AI, Maths & Human Conjecture

If mathematics is viewed strictly as a mechanical exercise of symbols and proofs—a perspective famously proposed by David Hilbert—it may indeed be vulnerable to disruption by AI. However, if mathematics fundamentally encompasses human understanding of truth rather than merely symbolic manipulation and proof , it is likely more robust !

This presents a noteworthy paradox: if mathematicians subscribe to a Platonic philosophy, suggesting that mathematical truths are independent of the human mind, they inadvertently increase the risk of obsolescence. It seems contradictory to advocate for the necessity of human understanding while simultaneously asserting the independence of mathematical truth.

Recently, I formulated a conjecture suggesting that because AI systems rely on data, they remain inherently limited by the fundamental incompleteness of data-based architectures, thereby ensuring that human cognition remains complementary. To preserve the relevance of the discipline, it may be necessary for mathematicians to move beyond purely mechanical proofs and prioritize a human-centric approach to mathematics. I evidently belive that still some of the most creative human ideas are beyond AI...Human Natural Intelligence will have to be improved for not being wiped out by AI.

Otherwise, the professional landscape for mathematicians may face significant challenges.