The world, as it's organized, we try lots of things that are bad ideas. While you're building this company that's going to fail, would an AI do that? Would an AI look in a situation and say, I have no idea how to make this thing work? So I won't do it, because I can't see the endpoint. But a human will say, well, frankly, I don't know how this is going to work either, but I think I can get funding. Maybe I can figure it out. There's almost nothing that an AI would do that's the same as what you would do.