Why Can’t AIs Like ChatGPT, Gemini, and Claude Solve 8.8 – 8.11? A Look into Their Math Struggles

Large language models like ChatGPT and Gemini excel in processing language but struggle with mathematical calculations due to the nature of their training data. They primarily learn from text, with some capabilities in handling images and videos, but their math skills are limited. Rather than performing precise computations, they predict answers, often turning to web searches which can lead to inaccuracies. The design philosophy behind these models emphasizes aiding in creative tasks like writing, where efficiency can significantly reduce effort. However, for mathematical queries, it’s more effective to use a dedicated calculator app. To these models, numerical values like 8.11 are treated as character sequences, making them less reliable for math. Thus, while large language models serve well for linguistic tasks, it’s advisable to approach their mathematical outputs cautiously and trust traditional calculators for accurate calculations. Mastery of math by AI is still a few years away.

Source link