Don’t get me wrong, I’m not suggesting the current capabilities are anywhere near replacing human productivity. Some things are 1 year out, some 5 (maybe self-driving cars by then? Mercedes has it on their roadmap for 2030 and they’ve historically been realistic), some 10+. But the pieces are in place and the investments are being made. The question is no longer “can AI really automate this?”, it’s “how do we get the dataset that will enable us to automate this with AI?”. And as long as Open AI keeps people’s eyes on their whizbang demos, the money will keep flowing…