Renowned mathematician Terence Tao has highlighted a shift in mathematics from an era of 'proof scarcity' to 'proof surplus' due to AI advancements. Large language models (LLMs) are rapidly generating proofs, while tools like Lean automate verification. However, the human ability to comprehend these proofs lags, creating an 'impedance mismatch.' Tao used the Erdős problem as an example, where a student generated a proof using ChatGPT in 80 minutes, but it took Tao 24 hours to verify and understand it, revealing new connections in the process. Tao predicts that academic evaluation systems will need restructuring, as understanding, rather than generating proofs, becomes the scarce resource. He emphasizes that the future of mathematics will focus on the ability to choose the right problems, verify, and digest results, rather than merely producing proofs. This shift is expected to impact other proof-based disciplines, such as theoretical physics and cryptography, as AI continues to transform the landscape of mathematical research.