I personally hate this train of thought. It sounds identical to the "all great things originally looked like toys" and then this gets twisted to the invalid conclusion that "all toys will eventually be great". His logic also seems to contradict their early validation mantra. I'd appreciate an even more strict filter on startup ideas, preventing so much waste on multiple versions of the same unneeded projects.
I think it's closer to the truth to say that most progress comes not from invention, but in recognizing the implication of things that have already been invented. That can be borrowing ideas from another domain. It can be combining two ideas that never met each other. Or it can be seeing wisdom in something people take as a frivolity (literally turning toys into something profound).
But we also already suffer from a misapplication of this. Everyone wants to be the X of Y (which is basically trying to apply someone else's winning strategy to a different domain), and they think every pair of ideas will be the next Reece's Peanut Butter Cup. But it's still the champ.
It doesn't contradict. Early validation isn't about whether people think something is good idea. It's about whether people will pay. The Mom Test makes this distinction clear.
And there's no waste on startup ideas. Searching for something requires looking in places where there is nothing. That's not waste. That's searching. If you know where to look it's not searching, it's a lookup.
In the end, there's a pretty strict filter on startup ideas: you have to be able to convince someone to give you money.
You need to try some things to see their potential. These people understand it causes waste. This is one of the major theses of Zero to One. It's really hard to know what works in advance. Power distribution of success cases pays for the losers. They know there will be losers, but they don't care.
