If I'm trying to solve a generic problem, I try to break it down into pieces which I have familiarity with and barring that I will look to google to fill the gaps. I might try think of alternative approaches and compare them, and reflect on individual steps to try and improve them. I'll review my work and test it as I go to ensure it is actually solving the problem.
Given some of the capabilities I have seen with chain-of-thought processing, embeddings, and vector databases - it does seem to me conceivable that computers could be made to do all of this. One gap maybe is being able to pick up that the problem statement is wrong, i.e. not solving the right problem, or being aware of subtle undocumented details about the system and the business that need to be factored in. The AI needs to actually know about these subtle details somehow in order to factor them in.
Perhaps that is the big problem that remains - capturing subtle details about the world.
Given some of the capabilities I have seen with chain-of-thought processing, embeddings, and vector databases - it does seem to me conceivable that computers could be made to do all of this. One gap maybe is being able to pick up that the problem statement is wrong, i.e. not solving the right problem, or being aware of subtle undocumented details about the system and the business that need to be factored in. The AI needs to actually know about these subtle details somehow in order to factor them in.
Perhaps that is the big problem that remains - capturing subtle details about the world.