This post discusses the capabilities and limitations of AI in performing mathematical tasks. While some commenters believe AI can indeed assist with automated mathematical work, there is skepticism regarding its deeper understanding of mathematical concepts beyond surface-level computations. AI models, such as Large Language Models (LLMs), might handle formal math with ease akin to smart graduate students, but struggle with more complex quantitative and geometric reasoning tasks. The general consensus among experts appears to be that while progress has been made, AI still lacks a foundational grasp of mathematical concepts that would allow it to genuinely understand mathematics like a human, particularly at advanced levels. This raises questions about the nature of 'doing math' versus merely mimicking solutions based on pattern recognition and existing data.