OpenAI’s latest version of ChatGPT chatbot, ChatGPT-4o, has improved reasoning abilities but tests showed mixed results. The bot struggled with logic puzzles and made errors in reasoning. The new version also features a humanlike voice and the ability to tutor a child on solving a geometry problem, but these features have not yet been released. The current version can analyze uploaded photos and do real-time language translations, but it still has a machinelike voice. Overall, the new version did not show significant improvements from the previous version and has faced criticism for not delivering promised features.