This demonstrates considerable enhancements in person choice and Total quality of open-finished outputs, showcasing improved alignment with user anticipations. None of the GPT-4o or Claude 3.5 Sonnets could reply this straightforward question properly. Only o1 was capable of finding the proper reply with no assistance. Allow’s see how Deepseek performs. https://x.com/kidtsang/status/1884008035535782292