The long-delayed GPT-5 may be significantly postponed, but it is expected to have a significant leap in performance, reaching &#34;PhD level&#34; intelligence.
According to media reports on Sunday, Mira Murati, Chief Technology Officer of OpenAI, recently revealed that the company's next-generation artificial intelligence model GPT-5, expected to be launched at the end of 2025 or early 2026, is poised to achieve PhD-level intelligence in specific tasks, marking a potential major leap forward in artificial intelligence technology.
It is reported that GPT-5, internally codenamed &#34;Gobi&#34; and &#34;Arrakis,&#34; is a multimodal model with 520 trillion parameters, compared to the previous generation GPT-4 with around 2 trillion parameters. This massive parameter scale implies its potential powerful capabilities. Murati likened the progress from GPT-4 to GPT-5 to a leap from high school level to university level, indicating a significant improvement in complexity and capability for the new model.
It is worth noting that this release time differs significantly from earlier rumors. The market had previously speculated that GPT-5 might be released at the end of 2023 or in the summer of 2024, but Murati's latest statement has pushed the timeline to the end of 2025 or early 2026, with OpenAI CEO Sam Altman emphasizing that there is currently no fixed release schedule.
AI expert Alan D. Thompson had previously predicted that GPT-5 might be publicly released in December 2024 (after the U.S. election), but the latest timeline adjustment indicates that OpenAI may need more time to develop a more powerful and intelligent system.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/df2dffe7-98ec-4abf-b03d-671c81eb929c.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/df2dffe7-98ec-4abf-b03d-671c81eb929c.png"/>
This adjustment reflects the company's emphasis on product quality rather than strictly adhering to the predetermined timeline.
Microsoft CTO Kevin Scott emphasized at the 2024 Berggruen Salon the significant leap in AI capabilities, with GPT-5 having the potential to pass complex exams, reflecting significant progress in reasoning and problem-solving abilities. Scott also pointed out that the barriers to entry in the AI field are decreasing, meaning that powerful AI tools will be able to be used by a wider audience.
Murati claims that GPT-5 will achieve &#34;PhD level&#34; intelligence in certain specific tasks. This statement has been supported by industry leaders. Microsoft CTO Kevin Scott stated that due to improvements in memory and reasoning operations, the next generation of AI systems will be able to pass PhD exams, with Murati emphasizing that this high level of performance will be task-specific

OpenAI

OpenAI CTO also revealed that GPT-5's performance will make a huge leap, reaching "PhD level" intelligence in certain specific tasks

GPT-5 significantly delayed? OpenAI CTO said it will be launched at the end of 2025 or early 2026