The architecture of computational intelligence is a marvelous amalgamation of mathematics and physics, but the equations and algorithms in the AI stack merely describe fundamental laws and not discover or create them.
It’s not a question of scaling current AI models. In their present state, they won’t yield new laws of physics or math.

The great Turing Test is no longer relevant; a new benchmark for computational intelligence is needed. The renowned @DrBrianKeating proposes ‘The Keating Test’ to assess AI’s ability to discover novel STEM laws.

Thoughts inspired by a discussion between Dr Brian Keating &  Prof John C. Lennox on the ‘Into The Impossible’ podcast