There aren’t many examples of people making money selling AI. However, there are several good examples of people making money selling companies that work on AI.
These examples, that involve great big corporate giants acquiring companies that develop AI for a myriad of applications and varied verticals, have set an interesting precedent: the purchased companies do not sell. They might not even have potential buyers for their product, or worse, they might not even have a clear target market! So not only do they make no money, but they oftentimes do not even have the semblance of a plan to make money.
One would think that the purpose of a company is to make money. Well, these corporate giants have so much of it that they are not buying a money making company, but an idea that comes packaged in company-like wrapper.
This is all fine, as corporate giants instrumental in effectively changing the world can afford to pay company-size price tags for maybe-not-even-commercializable ideas. The problem arises when corporate investors are laser-focussed on reproducing that precedent by checklisting companies they are considering for investment, by emphasizing their buyability (by corporate giants in an speculative short-term future) rather than their viability as money making entities.
Oh yes, that golden ticket, that 2-5 year ROI. Just imagine: As an investor, you find a little company with an AI idea, full of Ph.D.s busy with R&D projects, with no business guidance or experience, fresh out of university, not really trying to solve a problem that people are willing to pay for, some of them even trying to prove a theory in the mystical realm of quantum computing. You, as a VC, would throw so much money at them, that when (not if, but when) Google buys the little company for millions, your ROI would be the biggest slice of that pie. Just imagine!
The active imagination of investors leads them not to wait for these opportunities, but to actively search for them, consciously discarding perfectly viable little companies in their desperate search for buyable little companies.
This is also fine, because VCs can do whatever they want with their money (or at least whatever fits within the board’s guidelines for what they can do with their money). After all, they are in the business of making money investing in companies. The thing is that the paradigm around AI companies has changed slightly: From making money by investing in companies that make money (or have a solid plan to do so), which would typically make them attractive to subsequent buyers; to making money by investing in companies that might be bought before they even become a startup.
As fascinating as this is, the only small problem is the mixed message viable little companies get when potential investors are quick to dismiss them for a few Ph.D.’s that have never had a job with a hypothesis that barely mentions AI - and call THAT a company worth investing the big bucks in. The issue here is that little companies following this trend will not only be unviable from the start, but that they also face the great risk of being unbuyable. Being viable should not only be a safety net if proven unbuyable, it should be a company's raison d'être. Trends, ignorance and confusion will come and go, and viable little companies will always find a way to make money, even if it is by definition.