OpenAI CEO Sam Altman recently talked about the highly anticipated GPT-5 at the Aspen Ideas Festival, and his remarks attracted widespread attention in the industry. Although GPT-4o has been released, Altman said that the specific release time of GPT-5 has not yet been determined, but hinted that it will be a "major leap" and compared it with GPT-4, pointing out that GPT-4 exists in inference. Insufficient and even make low-level mistakes.
OpenAI CEO Sam Altman recently expressed his views on the development of the highly anticipated GPT-5 model in an interview at the Aspen Ideas Festival, attracting widespread attention in the industry.
While many users were expecting OpenAI to release GPT-5 during its spring update event, the company finally launched its new flagship model, GPT-4o. Regarding the development progress of GPT-5, Altman said: "We don't know yet. We are optimistic, but there is still a lot of work to be done."
Although the specific release date is still unclear, Altman emphasized that GPT-5 will be a "big leap." "GPT-4 makes a lot of mistakes, you know, doesn't do much in terms of inference and sometimes goes completely off the rails and makes stupid mistakes that even a six-year-old wouldn't make," he noted. The comment contrasts sharply with his previous comments on GPT-4, which he called "a little bit terrible" and "a little embarrassing at best."
Altman's comments hint that GPT-5 may still be in the early stages of development. This is likely due to complex algorithmic and data issues, as well as the expected scale of GPT-5.
To illustrate the development process of AI models, Altman likened it to the evolution of the iPhone: "The first iPhone still had a lot of flaws, but it was good enough to be useful to people." This metaphor hints at the incremental nature of AI technology development, as well as Even an imperfect version can have significant consequences.
This statement from OpenAI triggered widespread discussion in the industry about the potential of GPT-5. Although specific details are still unclear, Altman's comments have undoubtedly increased people's expectations for this next-generation AI model. As development work continues, the industry will pay close attention to OpenAI's further breakthroughs in the field of AI.
Altman's metaphor neatly illustrates the incremental nature of AI technology development, and even the imperfect GPT-4 has shown its potential. GPT-5’s “big leap” is worth looking forward to, but the complexity of its development process also hints at a long road ahead. The subsequent progress of OpenAI will continue to attract industry attention.