The AI is getting better at things like maths. I recently asked it about iterating the Burrows-Wheeler transform, and it appeared to really understand that. It's not super easy to reason about why it's reversible, etc. and I felt that it got it.
This is obviously not AGI, and we're very far from AGI as we can see by trying out these LLMs on things like stories or on analyzing text, or dealing with opposing arguments, but for programming and maths performance is at a level where it's actually useful.
My answer to this is - so what? What are the effects in the real economy?
Theres probably a good 2-3 years left of runway for substantial progress before things really fall off a cliff. There has to be some semblance of GDP (real) growth.
I think it might be possible to automate development of short programs. I also think it might be possible to reduce the confusion, misunderstandings and cliches, so that the models can work for longer stretches.
But people probably expect to get the next version for what they pay in subscriptions now, so I can't imagine much more revenue growth for the model companies.
This is obviously not AGI, and we're very far from AGI as we can see by trying out these LLMs on things like stories or on analyzing text, or dealing with opposing arguments, but for programming and maths performance is at a level where it's actually useful.