Sign up to save your podcasts
Or
Share 70 – AI as a co-founder
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
70 – AI as a co-founder
Can AI be a co-founder? Do you need a technical co-founder, any more? How about a business co-founder? What can AI do for you as co-founder? Will this become the “brave new world” of start-ups, small and medium businesses? For this and much more discussion, a no BS perspective on AI as a potential co-founder.
Our co-hosts:
Our show: Tech DECIPHERED brings you the Entrepreneur and Investor views on Big Tech, VC and Start-up news, opinion pieces and research. We decipher their meaning, and add inside knowledge and context. Being nerds, we also discuss the latest gadgets and pop culture news
Subscribe To Our Podcast
Bertrand Schmitt
Intro
Welcome to episode 70 of Tech DECIPHERED. Today, we’ll talk about AI as a co-founder. Can AI be your co-founder? Do you need a technical co-founder anymore? Or do you need a business co-founder anymore? What can AI do for you as a co-founder? Will this become the brave new world of startup and small businesses? Nuno, what’s your take on this topic? Have you started seeing that for early stage startups, the AI co-founder?
Nuno Gonçalves Pedro
Yeah. We start seeing this notion of people now—”Oh, I could be a single founder.” I mean, single founders have existed for a long time. We’ll come back to a little bit the taxonomies of founding teams. But definitely, is now a little bit of a trend where people are like, “Well, I don’t need a co-founder. I’m just going to go do my own thing.” Normally, the case is made more for, “Oh, I don’t need a technical co-founder. I can vibe code and put some stuff together and go through things.”
As we go through the episode today, I think we’ll go into the details on why technical co-founders might still matter and why there are certain areas of technical founding that might not matter as much where AI can really be your co-pilot, so to speak.
The Classic Case for Co-Founders
But maybe let’s start with what is the case for co-founders? Why do you need to have a co-founder? Why can’t you just do it yourself? Historically, there’s been really sort of an angle where there’s sort of these two entities in the founding team. Right? The business founder and the technical founder.
The business founder is the person that runs business related activities. If you’re doing, for example, enterprise software, SaaS, et cetera, your business co-founder is responsible for go-to-market like hiring sales—in particular, at the beginning actually being the person who does sales, establishing partnerships, creating and managing elements that are more related to admin with the help maybe of third parties around finance, et cetera. The business founder is more the person who’s focused on the business aspects of the company, which would be go-to-market, which includes sales, channel partnerships, marketing, et cetera. Might include, as I said, the admin side, et cetera.
Then the technical co-founder is more focused on elements that are connected to the technology stack, development of the code base. If it’s just software, development of software, could be more on the product side as well, someone who’s more of a product manager, et cetera. That’s why you need those two entities. Because you need these two entities, so to speak, these two people. Because you need someone who has more knowledge of how to develop a code base, how to get it off the ground, how to develop the MVP, the minimum viable product early on.
On the other side, you need someone who figures out: how do we get this thing to market, how do we actually deploy it, how do we monetise it, how do we create partnerships if they apply. That’s why we’ve had this classic case for founding teams. Now just to be very clear, it is also true that we’ve also had single founding teams for a long time. We have companies that have been founding only for one person. But at the end of the day, the ethos has been, let’s have two co-founders, one on the technical side at least and one or more on the technical side and one or more on the business side at the very least.
Bertrand Schmitt
That’s true. That’s what you have typically seen. Maybe, going back historically, maybe the most famous example of this technical co-founder plus business co-founder has been the founding team of Apple—Steve Jobs with Mr Wozniak. Steve, on the business side; Woz, as he was called, on the technical side. That has been that maybe that’s reference point for all of Silicon Valley for decades. At the same time, it has not been true for every successful company. If I take HP, for instance—if I remember well, they were both technical co-founders: Bill Hewlett, Dave Packard. If you take Microsoft, Bill Gates had, in some ways, a lesser co-founder, but both were technical. Paul Allen was technical like Bill Gates.
Nuno Gonçalves Pedro
I would say Gates was more incredibly technical, but he became a business guy. I think that was part of the secrets of Microsoft early on. He was very business-savvy in some ways. He was very technical. Until this day, I think he’s extremely technical and extremely intelligent, but he was definitely very business-savvy.
Bertrand Schmitt
He was able to wear two hats in some ways. If you think about HP, it’s probably a similar story. When we talk about the one-person founder alone by himself, the question would be, is it a technical co-founder or business co-founder or a mix of both? Someone who can do both even early on or who learned to do both pretty quickly. Probably, Bill Gates would be a good example if we assume Paul Allen was less critical.
Nuno Gonçalves Pedro
Yeah. I think that’s a good point, and you went back to one of the OGs like Wozniak and Jobs. I’d say Jobs was business clearly, great marketeer, et cetera. But he was also very deeply involved in product. Interestingly enough, think maybe not as involved on product early on as he became later on with NeXT and then his second coming, so to speak, to Apple. Where you get much more involved in product management and how products were deployed. But to your point, he wasn’t a technical co-founder. Clearly, that was not the role.
I think if we look at some of the great success like Google, Sergei and Larry were both very technical. I guess they brought in Eric Schmidt very early on to be the CEO. But at the end of the day, the definition of founders in Silicon Valley, maybe we’ll come back to that later, is at least you need to have technical founders early on. Then maybe the business people will come along or maybe one of the technical people will go more onto the business side. But you need to have this duality of role that you, Bertrand, we were just talking about. Someone who either becomes more business-savvy, more business side, but definitely someone who’s more on the technical side. I think that’s how the Silicon Valley has emerged over the last many decades. Right?
Bertrand Schmitt
Yeah. It’s interesting you brought in the story of Google with Eric Schmidt because you could argue they brought him in as a CEO, but he used to be CTO at—was it Novell? I forgot.
Nuno Gonçalves Pedro
Novell. Yeah. I think he was CEO of Novell at some point. Yeah.
Bertrand Schmitt
He was CEO at some point because I was going to say, clearly, he has also a technical background. Able to do the business side. No question. But he was also coming from the technical side.
Nuno Gonçalves Pedro
He was definitely technical at the beginning of his career, and he was also at PARC, Bell Labs, and a variety of other organisations. He was in software engineering at Sun Microsystems, to your point. Then he became the CEO and chairman of Novell. But he is a technical guy by background, for sure. No doubt. He has a PhD. But he was brought in to be the CEO. He was brought in to be more the business guy. I think, very honestly, Silicon Valley loves the story of actually the technical CEO. The person who has at least very product or deep product knowledge and understanding of technology, so it doesn’t get bullshitted necessarily by the rest of the engineering team, so to speak.
Now there’s notable exceptions to that. There are companies that were built very heavily on people that were incredibly business-savvy and very much on the go-to-market side. But, again, justifying why you have co-founders, why you have the technical person and the business person at the very least early on in the structure of the company.
Bertrand Schmitt
If we look at more recently, of course, we have the exception of Elon Musk starting and running so many companies in parallel. Some he didn’t technically funded. He was early on investor, took over very quickly. But he’s also an interesting example of a very deeply technical person with very strong business acumen, obviously. That’s another example. Do you have examples of many successful companies just started by a very business co-person in the tech industry? Obviously, outside tech, it’s a different story. Yeah?
Nuno Gonçalves Pedro
Not a lot, but I feel like in the enterprise software side, there are people that have managed to be great at doing it because they are amazing at building sales teams, and amazing at then building engineering teams around them. I’m not sure I can start giving as many examples as we’ve given on the other side of people that did have a very strong technical background so to speak. I don’t know if he’s a technical guy. I do think he had the bachelor’s in electrical engineering. One guy is David Duffield. He was the guy who founded PeopleSoft and then later on founded Workday.
Bertrand Schmitt
Yeah. I was looking for him, actually.
Nuno Gonçalves Pedro
David Duffield, to my knowledge, I think he has a bachelor’s in electrical engineering and then MBA. But I feel when he started his companies later on, he was very much on the business side. He had gone through marketing at IBM, and did a variety of other roles that led him to create PeopleSoft in 1987. My perspective is probably that he had become more of a business guy later in his career.
But to your point, I don’t think there is a ton of examples. Jobs, I guess, maybe was a great example of that because he was very business. He wasn’t as technical. I think he always had a great product mind, but he wasn’t very technical, from what we know about Steve Jobs.
Bertrand Schmitt
Of course, we have Larry Ellison, founder of Oracle. I think he did it alone. No? I don’t have memories of any co-founder of Oracle. He was technical originally.
Nuno Gonçalves Pedro
Yeah. I’m not sure if he had a co-founder. Actually, that’s a really good question. These stories are never very well told, so one never knows. My knowledge, I don’t think he was. I do think he’s he was very sales-driven from the outset when he was doing Oracle.
Bertrand Schmitt
No. Actually, he co-founded Oracle. Bob Miner and Ed Oates were the other co-founders.
Nuno Gonçalves Pedro
He founded Software Development Labs, right? Yes. Then they changed names at some point. They changed names a couple of times, actually, so he did co-found it. But he was more of the sales guy from what I understand. He was more the hardcore go-to-market guy, so to speak. Really impressive how he built that. I think that’s another good example of someone that was more on the business side that managed to scale a company dramatically. Really interesting.
Not a ton of examples, but probably if you guys are listening to this, and you have a lot of examples that you want to share, you share with us. Maybe there have been some certainly on the enterprise side, enterprise platforms, or Software as a Service side that have been incredibly successful. Do feel free to share with us on X or on LinkedIn.
Bertrand Schmitt
Going back to AI as a co-founder, do you see AI as a potential technical co-founder or more like a business co-founder? What’s your take?
Nuno Gonçalves Pedro
AI as the Technical Co-Founder
Let’s start maybe with the technical side, because I think that’s the most interesting one to debate. Could you be a business-savvy person that just uses artificial intelligence as your technical co-founder, and you don’t need anyone technical early on? There’s now a lot of discussions that people have built an MVP in a couple of weeks, raised $500,000, raised a bunch of money and went forward.
I’m not sure I’d call it an MVP, a minimum viable product, but building a proof of concept or something early on with the help of AI, I feel is eminently doable these days. With vibe coding, with all the tools that people have out there. I am still of the opinion, again, being a computer engineer by background, that at some point having technical founders on board or technical people on board very early on is pretty important. The reality is, we’ve had this discussion in one of our previous episodes, is we need to go back to the methodology of software development.
For those who are listening to us who are not very technical, for a very long time, just a very quick rehash, the world was based on this methodological approach to software development, which was called waterfall, where you would first think through design, the architecture of the system, and only later you’d code it, and then you test it, and then you deploy it. It looked like a waterfall. It looked like every stage has its own part.
Then at some point in time, I would say probably with the advent of the Internet late ’90s, we started going to a world that is definitely more going through a methodological approach that now people call broadly agile software development. Again, if you guys want to go back and listen to the different methodological approaches, we’ve had episodes on it in the past, so you can definitely Google it and search for it.
The world has gone to agile. What that means is that, with going through an agile software development where it’s a lot more iterative, where you’re deploying code a lot more iteratively, you might be under the impression that, actually you don’t need to architect systems very early on. You just need to fake it until you make it. You just need to develop a code base that you can use, and then you go and basically, it works, or it doesn’t work. I think that’s still a mistake. I think the world has gone too agile. We discussed that in one of our previous episodes as well.
Architecting the platforms and what you’re doing early on is important. I think that’s one of the key reasons why you still should have a technical co-founder, and it’s going to be very difficult for AI totally replace it. Because I think the parts of AI that you can use as, so to speak, as a technical co-founder are the parts that would be done by a relatively junior developer. If you’re not very good at asking AI for what you need in terms of code, the code you’re going to get is going to be crap. The architecture and the underpinning and everything that’s underneath it, the platforms that you’d use in terms of your cloud architecture, et cetera, will also be crap.
Basically, you can develop a proof of concept. Maybe you can even get to an MVP early on, but at some point in time, you’ll need to rewrite the whole thing. You need to get technical talent in, just to figure out, is this code base unique or not? For applications that are just software, basic apps, et cetera, potentially can get away with it early on, but at some point, you’re going to pay the cost of it—of not having had a technical co-founder, of not having had any help in terms of architecting the system early on, of not having had any senior software developer early on prompting the AI or working with the team or developing it him or herself. You’re going to pay that cost at some point in time.
Now for deep tech stuff or heavy-duty platforms, whatever, there’s no way in hell. Because that’s not something that AI can help you today fundamentally innovate on. It’s just the level of knowledge you need to have to build it. It’s intrinsically linked to a technical co-founder, so to speak. My perspective on it: it’s great to if you’re doing software, if you’re doing something that’s little bit simpler—be it consumer or enterprise software—maybe you can get away with AI as your co-founder early on. You’re going to pay the price later on, so having at least senior engineering team early on that can help you or can help you think through things is something that actually has a lot of value. I feel the more you delay it, the more you’re going to pay for it later on.
Can you raise money? I think it’s very difficult. Again, unless you figured out some angle on your go-to-market or the business side or something that allows you to have a bunch of traction, retention engagement early on because you figured out some angle, that’s great. But you definitely won’t have developed a tech mode. That’s the piece that at some point in time you’re going to have to pay for. What’s your perspective, Bertrand?
Bertrand Schmitt
I think as you say, it depends on the industry. If you are in deep tech or core infrastructure, it’s quite obvious that you need someone already technical to lead that because the technical side will truly be your key differentiating factor. You need that not just to raise money, but just truly successfully execute. If the product is a bit less technical like many SaaS products are, to be frank, I guess it’s a different story.
For me, the question would be what type of non-technical funder are we talking about that can benefit from an AI co-founder? If we are talking about the business co-founder who truly has no clue about how you build stuff, I guess it won’t really work. Yes, maybe you can fake it to go to the first level of the model, raise a few $100,000s, and then find your co-founder.
I will say where it gets more interesting is if you are technical enough yourself, have a good business sense, and just don’t want to have a technical co-founder for some reason because you don’t want to share the equity, you don’t want to share the control, or you just want to have some level of control on the whole business, but you don’t have the time, you don’t have the true expertise. I think that’s where AI can be quite interesting, actually.
I must say myself, I’ve been quite impressed of how I’ve been able to use AI recently to develop tools, to develop products, to build infrastructure in the cloud. Yes, it requires me to spend some time to understand what I’m doing because I was not going to just trust blindly the AI. I spent some time discussing with it options, and I must have been quite impressed how it was coming up with a level of expertise that is actually not so easy to find. Because when you build out infrastructure and stuff, you have to think at so many levels—from security, to Linux configuration, to network, to storage, to development stack. You’ll realise that it’s actually quite rare to find anyone, even a good CTO who has truly expertise on all of this stuff. That’s where you realise that there is a lot of value if you kind of understand what you are doing, and you don’t let your AI go wild. Because it’s very clear that if you let your AI go wild, step by step, it goes wrong. It’s very clear.
From my perspective, if you are technical enough—like a product manager, for instance, who did some engineering studies, learned some computer science. I’ve not really been deep technical, but understand how all the pieces of the puzzle fit together, have some understanding about what is state-of-art in the industry, but didn’t rise from the ranks of coding and developing or network administration, that sort of stuff. I think you might get real value from AI today as long as you are ready to control it relatively tightly.
For instance, vibe coding, I have absolutely no trust in that. I think it’s just plain crazy when you see how it’s working. But I think you can do some development where you are helpful significantly with AI. You can make the architecture decisions with AI. I think there is a lot you can do if you have some, I will say, technical flair, but not deep technical expertise.
Nuno Gonçalves Pedro
Yeah. I mean, it’s unavoidable that AI will take over a lot of coding right now, that people will be using AI agents to develop code for them. I think there’s Y Combinator Winter ’25 data that 25% of the startups are using 95% plus of AI-generated code bases. I think that’s a bit scary, but anyway.
Bertrand Schmitt
That’s scary for me, as of today’s technology.
Nuno Gonçalves Pedro
It’s very scary. Yeah. Back to the point, you will pay for it at some point. There’s another unintended consequence of everything is AI-coded, so to speak, which is we often forget that companies, when they’re early, they might have a very clear vision of what they want to become and what product they’re developing. But at some point in time, they might have to, shockingly enough, pivot and go to other areas. If we think about pivoting as part of the intrinsic development of a startup through its life, that you serve intrinsic, maybe at some point you’ll need to change, in terms of what product area you’re going after, even what business you’re going after, a lot of it sometimes actually is educated by the tech stack that you’ve built in the first place.
We’ve had amazing failures in companies, notable gaming companies that failed as gaming companies that led to the creation of great companies. I won’t flag App Annie, but App Annie came out of a gaming company. Slack came out of a gaming company as well. There’s all these things that sometimes happen, and it happens because someone developed code that was really, really good on something else.
That allows you to figure out maybe we should pivot. Maybe we should go after that space instead. Maybe we have more of a platform angle here rather than an actual app angle or vice versa. We’re more in an app angle than a platform angle here. That doesn’t happen if you’re just coding stuff that serves a purpose. There’s no actual quality of code that would allow you to, at some point, go and pivot to something else. I think that’s a little bit the unintended consequence of also having such huge dependency on AI-generated code bases and potentially not having a technical co-founder just because you’re like, I don’t want to deal with technical co-founders, and it’s too complicated, et cetera.
Bertrand Schmitt
Yes. I mean, you could argue before AI, what were people doing if they didn’t want to have a technical co-founder or couldn’t find a technical co-founder early on? We had many situations, I guess you heard as well, about teams that are outsourcing their early development to some sweatshop far away. We thought control was talking about what’s going on. In some ways, that’s similar to AI vibe coding today. You have no clue what’s happening. You ask for change, and you get some change, and all the stuff got broken. It was available in a different format, and definitely more expensive. You had the option before, and obviously, that was not a great sign. I guess some very early investor will accept it if there is a plan to change that, basically, and make a tech transfer of some sort.
Nuno Gonçalves Pedro
It’s very interesting because I’m going on two different portfolio companies of mine. I’ve gone through two different at the same time two different stages. One where the two co-founders weren’t highly technical. One of the co-founders was really a product person, though, product manager, et cetera. At some point in time, they believe we want to control more of our engineering team, so they started having more and more of their engineering team in house and less of it outsourced. Very interestingly, another co-founder, actually, co-founder who’s the CEO of a company, she is a computer science background, and she can code. She’s actually going the opposite direction now. She’s like, I want to outsource more and more of my team actually, because I can control quality.
I think what’s really important is can you control the quality of what you’re getting on the other end, back to your point. If you can have visibility on what you want, you can have some ability to have oversight of the code base that’s being generated, you know it’s solid, and you know it’s very strong, and the quality of the code is strong, I mean, it doesn’t matter if you outsource it or not.
To your point, you could use AI for a lot of stuff if you’re an individual contributor or if you’re a founder, et cetera. If you don’t have that knowledge, if you don’t have that understanding, if you can’t judge the quality of a code base and the architecture that your team has put together, then maybe you should have a technical co-founder at some point in time or a very senior hire very early on to the company that can help you make those decisions.
I feel AI—as a full on technical co-founder—maybe can get away with it early on, stage-wise, and maybe pre-seed. Maybe you’re early days of figuring stuff out, but at some point, you’re going to pay the price for it. Be it at the series C stage or a series A stage, you’re going to pay for it. If you’re really, like, I’m not very technical, and I don’t need a technical person, I feel at some point you’re going to pay for it. At some point, someone’s got to ask, where’s your senior technical person. That’s keeping the trains going, either CTO, VPN, et cetera.
Bertrand Schmitt
You talk about evaluating the quality of the code and everything. I mean, you could argue you could use AI for that. I’m sure AI these days is not so bad to evaluate code quality. How is it structured? How is it documented? How which standard is respecting? I think, actually, AI for this type of task is really not bad. You as a founder, that might help you understand that if you decided to use AI to build, you could use AI to review. If decided to outsource to build, you could use AI to review. By the way, if you use outsourcing these days, how are you sure that, I mean, it’s not a lot of AIs behind the scenes.
Nuno Gonçalves Pedro
Correct. I mean, you can’t be sure unless you have, again, some technical knowledge, or you’re checking the code in some way. I think my verdict on can AI be a technical co-founder is yes early on, but not long term if you are a company that has some level of technology differentiation, be it in software or physical tech. That’s my two cents.
Bertrand Schmitt
I agree with you. I can see AI as a shortcut early on as actually a way to improve things early on. Because one thing I’ve seen a lot in early stage startups is that technical co-founder who is good but not great, you’re a bit worried. That’s why you are like, actually, AI might help evaluate or might help a company, might help make better decisions. Someone who was mediocre or good become much better, thanks to AI. That could be another perspective.
Nuno Gonçalves Pedro
AI as the Technical Co-Founder
Maybe moving to AI as the business co-founder, what do you think? Can AI be the business co-founder? You’re technical, and I just want to have a business co-founder, but I don’t want to have a business co-founder. I’m just going to use AI for it. Can you get away with it?
Bertrand Schmitt
I think that’s an interesting perspective. I think it’s quite similar to the technical side. I think AI can help you a lot to improve things, make better things. For instance, designing part of the product, designing your logo, picking your colours. I think AI can help you get much better if you were a technical co-founder. Can it truly replace a more business co-founder? I’m not so sure because a business co-founder will what? We help find financing. We’ll help put in place partnership. We’ll help you close some big deals. This is stuff that you cannot really do with AI. There is some stuff, business side, that AI can help you with. That’s for sure. To help you set up a company, to help you do some legal reviews. I think there are a lot these days that AI can help. In terms of truly big value add, I’m not sure.
Nuno Gonçalves Pedro
I have a bias. I always felt the need for a business co-founder early on wasn’t as significant as others thought. Now you, I think I identified the key issue with not having a business co-founder early on, which is, for me really fundraising. Because everything else you can definitely get there early on through help on marketing, on content, on SEO, on your sales collateral stuff, on pitch decks, or whatever. This is true not only I think of B2B SaaS, but I think it’s also true of consumer. You can be a highly technical founder and just find a lot of stuff online. Now with AI, it’s even easier. You’re really interacting in a way that’s almost like a co-founder, so to speak.
I think the problem starts if you need to raise a bunch of money. If you start having to have a discussion around raising a bunch of money, then you have to have that salesmanship of pitching your company, identifying the markets you’re playing in, being very clear on what’s your go-to market, for example, again, you’re in B2B SaaS. That, I feel, is difficult to do if you don’t have the business co-founder. Now it could be that the technical person is very good and a bit of a hustler type. I mean, often type when we talk about co-founders, we talk about the business person as the hustler and the technical person is the hacker, so it’s the hacker and the hustler. It might be that the hacker in this case might be a hustler as well, but it’s very uncommon. It’s very uncommon that the person has salesmanship. The person whose technical has salesmanship. I think that’s the only key issue that I see.
That said, you can get away in my opinion, I think you can get away with it more being a technical founder that doesn’t have a business co-founder and for longer than the other way around, than you being a business founder and not having a technical co-founder. That’s at least my bias. You can get away with it longer and deeper if you are that technical founder yourself. That’s my two cents on this area—is you have the advantage still to be the technical person.
Bertrand Schmitt
I certainly agree. I think the usually, most of the moat of a company goes to the technical side. It’s not always the case, but it happens often enough. If you’re an investor, even if you’re a buyer, you will feel more comfortable when you see that in front of you, you have someone who understands the technology, who understands the pieces of the puzzle, and who you feel can deliver on it. As you say, I think there is a belief for investors that early on, do you really need that business co-founder? Because early on, for instance, maybe you are deep down just coding for two years to get your product out of the door, and you might have two years where you don’t really need that business co-founder. Could even be nearly a deadweight in some ways.
One thing I’ve seen also is a belief from some investors that I mean, as we talked during Google early days that, hey, you have the great founders, great technology. We’re just going to attack a CEO or a COO, and that will all work out. No big deal in a way. It’s easier to find that than to find the right product and the right business. Adding that later on, there is a market for that business person who wants to join a high-flying startup. I think that’s definitely how things have been working for a while in Silicon Valley. In the same direction, there has been a new approach maybe the last 10, 15 years where instead of replacing that technical co-founder, it’s been to make sure you nurture that technical co-founder. He’s becoming more business-savvy over the years, and there is no real need any more to have a business co-founder. You will have a COO, chief of staff, or whatever.
I mean, a good example, we now talk about Mark Zuckerberg from Facebook, but, I mean, he clearly came from the technical side. He stayed CEO, but he had some COO who helped him on the business side of things, but were still running the show. I think that’s another good example.
Nuno Gonçalves Pedro
Yeah, I think the points you’re making, Bertrand, are extremely valid. Maybe two things that I would say. One piece is we have had some companies, just to be clear, where there was huge distinctiveness. You could argue it’s more on the product technical side or less, but for example, I think the stuff that Nikita Bier did at TBH before they got bought, it was very centred around growth hacking, and we’ve talked about it at a previous episode.
To my knowledge, his background is more non-technical. It’s more on the business side. There’s sometimes you do have a business founder, co-founder, et cetera, that does identify an angle that is very unique either if it’s B2B, a go-to-market angle, or if it’s on the consumer side, some growth hacks or some way to really get in front of consumers, et cetera. We don’t want to downplay the role of a business founder here, and there’s a lot of knowledge that gets attributed in terms of distinctiveness to it.
That said, I feel this is probably an area where, again, the technical founders are going to get away more with not having the business co-founder than the other way around. We already mentioned at the beginning of this episode a bunch of cases where the team that started was really technical—I mean, Google, et cetera. Having the ability later on to bring in a senior person that acts more as the grown up and adult like we could argue Google did with Eric Schmidt back in the day, like you were just mentioning Mark Zuckerberg did with Sheryl Sandberg.
I think that’s definitely something that from the perspective of the play book of a company has its scales is more widely accepted, certainly in technology. I guess, as it goes, the tech mode is probably more definitive and more important than any other piece. Probably goes without saying that there’s a little bit of bias in our comments. We are both technical by background, but we probably feel that I’m speaking for Bertrand. Don’t know if you’ll correct me in a second, but we probably feel that AI can likely be a business co-founder to a technical founder or a human technical founder longer and more successfully than the other way around. At least today, we’ll see how things evolve over the next few years.
Bertrand Schmitt
Yes. Again, I can see AI as, if not a replacement, a booster. Can you actually create better teams faster because they are complemented by AI at very low cost? AI is there to help them make better decisions all the way. On the technical side, architecture choice, some coding, code reviews. On the business side, helping you generate better copy, better design, better logos, better this. I mean, create your photography for your website. There are so many things that now, for instance, if you’re a business co-founder, you could do your job much better. I think AI as a booster, as an enabler, that could also be a way to go.
Nuno Gonçalves Pedro
The beautiful thing about AI as a co-pilot thing is you obviously have immediate access to very distilled and synthesised information. Again, in the case of business best practices, that can be very powerful and very helpful. Even, again, if you do have business co-founders. Just checking on quick things. I mean, I use AI myself for my own stuff. Either it’s a social media post for Chameleon or leveraging other things. Obviously, I’ll rewrite stuff, and I’ll write it the way that I think it’s my voice. Having that at your fingertips is very powerful. In any case, the co-piloting is hugely valuable even if you decide to have a technical and a business co-founder along the way.
Bertrand Schmitt
Yeah, I can see in the past, if you are a younger start up, there are so many things, so many decisions that you have to take, and it’s very tough to find the right level of expertise. It takes time to find it. It costs usually significant money to pay for it. Well, now we say, hi. When I say now, these days, because I’ve just said two years ago, even one year ago, it was not the same quality—but now the quality you get in some of the answers you receive, if you ask the right questions, it’s getting quite impressive. I mean, the added value you can get to have a better architecture product, a more secure platform from the get go, better design. I just feel there is a lot more you can do, a higher quality you can do, thanks to AI, relatively earlier.
Nuno Gonçalves Pedro
Investor Sentiment & Founding Team Dynamic
Maybe moving to our last piece then, which is what is the investor sentiment now in everything that’s happening with AI as co-pilot, AI as co-founder, and how does that change as we look at founding team dynamics going forward where we’ve had the two technical co-founder, the technical and business co-founders, two technical co-founders, all sorts of founding teams, how does that evolve over time?
I would say on the investor sentiment, maybe just to lay the ground on my area, at least, of focus, we spend a ton of time really looking at tech and science modes in a company. Obviously, at the top end, we look at product modes, but then we go under the hood and try and understand technologically and scientifically what are the underpinnings of the product that this company is about to deploy or has already deployed. If we can’t find anything, if we can’t find modes, if we just go under the hood and there’s no technical element to it that’s very differentiated, you’re like, I’m not sure I’m got to invest in these guys. Why would I invest in this team or in this founder if it’s a single founder company? If it’s all AI as a co-pilot on the technical side, maybe it’s a little bit of a difficult decision for us to come and invest in the company.
Now, again, we’re series Seed, series A investors, so maybe at that point in time of the development of the company, we would expect that there are some senior technical talent at the table, even if it’s not a founding talent. That’s, I think, our skew on the technical co-founder discussion where AI is the technical co-founder.
On the business co-founder thing, we do appreciate, and we do spend a lot of time looking at the company and in particular to the CEO, founder-CEO position if the person can raise money for the company. We do spend a lot of time around salesmanship. Is this person that… After we give this person money, can this person go and raise more money in 18 months or 24 months, et cetera, from other larger investors. Does this person have ability to sell?
The story of the company that they’re building. Is this person, for example, if it’s in B2B, is this person able to actually sell, sell the product and go to market? Founder led sales early on is the name of the game. If the person, again, doesn’t have that savviness, that ability to articulate the story of the company, that edginess around either marketing or sales or something else, probably will also be a pass for us.
Where do we get to? I mean, I think it would be more likely that we invest in a highly technical team that has a very low business savviness, but has something very distinctiveness and differentiated on the technical side, then we would do the opposite. We would invest in a founding team or a founding person, a company that is incredibly business-oriented, amazing hustler or whatever—but has no technology mode whatsoever, no ability to create anything in the foreseeable future. I’d say that’s probably where our biases are as a firm today. Now it doesn’t mean we won’t do the second one, but it’s just more unlikely. I think that’s how we would frame it.
Bertrand Schmitt
I guess that’s also for me, what I’ve often seen is a discussion with the team, with the founder in these situations of, are you okay bringing in a technical co-founder? Are you okay bringing in an external CEO, or a CEO to manage the business side? Have you started discussion in that direction? Do you have leads? What type of people are you thinking about? I think that could be part of the discussion, depending on the stage of the venture. I must have seen that quite often, that discussion. Even if, from my perspective, I still have a preference with the founders—the technical founder, or relatively technical founder staying the CEO. Instead of bringing an external CEO, bringing more a CEO type of person.
In some ways, I prefer to see the CEO build up his game on the business side as necessary, versus the replacement like it used to be in the Google days. Personally, I’m a bigger fan of growing up a technical background CEO to stay as a CEO for the long run.
Nuno Gonçalves Pedro
Basically, we both have similar biases, I guess. When will we have the solo founder AI thing be happening all the time? I think pretty soon. We already have solo founders as a really significant part of the business. We already have right now, I’m pretty sure, the next big company that is a solo founder using AI as a co-pilot being created. I don’t think there’s anything dramatically different from all of that. I feel that’s where we’re at this point in time. We’ve had the single founder strategy. We now have a different angle to it in terms of co-piloting. It’s just a different co-pilot. Instead of advisers or outsourcing or other mechanisms, we have AI and AI agents.
Bertrand Schmitt
I think that the way it’s going, as you say, that we’re already solo founders. I think AI is just increasing the odds to have a successful solo founder. I think if you are technical enough, you can control AI and understand AI. I think it always goes back to some question. How mature is a founder? Can he grow into the job? Can he learn not just his side of the business, but the other side of the business? It’s always true that it’s tough to learn from the business side to the technical side. That is something that you will rarely learn on the job. As we said, unlike Steve Jobs demonstrated, as a more business person, you can learn becoming a good product person. You can learn to inspire and somewhat control more technical teams. That part, it’s definitely possible.
Nuno Gonçalves Pedro
Conclusion
In conclusion, do you still need a technical or business co-founder? We think probably not. Not certainly at the early stages of the company. You probably can get away with not having a technical co-founder or business co-founder. You’ll probably pay for it earlier on, if you’re missing that technical co-founder, so the technical co-founder does seem to be more important in some ways, and more difficult to replace even by AI than a business co-founder. But it’s not fully, in any case, yet a substitute for real human judgment. Having a dynamic where you have a team, and you have a co-founder that can help you, that can support you, that can take you to the next level, that can commiserate with you when things are hard is probably still better than just going through the use of an AI co-pilot or an AI agent, so we’re almost there, but not fully there yet. Thank you, Bertrand.
Bertrand Schmitt
Thank you, Nuno.
View all episodes
By Bertrand Schmitt & Nuno G. Pedro
5
2828 ratings
Can AI be a co-founder? Do you need a technical co-founder, any more? How about a business co-founder? What can AI do for you as co-founder? Will this become the “brave new world” of start-ups, small and medium businesses? For this and much more discussion, a no BS perspective on AI as a potential co-founder.
Our co-hosts:
Our show: Tech DECIPHERED brings you the Entrepreneur and Investor views on Big Tech, VC and Start-up news, opinion pieces and research. We decipher their meaning, and add inside knowledge and context. Being nerds, we also discuss the latest gadgets and pop culture news
Subscribe To Our Podcast
Bertrand Schmitt
Intro
Welcome to episode 70 of Tech DECIPHERED. Today, we’ll talk about AI as a co-founder. Can AI be your co-founder? Do you need a technical co-founder anymore? Or do you need a business co-founder anymore? What can AI do for you as a co-founder? Will this become the brave new world of startup and small businesses? Nuno, what’s your take on this topic? Have you started seeing that for early stage startups, the AI co-founder?
Nuno Gonçalves Pedro
Yeah. We start seeing this notion of people now—”Oh, I could be a single founder.” I mean, single founders have existed for a long time. We’ll come back to a little bit the taxonomies of founding teams. But definitely, is now a little bit of a trend where people are like, “Well, I don’t need a co-founder. I’m just going to go do my own thing.” Normally, the case is made more for, “Oh, I don’t need a technical co-founder. I can vibe code and put some stuff together and go through things.”
As we go through the episode today, I think we’ll go into the details on why technical co-founders might still matter and why there are certain areas of technical founding that might not matter as much where AI can really be your co-pilot, so to speak.
The Classic Case for Co-Founders
But maybe let’s start with what is the case for co-founders? Why do you need to have a co-founder? Why can’t you just do it yourself? Historically, there’s been really sort of an angle where there’s sort of these two entities in the founding team. Right? The business founder and the technical founder.
The business founder is the person that runs business related activities. If you’re doing, for example, enterprise software, SaaS, et cetera, your business co-founder is responsible for go-to-market like hiring sales—in particular, at the beginning actually being the person who does sales, establishing partnerships, creating and managing elements that are more related to admin with the help maybe of third parties around finance, et cetera. The business founder is more the person who’s focused on the business aspects of the company, which would be go-to-market, which includes sales, channel partnerships, marketing, et cetera. Might include, as I said, the admin side, et cetera.
Then the technical co-founder is more focused on elements that are connected to the technology stack, development of the code base. If it’s just software, development of software, could be more on the product side as well, someone who’s more of a product manager, et cetera. That’s why you need those two entities. Because you need these two entities, so to speak, these two people. Because you need someone who has more knowledge of how to develop a code base, how to get it off the ground, how to develop the MVP, the minimum viable product early on.
On the other side, you need someone who figures out: how do we get this thing to market, how do we actually deploy it, how do we monetise it, how do we create partnerships if they apply. That’s why we’ve had this classic case for founding teams. Now just to be very clear, it is also true that we’ve also had single founding teams for a long time. We have companies that have been founding only for one person. But at the end of the day, the ethos has been, let’s have two co-founders, one on the technical side at least and one or more on the technical side and one or more on the business side at the very least.
Bertrand Schmitt
That’s true. That’s what you have typically seen. Maybe, going back historically, maybe the most famous example of this technical co-founder plus business co-founder has been the founding team of Apple—Steve Jobs with Mr Wozniak. Steve, on the business side; Woz, as he was called, on the technical side. That has been that maybe that’s reference point for all of Silicon Valley for decades. At the same time, it has not been true for every successful company. If I take HP, for instance—if I remember well, they were both technical co-founders: Bill Hewlett, Dave Packard. If you take Microsoft, Bill Gates had, in some ways, a lesser co-founder, but both were technical. Paul Allen was technical like Bill Gates.
Nuno Gonçalves Pedro
I would say Gates was more incredibly technical, but he became a business guy. I think that was part of the secrets of Microsoft early on. He was very business-savvy in some ways. He was very technical. Until this day, I think he’s extremely technical and extremely intelligent, but he was definitely very business-savvy.
Bertrand Schmitt
He was able to wear two hats in some ways. If you think about HP, it’s probably a similar story. When we talk about the one-person founder alone by himself, the question would be, is it a technical co-founder or business co-founder or a mix of both? Someone who can do both even early on or who learned to do both pretty quickly. Probably, Bill Gates would be a good example if we assume Paul Allen was less critical.
Nuno Gonçalves Pedro
Yeah. I think that’s a good point, and you went back to one of the OGs like Wozniak and Jobs. I’d say Jobs was business clearly, great marketeer, et cetera. But he was also very deeply involved in product. Interestingly enough, think maybe not as involved on product early on as he became later on with NeXT and then his second coming, so to speak, to Apple. Where you get much more involved in product management and how products were deployed. But to your point, he wasn’t a technical co-founder. Clearly, that was not the role.
I think if we look at some of the great success like Google, Sergei and Larry were both very technical. I guess they brought in Eric Schmidt very early on to be the CEO. But at the end of the day, the definition of founders in Silicon Valley, maybe we’ll come back to that later, is at least you need to have technical founders early on. Then maybe the business people will come along or maybe one of the technical people will go more onto the business side. But you need to have this duality of role that you, Bertrand, we were just talking about. Someone who either becomes more business-savvy, more business side, but definitely someone who’s more on the technical side. I think that’s how the Silicon Valley has emerged over the last many decades. Right?
Bertrand Schmitt
Yeah. It’s interesting you brought in the story of Google with Eric Schmidt because you could argue they brought him in as a CEO, but he used to be CTO at—was it Novell? I forgot.
Nuno Gonçalves Pedro
Novell. Yeah. I think he was CEO of Novell at some point. Yeah.
Bertrand Schmitt
He was CEO at some point because I was going to say, clearly, he has also a technical background. Able to do the business side. No question. But he was also coming from the technical side.
Nuno Gonçalves Pedro
He was definitely technical at the beginning of his career, and he was also at PARC, Bell Labs, and a variety of other organisations. He was in software engineering at Sun Microsystems, to your point. Then he became the CEO and chairman of Novell. But he is a technical guy by background, for sure. No doubt. He has a PhD. But he was brought in to be the CEO. He was brought in to be more the business guy. I think, very honestly, Silicon Valley loves the story of actually the technical CEO. The person who has at least very product or deep product knowledge and understanding of technology, so it doesn’t get bullshitted necessarily by the rest of the engineering team, so to speak.
Now there’s notable exceptions to that. There are companies that were built very heavily on people that were incredibly business-savvy and very much on the go-to-market side. But, again, justifying why you have co-founders, why you have the technical person and the business person at the very least early on in the structure of the company.
Bertrand Schmitt
If we look at more recently, of course, we have the exception of Elon Musk starting and running so many companies in parallel. Some he didn’t technically funded. He was early on investor, took over very quickly. But he’s also an interesting example of a very deeply technical person with very strong business acumen, obviously. That’s another example. Do you have examples of many successful companies just started by a very business co-person in the tech industry? Obviously, outside tech, it’s a different story. Yeah?
Nuno Gonçalves Pedro
Not a lot, but I feel like in the enterprise software side, there are people that have managed to be great at doing it because they are amazing at building sales teams, and amazing at then building engineering teams around them. I’m not sure I can start giving as many examples as we’ve given on the other side of people that did have a very strong technical background so to speak. I don’t know if he’s a technical guy. I do think he had the bachelor’s in electrical engineering. One guy is David Duffield. He was the guy who founded PeopleSoft and then later on founded Workday.
Bertrand Schmitt
Yeah. I was looking for him, actually.
Nuno Gonçalves Pedro
David Duffield, to my knowledge, I think he has a bachelor’s in electrical engineering and then MBA. But I feel when he started his companies later on, he was very much on the business side. He had gone through marketing at IBM, and did a variety of other roles that led him to create PeopleSoft in 1987. My perspective is probably that he had become more of a business guy later in his career.
But to your point, I don’t think there is a ton of examples. Jobs, I guess, maybe was a great example of that because he was very business. He wasn’t as technical. I think he always had a great product mind, but he wasn’t very technical, from what we know about Steve Jobs.
Bertrand Schmitt
Of course, we have Larry Ellison, founder of Oracle. I think he did it alone. No? I don’t have memories of any co-founder of Oracle. He was technical originally.
Nuno Gonçalves Pedro
Yeah. I’m not sure if he had a co-founder. Actually, that’s a really good question. These stories are never very well told, so one never knows. My knowledge, I don’t think he was. I do think he’s he was very sales-driven from the outset when he was doing Oracle.
Bertrand Schmitt
No. Actually, he co-founded Oracle. Bob Miner and Ed Oates were the other co-founders.
Nuno Gonçalves Pedro
He founded Software Development Labs, right? Yes. Then they changed names at some point. They changed names a couple of times, actually, so he did co-found it. But he was more of the sales guy from what I understand. He was more the hardcore go-to-market guy, so to speak. Really impressive how he built that. I think that’s another good example of someone that was more on the business side that managed to scale a company dramatically. Really interesting.
Not a ton of examples, but probably if you guys are listening to this, and you have a lot of examples that you want to share, you share with us. Maybe there have been some certainly on the enterprise side, enterprise platforms, or Software as a Service side that have been incredibly successful. Do feel free to share with us on X or on LinkedIn.
Bertrand Schmitt
Going back to AI as a co-founder, do you see AI as a potential technical co-founder or more like a business co-founder? What’s your take?
Nuno Gonçalves Pedro
AI as the Technical Co-Founder
Let’s start maybe with the technical side, because I think that’s the most interesting one to debate. Could you be a business-savvy person that just uses artificial intelligence as your technical co-founder, and you don’t need anyone technical early on? There’s now a lot of discussions that people have built an MVP in a couple of weeks, raised $500,000, raised a bunch of money and went forward.
I’m not sure I’d call it an MVP, a minimum viable product, but building a proof of concept or something early on with the help of AI, I feel is eminently doable these days. With vibe coding, with all the tools that people have out there. I am still of the opinion, again, being a computer engineer by background, that at some point having technical founders on board or technical people on board very early on is pretty important. The reality is, we’ve had this discussion in one of our previous episodes, is we need to go back to the methodology of software development.
For those who are listening to us who are not very technical, for a very long time, just a very quick rehash, the world was based on this methodological approach to software development, which was called waterfall, where you would first think through design, the architecture of the system, and only later you’d code it, and then you test it, and then you deploy it. It looked like a waterfall. It looked like every stage has its own part.
Then at some point in time, I would say probably with the advent of the Internet late ’90s, we started going to a world that is definitely more going through a methodological approach that now people call broadly agile software development. Again, if you guys want to go back and listen to the different methodological approaches, we’ve had episodes on it in the past, so you can definitely Google it and search for it.
The world has gone to agile. What that means is that, with going through an agile software development where it’s a lot more iterative, where you’re deploying code a lot more iteratively, you might be under the impression that, actually you don’t need to architect systems very early on. You just need to fake it until you make it. You just need to develop a code base that you can use, and then you go and basically, it works, or it doesn’t work. I think that’s still a mistake. I think the world has gone too agile. We discussed that in one of our previous episodes as well.
Architecting the platforms and what you’re doing early on is important. I think that’s one of the key reasons why you still should have a technical co-founder, and it’s going to be very difficult for AI totally replace it. Because I think the parts of AI that you can use as, so to speak, as a technical co-founder are the parts that would be done by a relatively junior developer. If you’re not very good at asking AI for what you need in terms of code, the code you’re going to get is going to be crap. The architecture and the underpinning and everything that’s underneath it, the platforms that you’d use in terms of your cloud architecture, et cetera, will also be crap.
Basically, you can develop a proof of concept. Maybe you can even get to an MVP early on, but at some point in time, you’ll need to rewrite the whole thing. You need to get technical talent in, just to figure out, is this code base unique or not? For applications that are just software, basic apps, et cetera, potentially can get away with it early on, but at some point, you’re going to pay the cost of it—of not having had a technical co-founder, of not having had any help in terms of architecting the system early on, of not having had any senior software developer early on prompting the AI or working with the team or developing it him or herself. You’re going to pay that cost at some point in time.
Now for deep tech stuff or heavy-duty platforms, whatever, there’s no way in hell. Because that’s not something that AI can help you today fundamentally innovate on. It’s just the level of knowledge you need to have to build it. It’s intrinsically linked to a technical co-founder, so to speak. My perspective on it: it’s great to if you’re doing software, if you’re doing something that’s little bit simpler—be it consumer or enterprise software—maybe you can get away with AI as your co-founder early on. You’re going to pay the price later on, so having at least senior engineering team early on that can help you or can help you think through things is something that actually has a lot of value. I feel the more you delay it, the more you’re going to pay for it later on.
Can you raise money? I think it’s very difficult. Again, unless you figured out some angle on your go-to-market or the business side or something that allows you to have a bunch of traction, retention engagement early on because you figured out some angle, that’s great. But you definitely won’t have developed a tech mode. That’s the piece that at some point in time you’re going to have to pay for. What’s your perspective, Bertrand?
Bertrand Schmitt
I think as you say, it depends on the industry. If you are in deep tech or core infrastructure, it’s quite obvious that you need someone already technical to lead that because the technical side will truly be your key differentiating factor. You need that not just to raise money, but just truly successfully execute. If the product is a bit less technical like many SaaS products are, to be frank, I guess it’s a different story.
For me, the question would be what type of non-technical funder are we talking about that can benefit from an AI co-founder? If we are talking about the business co-founder who truly has no clue about how you build stuff, I guess it won’t really work. Yes, maybe you can fake it to go to the first level of the model, raise a few $100,000s, and then find your co-founder.
I will say where it gets more interesting is if you are technical enough yourself, have a good business sense, and just don’t want to have a technical co-founder for some reason because you don’t want to share the equity, you don’t want to share the control, or you just want to have some level of control on the whole business, but you don’t have the time, you don’t have the true expertise. I think that’s where AI can be quite interesting, actually.
I must say myself, I’ve been quite impressed of how I’ve been able to use AI recently to develop tools, to develop products, to build infrastructure in the cloud. Yes, it requires me to spend some time to understand what I’m doing because I was not going to just trust blindly the AI. I spent some time discussing with it options, and I must have been quite impressed how it was coming up with a level of expertise that is actually not so easy to find. Because when you build out infrastructure and stuff, you have to think at so many levels—from security, to Linux configuration, to network, to storage, to development stack. You’ll realise that it’s actually quite rare to find anyone, even a good CTO who has truly expertise on all of this stuff. That’s where you realise that there is a lot of value if you kind of understand what you are doing, and you don’t let your AI go wild. Because it’s very clear that if you let your AI go wild, step by step, it goes wrong. It’s very clear.
From my perspective, if you are technical enough—like a product manager, for instance, who did some engineering studies, learned some computer science. I’ve not really been deep technical, but understand how all the pieces of the puzzle fit together, have some understanding about what is state-of-art in the industry, but didn’t rise from the ranks of coding and developing or network administration, that sort of stuff. I think you might get real value from AI today as long as you are ready to control it relatively tightly.
For instance, vibe coding, I have absolutely no trust in that. I think it’s just plain crazy when you see how it’s working. But I think you can do some development where you are helpful significantly with AI. You can make the architecture decisions with AI. I think there is a lot you can do if you have some, I will say, technical flair, but not deep technical expertise.
Nuno Gonçalves Pedro
Yeah. I mean, it’s unavoidable that AI will take over a lot of coding right now, that people will be using AI agents to develop code for them. I think there’s Y Combinator Winter ’25 data that 25% of the startups are using 95% plus of AI-generated code bases. I think that’s a bit scary, but anyway.
Bertrand Schmitt
That’s scary for me, as of today’s technology.
Nuno Gonçalves Pedro
It’s very scary. Yeah. Back to the point, you will pay for it at some point. There’s another unintended consequence of everything is AI-coded, so to speak, which is we often forget that companies, when they’re early, they might have a very clear vision of what they want to become and what product they’re developing. But at some point in time, they might have to, shockingly enough, pivot and go to other areas. If we think about pivoting as part of the intrinsic development of a startup through its life, that you serve intrinsic, maybe at some point you’ll need to change, in terms of what product area you’re going after, even what business you’re going after, a lot of it sometimes actually is educated by the tech stack that you’ve built in the first place.
We’ve had amazing failures in companies, notable gaming companies that failed as gaming companies that led to the creation of great companies. I won’t flag App Annie, but App Annie came out of a gaming company. Slack came out of a gaming company as well. There’s all these things that sometimes happen, and it happens because someone developed code that was really, really good on something else.
That allows you to figure out maybe we should pivot. Maybe we should go after that space instead. Maybe we have more of a platform angle here rather than an actual app angle or vice versa. We’re more in an app angle than a platform angle here. That doesn’t happen if you’re just coding stuff that serves a purpose. There’s no actual quality of code that would allow you to, at some point, go and pivot to something else. I think that’s a little bit the unintended consequence of also having such huge dependency on AI-generated code bases and potentially not having a technical co-founder just because you’re like, I don’t want to deal with technical co-founders, and it’s too complicated, et cetera.
Bertrand Schmitt
Yes. I mean, you could argue before AI, what were people doing if they didn’t want to have a technical co-founder or couldn’t find a technical co-founder early on? We had many situations, I guess you heard as well, about teams that are outsourcing their early development to some sweatshop far away. We thought control was talking about what’s going on. In some ways, that’s similar to AI vibe coding today. You have no clue what’s happening. You ask for change, and you get some change, and all the stuff got broken. It was available in a different format, and definitely more expensive. You had the option before, and obviously, that was not a great sign. I guess some very early investor will accept it if there is a plan to change that, basically, and make a tech transfer of some sort.
Nuno Gonçalves Pedro
It’s very interesting because I’m going on two different portfolio companies of mine. I’ve gone through two different at the same time two different stages. One where the two co-founders weren’t highly technical. One of the co-founders was really a product person, though, product manager, et cetera. At some point in time, they believe we want to control more of our engineering team, so they started having more and more of their engineering team in house and less of it outsourced. Very interestingly, another co-founder, actually, co-founder who’s the CEO of a company, she is a computer science background, and she can code. She’s actually going the opposite direction now. She’s like, I want to outsource more and more of my team actually, because I can control quality.
I think what’s really important is can you control the quality of what you’re getting on the other end, back to your point. If you can have visibility on what you want, you can have some ability to have oversight of the code base that’s being generated, you know it’s solid, and you know it’s very strong, and the quality of the code is strong, I mean, it doesn’t matter if you outsource it or not.
To your point, you could use AI for a lot of stuff if you’re an individual contributor or if you’re a founder, et cetera. If you don’t have that knowledge, if you don’t have that understanding, if you can’t judge the quality of a code base and the architecture that your team has put together, then maybe you should have a technical co-founder at some point in time or a very senior hire very early on to the company that can help you make those decisions.
I feel AI—as a full on technical co-founder—maybe can get away with it early on, stage-wise, and maybe pre-seed. Maybe you’re early days of figuring stuff out, but at some point, you’re going to pay the price for it. Be it at the series C stage or a series A stage, you’re going to pay for it. If you’re really, like, I’m not very technical, and I don’t need a technical person, I feel at some point you’re going to pay for it. At some point, someone’s got to ask, where’s your senior technical person. That’s keeping the trains going, either CTO, VPN, et cetera.
Bertrand Schmitt
You talk about evaluating the quality of the code and everything. I mean, you could argue you could use AI for that. I’m sure AI these days is not so bad to evaluate code quality. How is it structured? How is it documented? How which standard is respecting? I think, actually, AI for this type of task is really not bad. You as a founder, that might help you understand that if you decided to use AI to build, you could use AI to review. If decided to outsource to build, you could use AI to review. By the way, if you use outsourcing these days, how are you sure that, I mean, it’s not a lot of AIs behind the scenes.
Nuno Gonçalves Pedro
Correct. I mean, you can’t be sure unless you have, again, some technical knowledge, or you’re checking the code in some way. I think my verdict on can AI be a technical co-founder is yes early on, but not long term if you are a company that has some level of technology differentiation, be it in software or physical tech. That’s my two cents.
Bertrand Schmitt
I agree with you. I can see AI as a shortcut early on as actually a way to improve things early on. Because one thing I’ve seen a lot in early stage startups is that technical co-founder who is good but not great, you’re a bit worried. That’s why you are like, actually, AI might help evaluate or might help a company, might help make better decisions. Someone who was mediocre or good become much better, thanks to AI. That could be another perspective.
Nuno Gonçalves Pedro
AI as the Technical Co-Founder
Maybe moving to AI as the business co-founder, what do you think? Can AI be the business co-founder? You’re technical, and I just want to have a business co-founder, but I don’t want to have a business co-founder. I’m just going to use AI for it. Can you get away with it?
Bertrand Schmitt
I think that’s an interesting perspective. I think it’s quite similar to the technical side. I think AI can help you a lot to improve things, make better things. For instance, designing part of the product, designing your logo, picking your colours. I think AI can help you get much better if you were a technical co-founder. Can it truly replace a more business co-founder? I’m not so sure because a business co-founder will what? We help find financing. We’ll help put in place partnership. We’ll help you close some big deals. This is stuff that you cannot really do with AI. There is some stuff, business side, that AI can help you with. That’s for sure. To help you set up a company, to help you do some legal reviews. I think there are a lot these days that AI can help. In terms of truly big value add, I’m not sure.
Nuno Gonçalves Pedro
I have a bias. I always felt the need for a business co-founder early on wasn’t as significant as others thought. Now you, I think I identified the key issue with not having a business co-founder early on, which is, for me really fundraising. Because everything else you can definitely get there early on through help on marketing, on content, on SEO, on your sales collateral stuff, on pitch decks, or whatever. This is true not only I think of B2B SaaS, but I think it’s also true of consumer. You can be a highly technical founder and just find a lot of stuff online. Now with AI, it’s even easier. You’re really interacting in a way that’s almost like a co-founder, so to speak.
I think the problem starts if you need to raise a bunch of money. If you start having to have a discussion around raising a bunch of money, then you have to have that salesmanship of pitching your company, identifying the markets you’re playing in, being very clear on what’s your go-to market, for example, again, you’re in B2B SaaS. That, I feel, is difficult to do if you don’t have the business co-founder. Now it could be that the technical person is very good and a bit of a hustler type. I mean, often type when we talk about co-founders, we talk about the business person as the hustler and the technical person is the hacker, so it’s the hacker and the hustler. It might be that the hacker in this case might be a hustler as well, but it’s very uncommon. It’s very uncommon that the person has salesmanship. The person whose technical has salesmanship. I think that’s the only key issue that I see.
That said, you can get away in my opinion, I think you can get away with it more being a technical founder that doesn’t have a business co-founder and for longer than the other way around, than you being a business founder and not having a technical co-founder. That’s at least my bias. You can get away with it longer and deeper if you are that technical founder yourself. That’s my two cents on this area—is you have the advantage still to be the technical person.
Bertrand Schmitt
I certainly agree. I think the usually, most of the moat of a company goes to the technical side. It’s not always the case, but it happens often enough. If you’re an investor, even if you’re a buyer, you will feel more comfortable when you see that in front of you, you have someone who understands the technology, who understands the pieces of the puzzle, and who you feel can deliver on it. As you say, I think there is a belief for investors that early on, do you really need that business co-founder? Because early on, for instance, maybe you are deep down just coding for two years to get your product out of the door, and you might have two years where you don’t really need that business co-founder. Could even be nearly a deadweight in some ways.
One thing I’ve seen also is a belief from some investors that I mean, as we talked during Google early days that, hey, you have the great founders, great technology. We’re just going to attack a CEO or a COO, and that will all work out. No big deal in a way. It’s easier to find that than to find the right product and the right business. Adding that later on, there is a market for that business person who wants to join a high-flying startup. I think that’s definitely how things have been working for a while in Silicon Valley. In the same direction, there has been a new approach maybe the last 10, 15 years where instead of replacing that technical co-founder, it’s been to make sure you nurture that technical co-founder. He’s becoming more business-savvy over the years, and there is no real need any more to have a business co-founder. You will have a COO, chief of staff, or whatever.
I mean, a good example, we now talk about Mark Zuckerberg from Facebook, but, I mean, he clearly came from the technical side. He stayed CEO, but he had some COO who helped him on the business side of things, but were still running the show. I think that’s another good example.
Nuno Gonçalves Pedro
Yeah, I think the points you’re making, Bertrand, are extremely valid. Maybe two things that I would say. One piece is we have had some companies, just to be clear, where there was huge distinctiveness. You could argue it’s more on the product technical side or less, but for example, I think the stuff that Nikita Bier did at TBH before they got bought, it was very centred around growth hacking, and we’ve talked about it at a previous episode.
To my knowledge, his background is more non-technical. It’s more on the business side. There’s sometimes you do have a business founder, co-founder, et cetera, that does identify an angle that is very unique either if it’s B2B, a go-to-market angle, or if it’s on the consumer side, some growth hacks or some way to really get in front of consumers, et cetera. We don’t want to downplay the role of a business founder here, and there’s a lot of knowledge that gets attributed in terms of distinctiveness to it.
That said, I feel this is probably an area where, again, the technical founders are going to get away more with not having the business co-founder than the other way around. We already mentioned at the beginning of this episode a bunch of cases where the team that started was really technical—I mean, Google, et cetera. Having the ability later on to bring in a senior person that acts more as the grown up and adult like we could argue Google did with Eric Schmidt back in the day, like you were just mentioning Mark Zuckerberg did with Sheryl Sandberg.
I think that’s definitely something that from the perspective of the play book of a company has its scales is more widely accepted, certainly in technology. I guess, as it goes, the tech mode is probably more definitive and more important than any other piece. Probably goes without saying that there’s a little bit of bias in our comments. We are both technical by background, but we probably feel that I’m speaking for Bertrand. Don’t know if you’ll correct me in a second, but we probably feel that AI can likely be a business co-founder to a technical founder or a human technical founder longer and more successfully than the other way around. At least today, we’ll see how things evolve over the next few years.
Bertrand Schmitt
Yes. Again, I can see AI as, if not a replacement, a booster. Can you actually create better teams faster because they are complemented by AI at very low cost? AI is there to help them make better decisions all the way. On the technical side, architecture choice, some coding, code reviews. On the business side, helping you generate better copy, better design, better logos, better this. I mean, create your photography for your website. There are so many things that now, for instance, if you’re a business co-founder, you could do your job much better. I think AI as a booster, as an enabler, that could also be a way to go.
Nuno Gonçalves Pedro
The beautiful thing about AI as a co-pilot thing is you obviously have immediate access to very distilled and synthesised information. Again, in the case of business best practices, that can be very powerful and very helpful. Even, again, if you do have business co-founders. Just checking on quick things. I mean, I use AI myself for my own stuff. Either it’s a social media post for Chameleon or leveraging other things. Obviously, I’ll rewrite stuff, and I’ll write it the way that I think it’s my voice. Having that at your fingertips is very powerful. In any case, the co-piloting is hugely valuable even if you decide to have a technical and a business co-founder along the way.
Bertrand Schmitt
Yeah, I can see in the past, if you are a younger start up, there are so many things, so many decisions that you have to take, and it’s very tough to find the right level of expertise. It takes time to find it. It costs usually significant money to pay for it. Well, now we say, hi. When I say now, these days, because I’ve just said two years ago, even one year ago, it was not the same quality—but now the quality you get in some of the answers you receive, if you ask the right questions, it’s getting quite impressive. I mean, the added value you can get to have a better architecture product, a more secure platform from the get go, better design. I just feel there is a lot more you can do, a higher quality you can do, thanks to AI, relatively earlier.
Nuno Gonçalves Pedro
Investor Sentiment & Founding Team Dynamic
Maybe moving to our last piece then, which is what is the investor sentiment now in everything that’s happening with AI as co-pilot, AI as co-founder, and how does that change as we look at founding team dynamics going forward where we’ve had the two technical co-founder, the technical and business co-founders, two technical co-founders, all sorts of founding teams, how does that evolve over time?
I would say on the investor sentiment, maybe just to lay the ground on my area, at least, of focus, we spend a ton of time really looking at tech and science modes in a company. Obviously, at the top end, we look at product modes, but then we go under the hood and try and understand technologically and scientifically what are the underpinnings of the product that this company is about to deploy or has already deployed. If we can’t find anything, if we can’t find modes, if we just go under the hood and there’s no technical element to it that’s very differentiated, you’re like, I’m not sure I’m got to invest in these guys. Why would I invest in this team or in this founder if it’s a single founder company? If it’s all AI as a co-pilot on the technical side, maybe it’s a little bit of a difficult decision for us to come and invest in the company.
Now, again, we’re series Seed, series A investors, so maybe at that point in time of the development of the company, we would expect that there are some senior technical talent at the table, even if it’s not a founding talent. That’s, I think, our skew on the technical co-founder discussion where AI is the technical co-founder.
On the business co-founder thing, we do appreciate, and we do spend a lot of time looking at the company and in particular to the CEO, founder-CEO position if the person can raise money for the company. We do spend a lot of time around salesmanship. Is this person that… After we give this person money, can this person go and raise more money in 18 months or 24 months, et cetera, from other larger investors. Does this person have ability to sell?
The story of the company that they’re building. Is this person, for example, if it’s in B2B, is this person able to actually sell, sell the product and go to market? Founder led sales early on is the name of the game. If the person, again, doesn’t have that savviness, that ability to articulate the story of the company, that edginess around either marketing or sales or something else, probably will also be a pass for us.
Where do we get to? I mean, I think it would be more likely that we invest in a highly technical team that has a very low business savviness, but has something very distinctiveness and differentiated on the technical side, then we would do the opposite. We would invest in a founding team or a founding person, a company that is incredibly business-oriented, amazing hustler or whatever—but has no technology mode whatsoever, no ability to create anything in the foreseeable future. I’d say that’s probably where our biases are as a firm today. Now it doesn’t mean we won’t do the second one, but it’s just more unlikely. I think that’s how we would frame it.
Bertrand Schmitt
I guess that’s also for me, what I’ve often seen is a discussion with the team, with the founder in these situations of, are you okay bringing in a technical co-founder? Are you okay bringing in an external CEO, or a CEO to manage the business side? Have you started discussion in that direction? Do you have leads? What type of people are you thinking about? I think that could be part of the discussion, depending on the stage of the venture. I must have seen that quite often, that discussion. Even if, from my perspective, I still have a preference with the founders—the technical founder, or relatively technical founder staying the CEO. Instead of bringing an external CEO, bringing more a CEO type of person.
In some ways, I prefer to see the CEO build up his game on the business side as necessary, versus the replacement like it used to be in the Google days. Personally, I’m a bigger fan of growing up a technical background CEO to stay as a CEO for the long run.
Nuno Gonçalves Pedro
Basically, we both have similar biases, I guess. When will we have the solo founder AI thing be happening all the time? I think pretty soon. We already have solo founders as a really significant part of the business. We already have right now, I’m pretty sure, the next big company that is a solo founder using AI as a co-pilot being created. I don’t think there’s anything dramatically different from all of that. I feel that’s where we’re at this point in time. We’ve had the single founder strategy. We now have a different angle to it in terms of co-piloting. It’s just a different co-pilot. Instead of advisers or outsourcing or other mechanisms, we have AI and AI agents.
Bertrand Schmitt
I think that the way it’s going, as you say, that we’re already solo founders. I think AI is just increasing the odds to have a successful solo founder. I think if you are technical enough, you can control AI and understand AI. I think it always goes back to some question. How mature is a founder? Can he grow into the job? Can he learn not just his side of the business, but the other side of the business? It’s always true that it’s tough to learn from the business side to the technical side. That is something that you will rarely learn on the job. As we said, unlike Steve Jobs demonstrated, as a more business person, you can learn becoming a good product person. You can learn to inspire and somewhat control more technical teams. That part, it’s definitely possible.
Nuno Gonçalves Pedro
Conclusion
In conclusion, do you still need a technical or business co-founder? We think probably not. Not certainly at the early stages of the company. You probably can get away with not having a technical co-founder or business co-founder. You’ll probably pay for it earlier on, if you’re missing that technical co-founder, so the technical co-founder does seem to be more important in some ways, and more difficult to replace even by AI than a business co-founder. But it’s not fully, in any case, yet a substitute for real human judgment. Having a dynamic where you have a team, and you have a co-founder that can help you, that can support you, that can take you to the next level, that can commiserate with you when things are hard is probably still better than just going through the use of an AI co-pilot or an AI agent, so we’re almost there, but not fully there yet. Thank you, Bertrand.
Bertrand Schmitt
Thank you, Nuno.
More shows like Tech Deciphered
View all
The Daily
112,880 Listeners
Eixo do Mal
16 Listeners
The Prof G Pod with Scott Galloway
5,559 Listeners
All-In with Chamath, Jason, Sacks & Friedberg
9,957 Listeners