OpenAI does not yet teach GPT-5

In the last year, everything related to AI seems to be moving at the speed of light. Since OpenAI ChatGPT debuted late last year, new AI companies and AI-powered mobile apps, web services, and features have been popping up every day.

In the past year, everything about AI seems to be moving at lightning speed.

But there is one area where the current leader in the field, OpenAI, is slowing down: Learning its next language model.

In a discussion at a conference hosted by the Economic Times (opens new tab) in India, OpenAI CEO Sam Altman explained that the company hasn’t even begun work on its next model, GPT-5.

«We have a lot of work to do before the GPT-5», —Altman said. «It’s going to take a long time. We’re certainly not close to it».

As recently as March, a thousand technology luminaries such as Apple’s Steve Wozniak and Elon Musk signed a petition calling for a six-month pause in learning AI language models «more powerful than GPT-4» which is the latest OpenAI model released. The letter calls for continuing AI training only when «we are confident that its effects will be positive and the risks —manageable.

Altman hinted that concerns about AI had some bearing on OpenAI’s decision not to begin work on GPT-5. «More security checks» need to be done, — he said at the conference.

But there are other reasons. In particular, many of the big platforms and tech companies that OpenAI has used to train its models aren’t thrilled that the company’s success is based on their content and freely available data. Companies such as Reddit have begun introducing expensive API subscription plans specifically in response to AI training.
As for when GPT-5 will be released, that remains to be seen. OpenAI didn’t release GPT-4 until March of this year. «I’d like to tell you about the timing of the next GPT» release, —Altman said.