I don’t know this company in particular, but this happens very often. I can imagine a business plan that said, “In the short term, we will pay people to do this task in order to build up a training data set so that we can train our machine learning algorithms.”
I actually just saw a company a couple of days ago where it was that exact same structure. Only they did it the way I think is right. In their venture pitch, they were very honest, they said, “Here is the current number of tasks that are completed every day on our platform, and here is the fraction of tasks that are completed by the A.I. versus the humans.” And you could see that the fraction is growing. So there is a level of disclosure that can make that plan ethical.
In start-ups you’ve invested in, or in your own companies, are there examples where testing something that doesn’t exist or painting a vision that you haven’t achieved was an OK and necessary thing to do?
It’s hard to pick a specific example because in every start-up, that’s intrinsic to the job. Again, it’s not about deception. It’s about the fact that you’re talking about the future and the future is always uncertain.
I’ll give you an example. I was once raising money for a start-up, and we had a hockey stick-shaped graph in our pitch that showed the number of customers we had and the revenue we had. And I remember showing it to an investor who said, “This is amazing. Congratulations. What are the units on this graph, is…