This image represents all of human knowledge before AI:
Press enter or click to view image in full size
Humans have explored and researched in many directions. There’s frontiers in every direction, but also many blank areas in between that we just don’t know much about. Often that’s simply not because we can’t explore it, but just because we haven’t had a chance to explore it yet.
Each square represents some piece of human knowledge, hard-fought and hard-won by centuries of researching and learning how to do things. Each square has a different color. There’s variation. Different humans and different cultures do things differently. Knowledge is shaped differently.
This is what that looks like after AI:
Press enter or click to view image in full size
AI is like a flashlight. Everywhere you shine it, it will fill things in. Make things better. But all in the same homogenous grey way. And never beyond the border of what humans already know, which I call Imagination Space in this diagram. AI can solve Millenium Prize problems, but can it find new ones? Can it come up with revolutionary new physics that humans haven’t yet shone their own light on? I remain skeptical.
AI is currently adding a lot of value to humanity, and that is because there are still so many gaps within the border of Known Concepts. Filling in those gaps, even if it’s just with grey goo, benefits humanity. But at some point that will no longer be true. When we reach that point, making those squares ever more grey will become a detriment to humanity, rather than a benefit.
..the irony will not be lost on you that I used AI to generate the two diagrams here. And it didn’t get it right the first time either. Through repeated iterations, more and more artifacts kept creeping in. The same thing will happen to the repository of all human knowledge if we let AI run wild.