AI Startups Are Breaking Fundraising Norms - Here's What VC's Are Watching

For early stage companies, there's always this tension around governance. I don't think most two person companies need a three hour monthly board meeting with minutes and attorneys. I also don't think no governance is the answer for most companies. The question is where do you want to put that slider at any point in time? And what I find is the slider tends to be set closer to the preferences of the people who have power and leverage at the moment.

It's the week of November 3rd. Welcome to the Learning Corner, a Precursor Ventures podcast.

Welcome to the Learning Corner by Precursor. I'm Mia Farnam and I'm joined by our GP and Managing partner Charles Hudson, where we will be talking through the articles and Goodreads that caught our team's attention over the past week. Our first piece is a post on X by Jason Lemkin sharing inside news that he watched three VCC deals this year where quote, the founders just decided to keep the money for themselves and not build the product. Ultimately, Jason said that these were deals that had a lot of momentum and all were in the form of safes where the founders took the position that they owed investors nothing as they never really had true converted equity due to the safe holder agreement. Jason ended the brief post saying that this is a reality of deals in the age of AI, essentially this truly dark side of explosive growth and hype in such concentrated period of time. So wanted to obviously hear your thoughts here, Charles.

It reminded me, I don't know, it was some time ago there was a company called Toptal in the recruiting space that also people got their feathers ruffled because if I remember all the facts right, the CEO was like, hey, I'm just never going to convert these safes, I'm not going to raise the price round. And I always wondered this, you know, if you think about it, when safes first started, we used to do, I mean, I feel old saying this, Mia. We used to do convertible notes. Yeah, and convertible notes had some good features, interest being one of them, like you know, getting something for your money over time. But also convertible notes had a conversion date which meant, hey, if at some point this hasn't converted, at least we have to have a conversation about what we're going to do next. That conversation could be let's just extend the notes, let's just allow them to stay or let's convert them or hey, give me my money back. But there was a forcing function for that conversation that was time based and I think for a lot of good reasons. I think YC said, look These convertible notes, most people don't really want their money back. The convertible note is just sort of this instrument that is cheaper and faster than a priced equity round. And safes make sense so long as you believe there will be a price round at some point in the future that will allow you to become an equity holder. But there are a couple scenarios you need to worry about. One is what if the company never raises a price to equity round and never exits. You end up with this like weird hybrid instrument. Or the more nefarious one is what if the founders take deliberate action to prevent or avoid ever converting because they don't want you to be a shareholder in the company or their control or governance provisions. And safes don't really give you a lot of the traditional investor productions. Yes, you will get your money back in the event of a dissolution or wind down or distribution. But I think it isn't going to be interesting to me to see what happens with these companies that and I don't know enough about the specific situations. Jason didn't go into details, but I can imagine a what if you're a company, you raise 500k on safes, you have some AI product that booms, you get to 10 million an ARR, and you're like, I don't want to raise venture capital. We've got 10 million in the bank, we don't have a new product. You know what I want to do? I actually want to dividend out this cash to common shareholders. And if you had a board or if you had preferred equity, like some of the reasons that we did all this other stuff was for investor protections. In a world where everybody agrees that like the safe is an interim instrument that will be used until the company can get to a price round event, everything works fine. When people start to think about safes as different and figure out like, well, how can I use this to my advantage? There are many ways beyond the ones Jason mentioned that a founder could use safes. But for me, the so what is we're not going back. I don't think we're going back to a world where we use convertible notes. I don't think we're going back to a world unless someone can make the world's cheapest price round process, which I don't, I don't think is going to happen anytime soon, even with standardized docs. I think this is part of the risk you take as an investor when you sign a safe is a belief that the founder has the right intentions about eventually having you become a shareholder. Because that's what most investors want.

We talked internally yesterday about the difference between nefarious activities on behalf of the founder versus just being irresponsible with cash management. Yeah, it's a fine line when you think about it. But then when you're talking about specific examples, I feel like it was easy for me to compartmentalize. Like that's nefarious, that's irresponsible. How do you guess? Hedge where things could possibly head? If you feel like you're already seeing inklings of similar types of behaviors in.

Either spectrum, think for early stage companies. There's always this tension around governance. I don't think most two person companies need a three hour monthly board meeting with minutes and attorneys. It's a gross misuse of time. I also don't think no governance is the answer for most companies unless the founders are incredibly experienced and can be trusted. And the question is, where do you want to put that slider at any point in time? And what I find is the slider tends to be set closer to the preferences of the people who have power and leverage at the moment. So in 2020 there was no governance because founders had all the power. Then in 2022 there was a lot more governance because investors were like I just got burned on a bunch of 100x ARR deals and now I have the power because capital is hard to get. So I'm going to be tighter on this stuff. I'm going to require more of companies reporting and burn ratio and all the blog posts that we wrote that nobody listens to any. And now we're back in a moment where founders have power. And as I talked about to our team, I think there's a really big difference between ambitiously getting an office that's too big for you because you deep down believe you're going to grow into it and you want to send a signal to candidates that we're a company that's going places and spending $50,000 on the corporate Amex at Nobu, taking all of your friends to dinner once a week for a year. Those two things to me are like categorically different uses of both of which you could argue are suboptimal. But one I think is coming from a place of putting the company first and the other one is probably closer to self enrichment.

Well, hopes that none of these deals fall within any of our friends laps. More to come there. Our next piece written by Ruben Dominguez on His substack the VC corner walks through 15 charts that explain how tech and Venture are changing in 2025. He shares 15 different graphics that show concepts, ChatGPT retention patterns, VC career progression and job opening rates mapped to foundational model launches. Ultimately, the piece is created to give a wide range of updates on the current state of venture through a more visual lens. I thought a lot of the charts were really interesting. The first one that jumped out to me was the Vibe Coding Growth Heat map. It listed really well known developer tools like Lovable, Replit, Cursor, Cognition and others on the vertical axis and then mapped across a 12 week timeline and beyond for growth. And a couple call outs stood out. Base 44 saw 900% growth but then cooled off within weeks. Lovable had a similar trajectory with more than 200% early growth but then significant stagnation by week 16. The big takeaway was that developers in this category tend to test pay and then churn faster than almost any other space. So retention is really going to be the key differentiator for whoever wins here. Curious if you had any thoughts on that specific chart there?

I was teaching yesterday and we had a guest lecturer who one of our students asked her the same question. She said we're always asking ourselves is this a sugar high business? And I think that's a good analogy. And right now we're in this very weird moment in Venture where privately every GP I know is we're not sure what the retention curves are going to look like for these products, but it's asymmetric. If these retention curves turn out to be terrible, we will lose our money, but we will have invested in a hot company. If the retention curves turn out to be good, then we've got a winner on our hands. And the asymmetry of being in the winners is so great that a lot of people are like the retention will be what it will be and we'll find out what it is when we find out. In the meantime, let's go. And my experience, and it's anecdata more than anything is the coding tools seem to be a little bit stickier and I think this intuitively makes sense, which is if you're a professional developer, my hunch is you are using an AI code assistant every day as part of your workflow, whether it's cloud code, whether it's cursor, whether it's whatever tool you're using. Whereas my hunch is that the Vibe coding app building things are more episodic and that unless you're a product manager who's constantly building prototypes, you know, I told him the day. It reminds me of like website builders. You know, every time you get web builders, we're going to democratize it and we're going to make everyone able to make a website. Most people don't need a website or if they need a, they have a website, they don't need to update it more than once a year. So a website builder creates more websites, but it doesn't necessarily create more valuable websites. And so I wonder for some of these coding tools, the churn isn't a product issue. It's just most people don't have an ongoing need to build stuff. They built something because they wanted to learn the experience. They built something because it was fun, but they don't need it every day. Whereas I use Claude Desktop 100. I'm like a seven of seven daily active user because it's something I use as part of my workflow all day long, multiple times per day. And I think we're gonna, I think we're gonna. And it's funny, we learn this with mobile too. Eventually we learn, hey, there's certain apps where good usage is daily, there's certain apps that are just episodic and there's nothing you can do by like growth hacking or weird dark arts. It's just that people only have so much need for this app and the question is, when they need it, do they pick yours?

Yeah. I was at a demo day over the past year and I was sitting with friends and we commented, I want to try all these tools, but I don't want to invest in most of them. And I think that's the reality of for every wonderful developer tool that you see in the news that like raises at a massive valuation, there are 20 of them that didn't make it through because it was more episodic or maybe didn't even reach that episodic inflection point at all and didn't have that heat in any way. The second graph that I really liked was around VC career progression, which highlighted that 65% of junior VCs leave the industry within 10 years and only 23% make it to partner. And while I know that some of that is obviously people being pushed out of the space because of performance, I've noticed a surprising number of peers that I know who have left or seriously considering leaving even when that path to partner is really available to them. And I was talking about this with a really close friend who was in venture, then left, went back to operating and we were talking about what this conundrum looks like and the decision Seems to come down to two key questions that we thought about. One, am I at a fund that can attract capital long enough for me to make partner if I earn that spot? And then the second question, which I feel like it's more pervasive, is do I respect and believe in my partners currently enough to actually want to be there long term from a quality of life life standpoint? So I wanted to know if you had any thoughts on this one as well.

I saw that chart and I was like, if you'd asked me what percentage, I would have said 10, not 23. Yes, I would have said it's like much lower. So that's a good thing, I think. I think when it comes to venture, I think we're still like a stone age business when it comes to succession, promotion, all these things. Part of this I think is we still have a surprising number of early stage firms that are on original leadership. And I think we will find out very soon who's good at, you know, some of the bigger firms, Sequoia, Kleiner Perkins, they've had to navigate leadership transitions. I mean, I think it's interesting, the second biggest firm in our industry, Andreessen Horowitz, has never had a significant leadership transition at the top. I don't think Mark or Ben have plans to stop doing what they do anytime soon. So that, that's one thing. And I think what I've learned is, and I think about this a lot, if you're going to run a venture firm, you have to do two things. You have to find really good people and you have to figure out how to create space for them. And sometimes the space that people want is more than the space that's available. And sometimes people get to the top and they go, well, I don't want to sit at that table. I want a different table. I want to be somewhere else. I've been interested in like just the number of particularly female GP friends of mine who've left the industry recently. Some to go to operating, some to go off and do their own venture firms, some just to take a break. And it's funny, the information feedback cycles in venture have gotten far more warped than when I started 20 years ago. 20 years ago, like companies went public, you kind of knew in five to seven years if you were pretty good. Also there was less capital and there was less hype. So yes, we had bubbles and yes, companies got ahead of themselves, but not like today. Today it's really like, I don't know how. If I were like an associate at a firm. I don't know how it would make sense of whether I'm doing a good job. What if I'm in a bunch of buzzy AI companies that are just buzzy but not good? Or conversely, what if I'm in a bunch of good companies that are compounding but nobody cares because they're not the thing du jour. And two years later when I leave it's going to turn out that like those companies were amazing and just nobody knew in the moment. And so I think this is a really big structural industry issue that we haven't really done a great job of training and developing the next generation of leaders. And I think that was fine when this was a cottage industry that didn't manage a lot of money. But I look at the amount of palace intrigue that happens at big PE firms buyout shops for my friends work. Yeah, you could argue it's political and it's maybe not a pleasant process, but it seems to be pretty effective at figuring out who is the relatively small subset of people who are interested in and capable of running these complicated asset management firms that are like running big consultancies or law firms or you're running a big white collar knowledge workforce. I would bet a lot of money that the next 10 years of promotion and talent management adventure will look very different than the last 40.

It just feels like a big re shift and then lump that in with what's happening on I think a lot of my more junior friends who sit at other funds who maybe were like one or two years into venture look around at the current landscape with an AI and say I actually want to go in that right now because I can take the risk. I've got the time. I want to run it this fast. I don't want to spend it being third party. I want to be out there building. And so I'm seeing people leave for that reason. A handful of them are top performers at their fund that I know that their firms. So it's lots of things happening at once. And it's one thing that I know a lot of our friends have talked about is mentorship in venture is such a fascinating concept. I think some of the best men mentors I've ever had have also been at the peer level at other funds. The only people who are living and experiencing venture in the same way you are are probably experiencing in today's world. People who were senior associates five years ago aren't facing what you are today. Our next topic is yet another public company announcing a massive Investment in generative AI Oreo maker Mondelez shared they are investing 40 million into a tool that they have been developing alongside Publicist Group and Accenture to cut ad production costs of marketing content by up to 50%. Like many other CPG in the wave of tariffs and shifting consumer budgets, marketing ad spend balloons and can eat a considerable amount of the bottom line for a company that relies on volume of eyeballs on their product via ads and media. They are going to start leveraging the tool to create social media for Chips Ahoy products and will start leveraging it for product pages on Amazon and Walmart this month. Similarly, competitors like Kraft and Coca Cola have also been public about their use of AI in advertising. In fact, Coke ran an AI created holiday ad last year that drew some criticism from consumers for lacking real emotion. I did actually watch this ad and to me it just felt more like a failure of solid storyboarding than an issue with the use of AI, though those probably are very connected. Is this just another case of a public company trying to signal innovation to shareholders? Or from your perspective, does it reflect a genuine shift towards overhauling cost structures in prep for the future?

I think it can be both. I think in the same way I rolled a lot back in the 2000s. It's like some companies like we're building a website. I'm like, well good for you. You should have a presence on the Internet. You also shouldn't get a cookie for building a website. You should, no pun intended, like with Oreo, but you shouldn't get a cookie. That's just part of business. And I think we're still at the curve where I think companies to some extent want to signal how forward thinking they are in two years. People, if you make the same out, some people are like, well what were you doing for the last two years if you're just not announcing that you're wellness? And I think it's interesting. You always get this cycle with big companies where when the individual products are immature, you go to your favorite consulting partner and have them help you build something that's bespoke, that meets your needs. And when I read it, I winced a little bit, which is not because the firms involved aren't good, but the history of these kind of things is always that they are an interim step toward a better future. And sometimes you have to take that interim step to get to the future. Everyone's like, I want to skip the like consultant, homegrown in house thing. And I'm like, well sometimes you don't get the luxury of skipping that you have to go through that to get to the other side because the thing you want doesn't exist yet.

Well, that was my next question is Accenture is obviously a well known consulting firm. Most people think that it's one of the first businesses to be disrupted by AI consulting. But what I'm seeing is kind of the opposite. You'll see a big AI project announcement and insert your consulting firm here at the helm of it and they're managing the large scale AI transformation projects across multimillion dollar deployments across the world. And so the question is, how sustainable is its model of big consulting firms acting as a go to AI guides for especially when the space itself is still early and evolving. And I've worked in consulting, so I like to know what the job is. If I was in consulting, I would want to be running all these projects too.

I think whenever someone says so and so's job is going to be taken by AI, I'm like, I always ask myself, does this person actually understand what that job is? I think if you think the job of consulting is to produce strategy slide decks, you don't actually understand what consultants do inside of companies. And so if you're just like, oh well, like I could produce this strategy slide deck with ChatGPT, why do I need to hire McKinsey or Deloitte or Accenture? You're not hiring them just for that. Hiring them to do a bunch of other things that are at the. We could, we could have a long talk about political air cover, but sometimes you do need humans to help you manage change management for other humans. And as far as I can tell, there aren't really great AI tools that can do that yet. So I think we're going to continue to find these ironies where people say, oh, AI is going to destroy this category of work and we're going to turn out that it like actually increases the demand for it for some period of time.

Last piece here is more of a news blast. Recent TechCrunch article by Marina Temkin announced that Roloff Botha, the senior steward of Sequoia, is stepping down as of Tuesday. So after serving in that top leadership role for three years, he'll be succeeded by Alfred Lynn and Pat Grady. While Roelof has held senior roles at the firm for years, overseeing the distribution of nearly 50 billion to its investors, he did formally step into the top seat during the venture market downturn in 2022, when portfolio valuations across Sequoia were being slashed. During his tenure, the firm wrote off nearly 200 million due to the FTX collapse, which is kind of pen something to note. And amid geopolitical pressure, spun off its China and India operations. Roloff himself has famously said, quote, we want the triumph of ideas, not the triumph of seniority. But this transition still signals a meaningful shift. It's a reminder that even some of the most iconic firms, they can fall out of favor. And changes are often underway behind the scenes to prepare for the next chapter and protect a legacy. So we obviously don't work for Sequoia, but just felt like it was something that we should bubble up.

In my entire venture career, they're the only firm I can think of that has always been relevant. There are other great firms like that have had, you know, their ups and downs and leadership transitions, but Sequoia has always been relevant. And as someone who runs a firm, it's almost impossible for me to imagine how you have a firm that remains. When I say relevant, I mean in the top three list of names that any founder would want on their cap table, regardless of company stage or strategy. To do that for 50 plus years is really hard. And you only do that if you've built an institution. And the other thing I'll say is this is like a general leadership principle. You don't get to choose the circumstances that are around you when you're a leader. It's the old, I think someone told me, like, being the president is an inbox job. You wake up with your set of priorities and then the world starts putting things in your inbox. And if I think about all of the challenges that he's had to navigate, it's obviously the downturn, liquidity, pressure. They restructured, basically it's more like geeky. The plumbing of how Sequoia works as an entity in the way that capital gets allocated across the firm. You've had some partners with pretty loud political opinions, some of which have caused some amount of polarization both within the firm, it sounds like, and outside of it. And the FTX thing, I think, wasn't so much a financial hit, but more of a reputational hit. And I think that's a lot to deal with in three years. And knowing the people at Sequoia as I do, I'm really confident that they will remain on top and that the new stewards will do a great job. And to me, it's just a reminder of how hard it is to remain on top in this job. It's so hard to do. And they have a remarkable history of doing this.

Well, we are all out of time today. Thank you, Charles, for stopping by. Thank you to all our new listeners and subscribers for popping in. We will talk with you all next week. Please like and subscribe in the meantime. Thank you.

Bye.