I am less bullish on how good it will be in 2029.

Even looking at red neck battle truck. The amount of time it'd take me to fix the annoying compositional mistakes, I'd rather spend sketching a truck from scratch.

Chat gpt 3 isn't that much worse than 4 or claud.

And before 2029 it will start costing an arm and a leg. It's basically only useful when it's funded by venture capital.

Have you read any ed zitron? I think he's too pessimistic, but I'm closer to that take.

Ai is far along the road to not have a product people want to pay for.

When I use it, it's fun and helpful. But for me, only because it's relatively free.