Each incremental model architecture can handle the last model's Second 90%.

Models are great at doing the First 90%, but not the Second 90%.

That's what creates the illusion of progress: you can create amazing demos, but then the grind of the Second 90% is excruciating, like trying to pin jello to the wall.

But Fable can tackle Opus's Second 90% much more autonomously.

So if you just wait a few months then maybe the model gets good enough to do the Second 90% of that project autonomously.