AI’s Struggle to Read Analogue Clocks May Have Deeper Significance

A new study reveals that advanced AI models like GPT-4.1 struggle to accurately tell time from analog clock images, especially when faced with unfamiliar designs. Fine-tuning improves performance on standard clocks but fails to generalize to modified versions. This highlights AI's reliance on pattern matching rather than true understanding, raising concerns about its reliability in real-world tasks involving unfamiliar visual inputs.