Is there a limit to Abstraction? (Other than the Poverty of our Imagination?)

And how to benefit AI from this question

Innovation Science points to abstraction as a productive innovation route. Recently popularized by Elon Musk, the diving into 'first principles' has proven itself innovation-productive, and this raises the question: how deep can we go? Is abstraction terminal or is it infinite? Abstraction requires imagination, a heavy doze of it. So when it looks like we discovered the bare essence of a subject, are we really at the bottom, or are we simply imagination-exhausted.

Group theory and symmetry are hailed as the mathematical abstractions that underlie modern physics. If we can dive a notch deeper, we are bound to discover even new physics. Is there a 'deeper layer'?

One must say 'yes' on account on applying abstraction in a self-referential way, namely abstracting the abstraction process. The common description of abstraction is that it is the process of washing out details and staying with what is left. The idea being that two subjects of different details when stripped of those details will appear very like and similar at their details-free level. This notion can be abstracted and described as the process of constructing similarity between two dissimilar matters. So abstraction will not be viewed as unilateral effort over a given matter but as a binary effort over two (and then more) matters, representing both as similar to each other.

This novel description of abstraction is great consequences in today's world of AI. Similarities are the building block of artificial intelligence. AI is so powerful because it gives answers based on extracting conclusions and inference from a large array of dissimilar matters which are recognized as similar owing to an effective process of abstraction.

Self applying this new definition of abstraction to "Artificial Intelligence Assisted Innovation (AIAI)" we have discovered new grounds and achieved new efficiency. Highly recommended!