New benchmark study results show leading AI models still lag humans in visual math reasoning.
Read Full Story at Decrypt.co →