knowledge to practice in a real context. And so, you know, I think though, there'll be a lot of opportunities to get closer to that of real world implementations and that interviews will change fundamentally from, you know, just asking a bunch of questions based on bullet points on a resume and be much more tangible. And I also push on universities on this.
36:12 Speaker 01 People need to build portfolios and have real experience so they can hit the ground running because if that bottom rung is, is gone where it's like, oh, just raw talent that I'm now going to train. If you have to have already almost been trained, right? That is, you know, you can have your resume and your GPA, but you also need to have your portfolio of things that you have built to your point.
36:33 Auren Hoffman? Have you have kids in high school or those kids, you know, if you're giving advice to kids in high school or kids in college, like how, what advice would you be giving them?
36:40 Speaker 01 I think the importance of applied learning is, is, is much higher today because I think fewer people are hiring just, and so that's internships, hackathons, various projects, things that they can show that they can hit the ground running and, and go a level higher. And so I think you're going to need to see an infusion. That's my hypothesis, we'll see. But of, of education stretching into that vocational learning in whatever the function so that you don't just have a raw talent with a degree, but you have somebody who can be productive on day one and, and frankly, start at the second or third bar and that maybe they don't get paid, you know, for that, that sort of the learning to that first
37:25 Speaker 01 bar. It becomes part of their educational package and, you know, there may be partnerships with corporations or whatever that, you know, and you've seen some, you know, it's interesting. My daughter is doing some interviewing and, you know, they have online, you know, interviews that they're doing where you actually have to do work. And there's some companies that have created modules that replicate the actual work, right? And these are free things that you can, you can go through. As part of your application process and then you get scored. And so you're starting to see people build up those and you can now put it on your resume. And I think very much like people put their GitHub profile, you know, on their resume today.
38:04 Auren Hoffman? You see someone like they've got this like they're writing, it could even mean they're writing something on the Substack or they created a game and you look at their game and it's like, oh, that's kind of cool. Or they created this cool video about something or whatever it is. It's just cool to see what they, they did something, they actually built something. I mean, back in the day it would be like you, you, you started a club at your school or something like that, right? Or I don't know, and it could be a club to play squash, but at least it was like you had to organize people and you had to get the, like that shows a lot of, of gumption to be able to do those things.
38:36 Speaker 01 And I think, and you'll be able to use AI tools in order to, to audit like the technical choices you may have made or the code quality or whatever the work output is, you know, whether it's a written format for marketing or, or a campaign that you did and how effective that was. I think you'll be able to get a much better quantitative sense to sort of pair with the qualitative elements that, that someone has on a resume. And that'll become a more important part of recruiting and, and getting jobs over time and building, building a, building a resume will include, you know, those, those types of portfolios.
39:08 Auren Hoffman? How are you personally using AI like in your personal life?
39:14 Speaker 01 For me, I'm launching some new products and the ability to, that have some complex and actually it was a fun product. So when you think about tax implications, structuring, all these types of things, it's amazing how you can use AI to be like, okay, wear the, wear the hat of a tax attorney, wear the hat of an accountant, wear the hat of, of this type of LP or that type of LP. And you can, the pace at which you can iterate on what would have been many months of conversations episodically with these different personas, some of which would be paid, some of which would be networked, et cetera. You could now dramatically accelerate that while you're in the car commuting to work
39:57 Speaker 01 and you're riffing on tax law or whatever, which I live a very exciting life. And the pace of iteration is just incredible. And it just like, you know, many tasks have been accelerated. This really allows me to iterate and go down all sorts of rabbit holes. The other thing is like, just in processing news. I love, you know, getting like AI summaries of all the stuff going on in the news. And then like, like a perplexity I use for this, and then I can dig in and go really deep on something that wouldn't have been in a newspaper article. So I can get a summation of it. But then I can take it down different dimensions that wouldn't have been in, in sort of a piece of news historically.
40:40 Speaker 01 And so you can both go shallow and deep at the same time, whatever interests you or might be relevant.
40:46 Auren Hoffman? What I love about it is it's just like, it's almost like you need the idea to do it. Like I was, I was just thinking about this. I said, tomorrow, I brought up a little time on my calendar. Tomorrow, I am just gonna, I'm just gonna, I'm creating an email address and all my like stupid Substacks and gazillion other random newsletters that I have, which already get filtered in some folder. They're all gonna go to this email. It's gonna repurpose them. It's gonna redo it. It's gonna learn all my stuff. And it's just gonna help me learn much faster, right? And it's like, that's something like you would never do in the past. It would just take too long to do it. And it's like so unique, like what I want is so unique to me. No one else is gonna want this thing. Right? It's gonna be, it's just gonna be for Warren.
41:26 Auren Hoffman? But like in an hour, I can just delight myself for the next five years of having this like awesome thing for me.
41:32 Speaker 01 Yeah. I mean, look, they're incredibly powerful tools and it just depends on how, how much agency you have. There are people who will use these tools to, to sort of take shortcuts to getting to answers and, and not go super deep. And there'll be people who will use it to follow every ambition and curiosity that they have. And I think it's, it's ultimately for that latter persona, like a massive unlock and, and democratizing force on education and access, right? And for other people. You know, it, you know, it may, it may stunt the sort of their, their first, first principles learning. And so it's just a, you know, it's an incredible tool, but I think people really
42:16 Speaker 01 need to lean in to, to the change and the opportunity that's around that. And, you know, if you do that, I think it's a huge unlock of opportunity. And if you don't, it, it could be a very threatening sort of, you know, component of your life that, that makes, you know, career paths much more stunted. So, you know, I, I think it cuts both ways, but it's, for me, as somebody who's sort of insatiably curious, it's a, it's just a really exciting moment.
42:45 Auren Hoffman? Yeah. And it allows you to be action oriented, which is super cool and super fun. Last question we ask all of our guests, what conventional wisdom or advice is generally bad advice? And you're not allowed to say follow your passion because everyone else, we already have like 30 other people have said that.
43:00 Speaker 01 Unfortunately, I think almost all the general advice is bad advice.
43:04 Auren Hoffman? And is that just because that's always been true? Or is that because we're in this unique time right now?
43:09 Speaker 01 Particularly now, that I think that you used to, you want to follow a major and everyone has like this very specific tool, toolkit that they graduate with and they all look the same. So an employer could be like, I'm hiring an accountant, I'm hiring a lawyer, I'm hiring a this, a that, a whatever. And I think that is a complicated path today. I think the most interesting thing or the intersection of two things that don't look like other roles and it makes you unique and therefore opens up a lot of opportunities for the right type of, of things. So I think people need to find non-traditional combinations and that almost all the traditional combinations are actually going to be more limiting today.
43:53 Speaker 01 Whereas before it, it meant that you hit a set, a standard.
43:57 Auren Hoffman? Yeah.
43:58 Speaker 01 Right? Today, too many people may have exactly the same profile and that, that make, make it harder to stand out.
44:05 Auren Hoffman? It also might you not just stand out, but it must allow you to understand things better in a way, you know, these two overlapping Venn diagrams gives you a unique perspective, unique perspective on something that other people just don't have.
44:17 Speaker 01 A hundred percent as a founder. I mean, I just was talking to some founders and it's the intersection of two things that haven't historically been inside a single organization. That is, that is a huge opportunity. And I think it's almost like this, you being bilingual, right? I speak.
44:32 Speaker 01 You know, I'm a, we had a founder who was like a professional game developer, but also was, you know, a pediatric surgeon, right? And then he created something at the intersection of like VR applied to surgical training.
44:45 Auren Hoffman? The beauty is you could be like the thousandth best, you know, game developer, not even, you could be like the, the hundred thousandth best game developer and the 10,000 best pediatric surgeon, right? So you're not like a winner in either field, but you put them together and you are by far the best in that particular, you know, zone. By the way, this was incredible. Thank you very much. We, we had a little technical difficulty at the end that we will edit out. But thank you very much, Chris Farmer, for joining us on Summation. I, by the way, I follow you at Chris W. Farmer on X. I definitely encourage our listeners to engage with you there. This has been just a ton of fun.
45:23 Speaker 01 Awesome. Thanks so much. Same for him or you. And it's always great to chat and you're always a unique thinker and appreciate you, you sent up this time.
45:29 Auren Hoffman? One more thing before we go. I read a blog called Summation. It's the same as this podcast. The blog is about non-obvious idea sharing on business, talent, data, longevity, and random contrarian takes. If you like the conversations on this show, you'll probably like the blog. It's free. New content comes out twice a month. You could subscribe at orin.substack.com. That's orin.substack.com.
Previous GTMnow Podcast | Your First VC Meeting Will Be Agent-to-Agent Next Charles Duhigg on the writing AI can't touch, what GLP-1s really do to your brain, and the serendipity tax Page updated October 30, 2026 · transcript and summary are machine-generated