"Creative blocks - The very laws of physics imply that artificial intelligence must be possible. What’s holding us up?"

https://aeon.co/essays/how-close-are-we-to-creating-artificial-intelligence

which warned about the trap of believing science is based on induction. But the article also left me wondering about whether induction has any role in AGI other than for classification systems. IOW ML might best thought of as advanced "curve-fitting" and, while it has a place in AI, it is a small well-defined place. ML would not give us AI.

Following up on Karl Popper's criticisms of induction referenced in Deutsch's post, I turned to Philosophy Bro:

"Karl Popper's 'Introduction to the Logic of Science:' A Summary"

https://www.philosophybro.com/archive/karl-poppers-introduction-to-the-logic-of

and was relieved to find that I need no longer worry about the problem of induction.

It appears the mass rush to ML is a mania to be soon satiated by the limited amount of unexamined information available. In the end we will have many datasets, each with a set of ML classification algorithms, each an island of information and no overarching theory or framework, thus leaving a greater gap toward AGI even more obvious than before.

So my question is: "When are we going to put aside the ML madness and get back to working on AI?"