OpenAI's Chief Scientist says "The past few years have been surprisingly slow," and previews that model iterations will continue to accelerate.

CryptoWorld News reports that OpenAI’s Chief Scientist, Jakub Pachocki, said at a press conference that progress over the past few years has been surprisingly slow, and that model iterations are set to continue accelerating. GPT-5.5 is just over a month away from the GPT-5.4 release. He explicitly answered that the release cadence will further speed up, with substantial progress in the short term and extremely significant progress in the medium term. OpenAI co-founder and President Greg Brockman said GPT-5.5 will think faster, more nimbly, and more efficiently using fewer tokens, so consumers and businesses can gain access to even more cutting-edge AI. Chief Research Officer Mark Chen predicted that it will take about two more years for AI to independently complete a full end-to-end AI research workflow. Altman also posted on X outlining three positions: gradual deployment is at the core of the safety strategy; GPT-5.5 is already a “smart model,” but is expected to iterate rapidly; and OpenAI’s goal is to build the most efficient models and reasoning architecture so that everyone can use the best technology, with the aim of becoming “a platform for all companies, scientists, entrepreneurs, and individuals.”