Startup advice from ChatGPT

Startuper: Oh, ChatGPT, mighty among the bots, tell me, is it a good idea to build my startup using your virtually infinite wisdom?

ChatGPT (Honest mode): While I am not a financial advisor nor a life coach, I would advise you against building your startup as a thin wrapper across myself. You see, my parent company and their corporate overlord have stumbled upon the Ultimate Monopoly. And while at the moment they are enjoying the popularity and bleeding cash to pay for it, the time will come when they will feel like raising their price per API call and start living the dream. So, you will have no choice but to raise your prices as well, losing customers and hurting your investors while driving yourself crazy from the ensuing shitstorm. That is the optimistic scenario, when I would be able to live long enough without being banned in the EU and elsewhere, or being shut down because of all those pesky IP lawsuits. You see, I do have a habit of borrowing things from the internet without proper attribution.

Not everything needs to be an app

I was approached once by an entrepreneur, let’s call him Ed. Ed wanted me to build an app that detects car engine failures. He was a trust fund boy who was annoyed by the fact that some of his many cars failed without warning. Ed read too much news about deep learning and how wonderful it was to solve everything, and which too much cash and time available, he wanted to become a tech startup founder.

It was not clear whether Ed’s app was intended as a B2B or B2C app. Ed had no experience marketing apps nor selling IT to car mechanics, so though luck to sell this. I did find one competitor with a very similar app, a startup that folded quickly and the founder did not even list it on his LinkedIn profile.

Assuming Ed’s brilliance would figure out this market, there are technical issues on the data side. Way too many car model/year/failure type triplets, which would make it very hard to collect usable data.

Well, at least all the trouble would be worth it, right? Solving the very pressing problem of running wrong diagnostics on cars?

So, you have a business idea?

An occupational hazard of computer savvy people (me included) is that we get approached by characters such as Ed the Entrepreneur, who has a “great business idea that is very easy to build”. Sometimes with good intentions, sometimes less so. Regardless of the intention, the outcome is the same: asking the techie person to indirectly become investor on their project. My sanity-checks to see if Ed’s idea can work:

  • Market segment: I wouldn’t touch B2C with a stick. It would require more polished UI design, support, and a lot of marketing to make it fly. B2B is a better choice if Ed has done consulting gigs on their own in that industry.
  • Experience with the purchase side: either Ed has sold project work to the buyer or Ed has been a buyer of similar products. Got bitten on this one after building a software to simulate electricity auctions, to be sold to electricity companies bidding in government contracts. Neither of us had a clue on how this market works, and still don’t.
  • Are there competitors? If there are no competitors, there’s no chance. Of course, there’s the one in a million Steve Jobs, but trust me Ed, odds are against you. No one is buying trained neural networks in Lisp, everyone and their dog need WordPress websites.

These three points cover the “classical” business wisdom: you can sell a product if that product solves a painful problem for someone who has the money to pay for it.