Can you really build a billion dollar company by yourself with AI?
Though nobody has seen one yet, solo unicorns are all the rage. But can AI turn individuals into whole companies? And if so, is it a good idea? Some thoughts from the vibe frontier.
Hey everyone, greetings from sweltering Zurich. I can tell you I’ve never been happier to be living in a city with its own lake than I am now. Hope you are doing well wherever you are.
Today I wanted to write about something a little outside The Assetizer’s normal remit: the idea of the solo unicorn. If you are following the tech startup space you have probably come across this.
For more on GenTwo, see: www.gentwo.com
We launched a new product today, the world’s first universal AMC creator. Learn about it here. And try it out here.
One for all and no one else
A solo unicorn is a billion-dollar company founded and run by one single person. No employees, just an army of AI agents. It’s the logical conclusion of the whole vibe coding movement.
I find this idea tantalizing and crazy at the same time.
Tantalizing because wouldn’t it be nice to be that one person. And crazy for a whole bunch of reasons – the craziest of all being that it’s actually not really a crazy idea.
You may have heard about Base44, a vibe coding side project turned bootstrapped startup that went from nothing to an 80 million dollar exit in six months (it was bought by Wix) .
Base44 is neither solo nor a unicorn (there are 8 employees apparently) but as I understand it was heavily reliant on GenAI to get off the ground. And the speed at which it did so is breathtaking. According to TechCrunch:
In its six months as a stand-alone company, Base44 reportedly grew to 250,000 users, hitting 10,000 users within its first three weeks. According to Shlomo’s posts on X and LinkedIn, the company was profitable, generating $189,000 in profit in May even after covering high LLM token costs, which he also documented publicly.
Granted these aren’t ChatGPT numbers, but still: this is happening.
On Tuesday’s podcast I got into the subject with Erik Wiesner-Mostenicky, a principal at Fidelity International’s Venture Capital team.
He is bullish on the idea, telling me that you can “already see smaller companies creating incredible amount of output on the revenue side with just having one founder operating in that way.”
And it’s not just him. I recently had lunch with David Roon, an old ConsenSys buddy of mine, and he is talking seriously about a new kind of very early-stage venture capitalist approach where the VC’s function is to help solo founders with the vibe coding part (because you still need to understand how that works, and here a helping hand is likely to be useful; that is what Base44 is, an app to help you vibe code).
David is a technologist, and one of his points is that these agents let you prototype at light speed. You can get to better quality, minimum viable products much faster. You can test and iterate at a higher level. And you can get a sense of what you really have (get the vibe of it) in a much more concrete way than you could before. And of course, once it all works, you can scale it like mad (though not for free, see below).
Them that’s got shall get
We’ve certainly gotten into the vibe here at GenTwo, where everyone is being encouraged – and given the tools and the space – to build their own agents to automate their work. And I mean everyone, from the devs and the UX folks right though legal, compliance, key account, finance, sales, you name it.
Heck, even marketing and communications.
My boss talks about building a “hands-free” marketing department, and we’re on it. As the comms guy, my bit is to try and automate all the tasks around communications, content and PR. That’s the whole gamut from research, media monitoring, client and competition monitoring, through to content ideation, scheduling, drafting, editing all the way to publishing, gathering and analyzing stats, and the rest of it.
And there’s the rub.
I can do all this – or at least have a chance of doing it – because I have decades of experience in this business and know how all these different parts of the process work.
As I told Erik on the podcast, sometimes I feel like a junior comms person’s worst nightmare. I can augment myself because I have something to augment (as modest as that may be). What if you don’t have that kind of experience to build on?
My point is that in the end, it is still going to be experience and good and novel ideas that will win. As well as those who get in early and move fast. That isn’t necessarily new.
I also can’t help thinking that if anybody can build a company by themselves, then likely everybody will, or at least will want to. This sounds nice in theory. In practice, it’s likely to lead to an AI-augmented race to the bottom along the lines of what we’ve seen before.
Right now, we are swimming in AI content. In the future, will we be swimming in AI companies? And if so, what are their chances?
Some final thoughts
So do I think a one-person unicorn is possible? Yes.
Do I think it likely? I am sure someone will do it, if for nothing else than to show it can be done (well, that and the money).
That said, I think the most prevalent outcome will probably be many highly-valued companies with very few employees, but not none. We’ve already seen this trend in tech. And we can certainly ask if this is a good thing or not for society as a whole.
Keep in mind that using AI doesn’t mean you have no costs, either. It just means other types of costs. As Erik pointed out, LLMs are expensive. So basically you are paying the model instead of people.
I get the sense that if you are a VC or an investor you will need to be closer to the ground and be able to move real fast. Wix got Base44 most likely because its founder was very vocal on LinkedIn. He was broadcasting. Others may not be.
Also, my chat with David left me thinking that this type of world might call for a different kind of venture capitalist, or rather, will favor venture capitalists with a tech bent, those who can really guide someone through the process of developing the idea from the earliest stages.
Last but not least, this being The Assetizer, I think there could be an assetization angle here. One of the big use cases for an AMC is fundraising, particularly for smaller companies. Instead of traditional raises, clever asset managers who can source nascent unicorns could crowdfund their early development with a tailored product.
Just a thought.
And finally: as it turns out, we launched a new product today that could let you do that fairly easily. It’s called The AMC Creator, and you can learn about it here.
Next week we’ll be deep diving into it in The Assetizer as well.
All the best,
Tom
Why does this scare the shit out of me?