DN Briefing #1: Character AI Goes Bust, Oura Pulls IPO & Airtable's 55x vs 1x (w/ Martin Tobias)

Martin Tobias - Incisive Ventu:The other lesson is that the the only people that made a lot of money were the people who wrote the seed check who was r were my friends. And they made 55 times their money. The last three round investors at Airtable who went into the C D E got 1X their money.

Niklas:Hi and huge welcome to you, my lovely listener. So glad you're here. Today you are joining me for a talk with Martin about what's going on in the venture landscape. And boy are things happening. I just read that character AI has filed for bankruptcy. Martin, so nice to have you first of all.

Martin Tobias - Incisive Ventu:Hi, thanks for having me, Nicholas.

Niklas:Yeah, so character AI, for those who are not particularly familiar with it, it's actually one of a company started by one of the original authors of the Transformer paper. They built like a wrapper around LLMs. You could talk to your favorite like famous people or some somebody like that. They had a very large number of users, I think around twenty million at at the peak. Then Google bought a license, I think, for something like two billion back in the day, hired back the original guys from character AI, and now they are bankrupt. They were a rapper. what does it mean? What's your feeling and how does it relate to the general venture space here?

Martin Tobias - Incisive Ventu:Well, I I think it comments on two things which I've said before on X and in my LP updates. One is I am not sure, in fact, I'm fairly certain that most of these YOLO, this guy was a good guy, so we're gonna give him a giant fucking pile of money just because of who he is, deals in venture are gonna turn out badly. and the reason is is because the for the second thing, which is you have to solve in venture a real problem that somebody's willing to pay for, not just do something cute. And character AI was two things I I don't do, which is a fancy person starting a company on a on vibes, number one and number two, to consumers. And consumers at the end of the day are not used to paying a lot of money for things. That's why they use free email at from Google, they use free search from Google and things like that. So what they started was a company which was vibes and not solving a real problem for consumers who don't pay for anything. And it was cute, and you could talk to people, but it was at the end of the day a rapper solving a non-real problem for people who didn't pay for anything. it was obvious to me that character would go to zero from the beginning. Now, they got a little bit of money from Google because there was some technical things that was maybe a little bit whatever, but it it was not a real venture scale. problem. and but the more important thing I I think you know because there will be people who create valuable things in consumer I I personally don't invest in consumer for in part for this reason that the majority of consumer things are really hard to do at scale to get consumers to pay for at scale. They're easier to get business people to pay. But the real problem is that people who have a fancy resume they wrote the paper they were the first guy they were Elon Musk's roommate They were who whoever. In today's market, venture capital can get a lot of money for very little things. And you know, that's how the venture capital game works. The majority of those turn out badly. And this is an example of that. This person may have been a great researcher who knew a lot about the technical details of his paper. He was not a good businessman. He did not create a business.

Niklas:Then on the other hand we have, and you have tweeted about that, another company that I think is making money, which is Aura actually, who

Martin Tobias - Incisive Ventu:Yes.

Niklas:just canceled their IPO. What's the background on that?

Martin Tobias - Incisive Ventu:Well, I have my aura right here. I was charging it up. I I've been a user of Aura for a long time. I've been a a shareholder I was friends with the original founder. Harpreet. it was an angel investment. I I do not in my fund invest in these kind of things because it's consumer, it's hardware, all sorts of things that make it a very, very hard business. But they have turned out to be the most useful, the most valuable one in that space. I at one time I tried all of them. I was wearing, you know, the three different things on my wrist, the the whoop and Apple Watch and all different kinds of rings. I A B tested like twenty. products or as the best product. and it they have good patent protection. They actually put a lot of their competitors in the US out of business because of patent protection. But the challenge has always been it's been a hard thing to sell because you've got different sizes, you got an inventory problem, and it it is a little bit expensive relative to what it can do, and you can get a decent version of what they do free in your Apple Watch. So the question was always what is the market size for people that are willing to Pay a premium for a better 10-20% better product. So that was one question. And other consumer hardware devices in fitness did not go well. GoPro was down 97% from the IPO price. Peloton went to shit. And so there was that overhang of you know, do these consumer health hardware things thing. But the worst thing, and the thing that was probably the nail in the coffin for them was and and they all Are profitable, they are growing, they have a good brand. they were gonna sell about 2.2 billion dollars of shares. About 37% of those were gonna be shares that the company sold to fund their operations, but about 63% were insiders selling, and that is always a bad sign for IPO buyers because they're basically being asked to be bag holders and give a whole bunch of money to the early investors, and frankly, it would have Have been me as an early investor getting some of the money too. So a lot of the venture people are like, wow, that would have been great if that thing got public because we could have sold our shares and gotten out and made some money. But the the the other side of this IPO thing, which the venture people sometimes forget, is that somebody, a public buyer, has to buy those shares and agree that there is upside from here. And one of the negative signals in a IPO or in any kind of exit is if the majority of the money is going to pay off prior investors versus going to grow the company because then there is less potential upside to grow the company because there's less money to grow the company. And I think those combination of factors the nail on the coffin being the majority of the offering was going to pay off venture investors is why they couldn't get it done. the CEO said something which didn't make any sense to me. He said our offering was five times oversubscribed. If it was in fact he would have done the offering because he would have been able to sell all his shares. The only reason to pull your offering the day you were supposed to price it is that there were not enough fucking buyers. There is no other reason. And the reason there were not enough buyers is that primarily I don't think they like if they had been selling two point two billion dollars of shares that were all going into growing a profitable company that was growing and very profitable, I think it would have been an incredibly successful IPO. And then everybody else like me was you know on a lo on a lockup. And that's exactly what happened. you know, you look at what happened with like SpaceX. Spa SpaceX's IPO, one hundred percent of their whatever it was seventy billion dollars was primary shares sold by the company to go into the coffers of ac of of SpaceX to grow the company. And all of the existing investors, they had this tiered lockup thing, which is a second thing which was pressure on the company. But a lockup is actually a preferred way for early investors to get liquidity later than right at the IPO and taking your money up front. And I believe that that there are Good ways I've taken two companies public. There are good ways to get companies public in bad ways, and a very bad way is to try to take all of your early investor money off the table right at right up front that and and or approve that. So while the IPO window is open, there are good and bad ways to do IPOs.

Niklas:Yeah, I think SpaceX is also a really interesting example because what they did wasn't only really smart in that sense that you just described, but they only sold a very small percentage of the company at that point in time, actually. So I think they

Martin Tobias - Incisive Ventu:They they they

Niklas:they sold like three percent. And so they now have a lot of equity actually to use it to buy other companies. We have seen that with Kerz already. But like SpaceX through the valuation, they are now able to to use like their to do. obviously acquisitions by just paying with their stock and are able at a relatively high valuation acquire other companies. And I think that's a really smart move actually.

Martin Tobias - Incisive Ventu:Yeah. Th that is, and that is one of the the good reasons to go public is to get a public currency. And that's exactly what, you know, SpaceX did. They bought they they were the only company on Earth who could pay sixty billion dollars for cursor, which was a wrapper that was declining, that was losing market share, but had a giant number of amount of data that allowed him to retrain his own model to catch up to Claude Code and Codex. And And because sixty billion dollars was a tiny percentage of his market cap, he could afford to overpay and have a good outcome for everybody. And that's one thing you want in a well-functioning venture market is people with very valuable public currencies that can use them in this way. The other one, which is using a public currency correctly and and and benefiting from it is bending spoons. And I don't know if you saw Jason Lemkin's thing, I re-posted it, but bending spoons is doing a typical well played roll-up strategy. I don't think you'll see X, you know, SpaceX doing as many acquisitions as you know, the typical roll-up people because it's really about Elon's vision and his ability to build products themselves. So he will do add-on things where he can get a little bit of speed like cursor, but I don't think you'll see them f just fundamentally buying random businesses for the arbitrage of the cash flow like Bending Spoons is. But if you look at what Bending Spoons does, they're buying In the public market, they're getting paid by the public market by shareholders roughly 8x revenue for their combined businesses. And they are buying businesses at 2.5 to 3x revenue. So they are immediately so spay vending spoons bought Airtable for $2.4 billion or something like that and added 10 billion of market cap to their company. So they basically made seven and a half billion dollars in share value by buying a company. Which is good for them.

Niklas:And it's it's impressive. Yeah, it's a I mean it's a really like the play to put it. What does bending spoons do? They acquire companies at two, three X ARR, they take the development to Europe, they put it to Italy, they fire the the developers from the team, they really got the organization and they do it like as an arbitrage deal on cash flow. You can say like a typical ZARS, you could probably operate at an at eighty percent margin, I would think.

Martin Tobias - Incisive Ventu:Exactly.

Niklas:Read if you if you do it well. And then it's like you pay three X, for example, then you need like four years of consistent revenue, maybe five, to pay it back if you assume no gross, with growth less. And the market actually thinks, yeah.

Martin Tobias - Incisive Ventu:But they but but they even got it paid back on day one because of the increase in the market cap of the holding company. right. So the i it is a well proven model. They're they've been doing it for a long time. I heard the CEO on the all in podcast. I think that's good. I also think it it exposes a a another thing. You know, again, you know, the these are examples of later stage liquidity, which a good, well-functioning venture market needs, but it is also you know, repricing some of these assets at cause the other side of the air table story was less than a year before they sold for two point four billion or whatever it was, they did around at eleven billion. So the reality is is that the late stage private markets are way ahead of where the public markets are pricing these assets. And and we have, you know, 1400 late-stage unicorns which are pri have been priced by the private markets and the public markets are pricing these things at a significant discount. There's there's very, very few public market premiums. I mean people like SpaceX are getting a huge premium. But most software companies, most companies that venture companies are gonna sell to are getting not great public mul multiples and are are gonna end up getting repriced. Now the the lesson from for venture capitalists and for startups is exics are available they're available probably at a discount to what you raised money at one or two years ago but You are gonna have to sell the company in the market that you have, not the market that you want. And trying to catch up to a prior valuation is a losing strategy and eventually these things are priced the the the way that they're priced by the public markets and you're y many people, many many management teams are gonna just have to do that or the shareholders are gonna press for it just because they've been in I think Airtable that was twelve twelve years or s or fourteen years since the first check in and those investors just finally needed their money out. The other lesson is that the the only people that made a lot of money were the people who wrote the seed check who was r were my friends. And they made 55 times their money. The last three round investors at Airtable who went into the C D E got 1X their money. So the the lesson is the earlier you invest in in companies, regardless of the exit landscape, whether it's good or bad or a push, you are gonna make the most money versus the later stage people. So a lot of people in the last year or two have gotten really enamored with doing these late stage. stage deals, but you're gonna see more of these things happen where these late stage investors at best get one X their money back. There were people who were in for five years at Airtable and got one X their money. This is not a positive return for those investors. They could have had their money in T-bills and made more.

Niklas:Yeah, exactly. Or probably if you look at the S and P five hundred it would have been significantly better if you look at data.

Martin Tobias - Incisive Ventu:S P five hundred would have been great. You're not beating that. So w what what I think all of these things g get or teach the venture investors again, which you learn the hard way all the time, is you really need to invest early. You need to invest for the long term and you need to invest in something that is building something that is actually that somebody's willing to pay for that solves a real problem. Character AI didn't solve a real problem. It was a nice to have. And it was a very thin wrapper. Airtable was was very good at the time, took their foot off the gas on innovation. other alternatives got better. The only thing they had is a bunch of long-term enterprise contracts that didn't want to change. So they had some cash flow, but they had very little growth. So their exit was not in very great because they had basically flat. Their growth. There's a buyer for a flatline growth, but that is a PE cash flow-driven low multiple exit environment. it is very, very hard over the long term to continue to grow or or to have accelerating growth. You know, why is SpaceX you know been able to get through its IPO lockup and and still have a decent share price? Because They have three or four businesses, all of which could be hundred billion dollar businesses within the next year or two, which are all working. Their data center business, their launch business, their AI business. Like they and it's run by Elon Musk, who has been able to pull rabbits out of the hat a hundred times. Those businesses are maintaining a premium, but unless you can grow, you know, a hundred percent a year over a ten, twenty year period, it's a very hard exit environment right now.

Niklas:What is interesting though, as a last thing to bending spoons is I think after the acquisition they roughly add like nine X ARR to the market cap every time they buy a company. Like so they buy it for two to three, so an extra six X on top, which investors think they can handle this company a lot better actually, obviously, than the market things of the currency. I think that's a really interesting point to be honest.

Martin Tobias - Incisive Ventu:That that is, and I think that right now the market is optimistic. but you know you I I think you'll also see a correction in that arbitrage over time because I doubt That they are going to get the kinds of efficiencies that would justify an an 8x multiple instead of a 2.5x multiple over the next year or two. I think it's going to be. For example, I canceled my air table subscription. And I think there's going to be more people that do. If they if they are not able to And you know, I mean the the the the guy who proved that you could run these businesses very efficiently on all this is is again our friend Elon Musk. You know, he bought Twitter, he fired eighty percent of the people. but it was a very bumpy ride. you know, a lot of investor or advertisers left public, blah blah blah blah blah. He had to restart, he'd get people to pay for this and pay for that. And I mean, in the end, then Who did he sell it to? Himself. at a premium. there's very few companies that can through force of will, you know, make a business from a a a crappy business into a a good business and then sell it to himself. would anybody else have paid, you know, what he sold it to himself for? Probably not. So that's an N of one. And I think bending spoons as they get to scale now. the question is, I mean, they've been they're doing at scale kind of what constellation software has been doing for smaller software companies for a very long time. And s and constellation software has 25 years of proving that they can do this really well at smaller acquisitions. And Bending spoons has maybe I mean they went public this year. So the question is what does it look like two or three years from now, after they do a couple more of these, do they still get that are they able to deliver the efficiencies that did d deserves that premium? Right now that premium is still very much in the eye of an optimistic shareholder. I I I hope they do, because that would be a very great thing for the venture market to have. the venture market needs more options for for companies that that have decent but not great businesses to be able to get their investors s early investors some liquidity. And bending spoons along with Constellation Software and Trinity Software, there's, you know, Main Street, there's a there's a a handful of companies doing this and they all need to be successful and have replicatable models and be able to attract capital in order to provide liquidity for the the the venture people who took the early risk.

Niklas:Yeah, and what what is also I think true, usually in these acquisition markets they should be highly competitive. So and therefore margins should be competed away over time. So if there's kind of like a three X, two point five X to eight to nine X, op opportunity is still there. Currently it should be competed away over time. So either the acquisition prices should go up

Martin Tobias - Incisive Ventu:Exactly. Like somebody also Yeah, yeah. Somebody also come in and pay, you know, a little bit more. And he did say, I mean the the CEO of Bending Spoons did say that there were com for both Miro and Airtable, they were bank run processes. There were multiple bidders. And you can imagine, you know, Vista, Tommy Bravo, traditional PE fuckers were in there too. and you know, these companies will have options and I I think those the the owners of Airtable did the right thing because selling to a public company, which is basically a PE company in a public company Gives you public shares. If they had sold to Tommy Pro Bravo, they are waiting again another three years until Tommy Bravo can get some liquidity. So the venture market wants all of these options on the table for their companies. PE, public roll-ups, spacks, direct IPOs. You've got to have all of those options working. And depending on how good your business is, it's gonna fit into each one of those. There's even one, you know, I have two companies that are working on it. and and and and and and there are companies that are buying the GitHub repos of bankrupt companies, sometimes now for more money than the IP or anything else is worth, because they're these AI companies are buying Slack databases, GitHub databases, Just basically Discord, you know, c servers. they need training data. and that's an interesting on the low end, say your company failed. it used to be I I had a company that that failed, the only asset they were able to monetize was their domain name. But now if you had the ability to monetize your GitHub repo, your Slack channel, and all this other stuff through training data, okay, now all of a sudden there's a little more money there. And for a well-functioning venture market, you need exit opportunities at every scale of potential exit, including failures, you know, aqua hires, y you know, roll-ups to PE, roll-ups to public. IPO success, you need all of these markets functioning. And the good thing for venture investors is it there are more options out there at all levels. the hard thing is that they all of those numbers are not gonna be are are are at a discount to what the venture market is willing to pay For companies that they believe will continue to grow. And that's why the private market

Niklas:And that's I think

Martin Tobias - Incisive Ventu:is priced ahead of the public markets.

Niklas:And that's interesting I think because we are also looking at a possible IPO upcoming which is Entropic. I think it's n after

Martin Tobias - Incisive Ventu:Mm-hmm.

Niklas:SpaceX probably the hottest topic we currently have in the market. What do you think? Will they go public this year?

Martin Tobias - Incisive Ventu:w I hope they go public. I think it's gonna create a lot of money, unlock money for venture capitalists to recycle into other companies, plus create a bunch of millionaires that might start investing in companies and and and restart recycle some of that money into new innovations. So I think that's the good reason. I'm a little f I I I'm more and more convinced that they will go public, but I'm a little surprised by their public statements. I mean I I don't understand Dario going up and throwing up his hands and saying I created a fucking Frankenstein and please help me with this and you know it's gonna kill us. and then he please buy my shares and make me a billionaire. I I don't understand that at all. But one thing I tweeted about like yesterday, they just made me an offer to prepay some credits at a huge discount. and that gave me more conviction that they're gonna go public because they're trying to stack the books. they're growing well and and all their revenue is clawed code. and But they want to get prepaid credits jacked up so they could crush their first quarter. Cause to have a successful IPO, I think there's enough people that it will buy the shares, but then the question is Do you continue to grow after the first quarter and at the lockup are you showing growth? And one way to do that is to prepay, get a bunch of prepaid revenue on your books, which they're trying to do. So they tried to get me, I had like $74 in API credits, and they're like, if you bought three hundred and seventy-one dollars and forty-two cents of credits, which I thought was a weird number, you know, we'll give you a 25% discount, which is the Biggest discount they've ever offered in prepaid credits. So it's pretty obvious to me that they're trying to get, you know, 10 or 20 billion dollars in prepaid credits so they can pack their books. So when they go public, they can have more confidence that their first quarter is going to be a good quarter. So I I'm pretty certain that's going to happen this year, just in part for that.

Niklas:It's also interesting because it's obviously a way to have some liquidity inside the company. So if you prepay credit, it's kind of working capital that you that you have available to to deploy inside. So

Martin Tobias - Incisive Ventu:Absolutely.

Niklas:that is that is also if they train large models, they that's still expensive. They have to put the mu money up front. They will probably not do another private round. I would be surprised if they tried that before the IPO. So they will somehow also need capital to bridge. That time. Then again, I'm yeah.

Martin Tobias - Incisive Ventu:Yeah, I mean they need th th them and ChatGPT and and SpaceX, that's the other reason why they went public. SpaceX is gonna do like another private offering for another like fifty billion or something like that. These companies have announced you know, a trillion dollars in capital expenditures. and the private markets cannot provide that. So they they ha they need the money to grow. And and actually that's actually, you know, a another reason I think why the IPO is gonna happen is that they're not gonna have secondary sales happening at the IPO. I mean all of the money that Anthropic and openAI raise is gonna go to growing the company. And and investors like that. I mean you can argue as to whether they really need that much or they can make a good return on the money and whatever, but right now it is pretty clear that we are super short on compute capacity and these guys are c capacity constrained. They they cannot produce tokens fast enough. And and at some point that'll stop. But right now, and I I actually think for the next year to year and a half, I don't care how fast you build compute, you are in a compute constrained, over demand, under supply environment. I don't care who goes public or how much money they raise. There's still let's say they raised a hundred billion dollars. you can't build data centers that fast.

Niklas:Only Elon. I mean Elon kind of showed what's the the fastest you can do, I think. Like his six months, yeah.

Martin Tobias - Incisive Ventu:E Elon is the fastest data center builder in the world. He can build a data center in a year. Everybody else spends five years. So that's the other problem. Is and and then the question is, who would you rather give your money to? Would you rather give it to Elon or would you rather give it to you know Microsoft or Chat GPT that takes longer? I actually think that it's a harder story for anthropic and those people to because they cannot build data centers fast enough and and they've proven that they can't. So you can give them money, but it's gonna be a worse return on investment than giving the money to Elon who could build it faster and then maybe they sell it back to Anthropic. So that's it. So and the funny thing is Anthropic needs Elon because he he he will build the data centers that they could lease.

Niklas:Yeah, and beyond that, so what is also interesting regarding Dario's comments, I think, and it has like once again two sides actually. So first of all, I think he's very vocal about this job loss topic. And I'm always when I hear that, I'm always wondering, you know, at the end of the day, somebody has to pay for the services. If people has no have no money, it's a r huge problem for the economy and for publicly traded companies at the end of the day who are measured on revenues. So I think there there's a point that that actually I'm really wondering what will happen when he when Anthropic goes public, because from a at least in my mind, from a rational point of view, as a CEO of a publicly traded company, it's very hard to make the same comments that he's making today from a as a CEO of a private company. What do you think about that?

Martin Tobias - Incisive Ventu:Well, we'll see. you know, it didn't stop Elon from making stupid comments and getting fined by the SEC, but the funny thing is I don't think Dario is is any kind of an Elon. I actually reading his comments, like if if you read what they're what They think they're building this whole effective altruism and this, you know, we're creating a sentient being that's gonna be a benevolent overlord and a new religion and all this bullshit. it's kind of a dystopian idea. And I don't really like it. and I hope it doesn't happen that way. And and I'm actually, you know, very optimistic that things like Grok, as they get better in programming, I'd rather have I personally would rather have Elon Musk writing my code than Dario, based on on on what he he has said. I I think that he he is the kind of guy that after he goes public will probably tap down his comments and and be more of a regular public CEO. I don't think he'll be somebody like Elon that just continues to do whatever the fuck he wants and says whatever he wants because he's the ri So Dario will not be the richest guy in the world. Elon will still be the richest guy in the world. So

Niklas:Yeah, that I also believe in. And I also I mean, I followed his like Tesla in the early days already when it was small. And there was a strong debate about electric cars and hydrogen back then. And Elon was very vocal why hydrogen is really stupid. Like because it's also just

Martin Tobias - Incisive Ventu:Yeah.

Niklas:an energy storage which is significantly less efficient. Like and I found this really interesting. Yeah.

Martin Tobias - Incisive Ventu:Yeah, it on a first principles basis, he had a physics argument as to why his approach was better and it was right. Now, you know, I I you you don't see that in in in in Dario. I don't I don't I don't think you they have a fundamentally superior approach to this. And by the way, the approach that they th say is superior with basically having the algorithm be smarter than you. He's he's very much one of these w one of the downsides I think for well it depends on where you are on the political spectrum, but The vast majority of people at at at Anthropic are part of this effective altruism thing. And if you take even you know, ten or twenty percent of the net worth that's gonna be created by that and put it into philanthropic to fund some of the things on that the democratic socialists care 'cause they're mostly like democratic socialists, and they believe that a benevolent overlord like the government or their a their AI can do a better job at running People's lives and the economy than people can. That's fundamentally what Anthropo the majority of people at Anthropic believe. And they're gonna fund a lot of that stuff. So you're gonna get more of this democratic socialism, you're gonna get more of this, you know, centralized decision making is better, you know, because it's benevolent. I I think you're getting exactly the opposite out of people like Elon Musk, which is let everybody do their their own thing and enable people. th this is gonna be a a a huge debate and I hope the one that wins is the let's enable people to be better versions of themselves instead of to tell them, you know, that there is this benevolent thing that knows better than you do how to live your life. That's the thing that worries me

Niklas:Yeah and if

Martin Tobias - Incisive Ventu:most about Anthropic, is that they fundamentally believe that what they're creating will be better at running your life than you are.

Niklas:Yeah, and that's why I think Elon already when he co-founded OpenAI back in the day or gave them the first money, he was he had a very clear vision about what he wanted AI to be. If there is smart AI than than humans. Like it should be to the benefit of all of humanity, these these topics. And I think that is in general still true. If we look at the age of AI, I think, and that's the last topic for today, it's which modes are actually remaining? So I think we're seeing that software, at least creation of software, is going towards zero in price if you are careful. But it like the the the creation goes down. So we have data remaining, we have network if effects remaining. What what do you think when you look at also investments that you do today versus like a year ago?

Martin Tobias - Incisive Ventu:Yeah. Well, I mean I I I think it's the same as it was a year ago. I actually I'm I'm just raising my my second fund. I think it's the best time to invest in B2B software in my lifetime. but it it it it it it is more challenging because you've got to figure out exactly what your your your moat is. And network effects are remote integrations, but the the the real moat in B2B software, and this is again the reason why I do B2B software versus consumer and stuff like that, is you've got to solve some old To do it for people that are willing to pay for it. And the people who are willing to pay for it over time are people that you understand their problems better than the next guy. And that your understanding can compound over time. So what that where that leads you to is a lot more vertical market software in niche industries. Those niche industries have to be big enough. So Anthropic published a report about four or five months ago. Months ago that showed the 10 most at-risk jobs for 2 AI, that which are going to be changed the most by AI. Number one, being a computer programmer, number two, being a customer service agent, number three, being a salesperson, and you just go down the list. And if you look at Anthropic, Grok and OpenAI's product strategy, they are very clearly going to have as a core part of their models general harness. That can be very good at those 10 jobs, which is why I never invested in Replit, Bolt.new, Lovable, all of these coding things, because I figured that coding is a core thing the models are going to do themselves. They're going to be very good at it. And by the way, when I started vibe coding, you know, a lot more sort of last year, I started with Replit. I was paying $2,000 a month. I'm now on Claude Code. I'm paying $200 a month. That is was in less than a year. You got a 90% reduction in cost because you got rid of the wrapper and I think that's gonna continue is that in in those horizontal things like coding, customer service, sales, the models are gonna be really, really good at you know bundling 90% of the features you need in those things in a $200 subscription. So those businesses are shit. So what you've got to do is go into markets where there are different kinds of workflows that require different kinds of integrations and different kinds of data that are far enough away from the models but still big enough to create a venture-style return. And I think that's where you need to invest, and that's where that you're gonna get you know, a lot of new software. Like if you look The at the the B2B SaaS business today, it's about a $300 billion business. But labor that uses sits in front of computers and uses that SaaS software is a $4.5 trillion expense. And for the first time now with agentic and certain vertical things, you can actually shift some of that $4.5 trillion employee expense into software. So what that means is that that software expense is gonna grow while the labor expense goes down. And yeah, that's a little bit of job loss, but it actually I think creates greater revenue for the companies and greater opportunities for new kinds of jobs, just not the old kind of jobs that computers can do better. You'll have new kind of jobs. In my I idea, you know, you have a net you know increase in in in overall jobs just like every technology that's ever come along, we've had a net increase in jobs and the product Of the prior ineffective jobs, you know, just gets better. So where I think where you invest in venture today, where I'm investing, and there's going to be lots of places. For example, I don't invest in space. I think people are going to make a lot of money in space. I don't invest in biotech. People are going to make a lot of money in biotech. I don't invest in robotics. People are going to make a lot of money in robotics. So I'm just talking about in software. I think B2B software is the place to be and investing in vertical market software in niches that are far enough away from. from the main models to still be big enough to get venture scale outcomes. That's what I think the opportunity is.

Niklas:Martin, thank you so much for being on this episode and see you next time.

Martin Tobias - Incisive Ventu:Okay, thank you.