ChatGPT wants to become as sharp as a doctoral student

Mira Murati, OpenAI’s CTO, gave some insight into ChatGPT and hinted that the upcoming GPT-5 language model will be as expert as a PhD student.

This start of summer is busy in the artificial intelligence sector. Between updating and creating a business, everyone is doing their part to be at the forefront of innovation. Today, OpenAI is speaking out about the feats that its next version of ChatGPT could achieve through its technology manager.

ChatGPT at university

In an interview with Dartmouth Engineering posted on YouTube, Mira Murati, CTO of OpenAI, talks about the evolution of GPT language models (LLMs). Last May, a blog post suggested that OpenAI was working on an ambitious model: “ we anticipate that the resulting systems will take us to the next level of capabilities on our path to AGI.“. Artificial general intelligence (AGI) capable of doing almost everything is arriving step by step.

“If you look at the trajectory of improvement, systems like GPT-3 were perhaps of child-like intelligence.”explains Murati. “And then systems like GPT-4 can resemble the intelligence of high school students. And then, in the next few years, we plan to obtain a system with intelligence equivalent to a doctorate, on specific tasks.”

The artificial intelligence industry is rapidly evolving and improving. Asked about the future availability of this model, Mira Murati is measured. This new model would not be released before the end of 2025 or the beginning of 2026. Enough to work on the regulations to be put in place on these new computer assistants in the professional environment. This OpenAI manager stating that “These systems are already on par with humans in specific tasks” .

Words which echo other missions such as those of Safe Superintelligence which declared on June 19 “Superintelligence is within our reach”pursuer“We plan to advance our capabilities as quickly as possible while ensuring our security always stays ahead”.