Mathematical Thinking and AI-Assisted Discovery
Mathematical Thinking and AI-Assisted Discovery examines how artificial intelligence is changing the exploratory side of mathematics. The article frames AI as a discovery amplifier rather than a source of mathematical authority, showing how systems can generate examples, suggest conjectures, write candidate programs, search large spaces, propose proof strategies, assist formalization, and connect ideas across domains. It distinguishes AI-generated output from mathematical knowledge by emphasizing verification through counterexample search, evaluator design, proof, formal checking, and human interpretation. The article also explores program search, geometric reasoning, proof assistants, education, credit, reproducibility, and the ethics of AI-assisted discovery. Rather than treating AI as a replacement for mathematicians, it argues that human judgment becomes more important: framing problems, choosing representations, testing claims, recognizing significance, documenting workflows, and deciding what a discovery means within the broader structure of mathematics.









