OpenAI Chief Executive Officer Sam Altman poured cold water on hopes for the next major release of ChatGPT, which launched this year.
"We're going to be releasing some really good versions later this year! We won't be calling it gpt-5, though," he said at this week's Reddit AMA.
OpenAI will focus on releasing GPT-o1The first of these was codenamed "Project Strawberry," a new AI model from OpenAI that is slower but designed to "think" about a question before answering it. The idea is to make something useful for specialized use cases like science, math, and academic research.
While specific details about ChatGPT's future are scarce, Altman predicts that ChatGPT's "next big breakthrough" will be "agents," tools that can perform tasks independently and interact with the world without human intervention. For example, booking your flight or concert tickets or answering IT help desk service queries.
Srinivas Narayanan, OpenAI's vice president of engineering, also said that in the future he hopes ChatGPT will be able to "better understand my personal information and take action on my behalf".
Altman also hinted that he might one day open the door to "not safe for work" adult content, which is currently blocked by the tool. "We fully believe adult users should be treated like adults," he told one Reddit user. "But it takes a lot of work to get it right, and right now we have more pressing priorities, adding, "We hope to get it right someday!