ChatGPT 5 will be as intelligent as a doctor of science
OpenAI CTO Mira Murati returned to the Dartmouth Engineering school as a graduate student and talked about the upcoming ChatGPT 5. She mentioned the release date and the possibilities of AI at the PhD level.
Mira MuratiChief Technology Officer OpenAI gave an interview to Jeffrey Blackburn, from Dartmouth Engineering, in which she presented the possibilities of artificial intelligence and provided some information about the upcoming ChatGPT 5.
Murati announced when the next version of OpenAI’s generative artificial intelligence can be expected to be released. Work on GPT-5 will last nearly 1.5 years – so you can expect this to be the case end of 2025 or beginning of 2026.
In terms of the sophistication of ChatGPT 5 compared to ChatGPT 4, Murati describes the progress by comparing AI with human development from high school to university level: :
If you analyze the improvement path, GPT-3 systems demonstrated toddler-level intelligence. Systems like GPT-4 have the intelligence of a smart high school student. Over the next few years, we are targeting the intelligence to perform specific tasks at the Ph.D. level. The situation is changing and improving quite quickly.
Particularly important is the fragment about intelligence to perform specific tasks – AI is developing in various aspects and reaching the level of “human intelligence”, but only in certain tasks. As the director of OpenAI points out, implementation of many tasks is still beyond the reach of ChataGPT.
The conversation devoted to the impact of artificial intelligence on humans and their environment lasted almost an hour, and the entire recording can be watched on YouTube: AI Everywhere: Transforming Our World, Empowering Humanity.
