Translated data: LeCun and xAI co-founder Christian Szegedy have disagreed over the reasoning capabilities of GPT-4, sparking a heated debate. LeCun argues that GPT-4 has significant flaws in its reasoning abilities, which has led to widespread discussion. Szegedy suggests potential pathways to breakthroughs in reasoning capabilities. Some netizens compare GPT-4 to human thought, questioning the objectivity of human intelligence. A paper by Subbarao Kambhampati of ASU University found that GPT-4 performs poorly in planning reasoning, with only a 12% accuracy rate.