No ChatGPT-5 yet, OpenAI's CEO Sam Altman confirms
OpenAI CEO Sam Altman has clarified that the company is not training GPT-5 yet, putting an end to the rumours. During a virtual interview at the Massachusetts Institute of Technology (MIT), Altman revealed that the company is working on upgrades and updates for GPT-4 instead. Altman also addressed concerns related to ChatGPT and emphasized the importance of taking time to study the safety of the model. He mentioned that OpenAI spent over six months training GPT-4 before releasing it to the public.
With the launch of GPT-4, the company has not shared any information on the AI bot’s training data or its architecture and construction. This lack of transparency has raised concerns about the safety implications of large-scale models like GPT-4. The company's report published alongside the GPT-4 release did not include details about its architecture, model size, hardware, training compute, dataset construction, or training method.
Altman affirmed that OpenAI plans to continue its transparency practices going forward. However, he also mentioned that the open letter, allegedly signed by Elon Musk, which demanded a pause in developing GPT-5, lacked most technical nuance about where the pause is needed.
Altman asserted that "moving with caution and an increasing rigor for safety issues is really important." The company is doing further work on top of GPT-4 that is of interest to address safety concerns related to ChatGPT. Even though there is no ChatGPT-5 in training, it is only a matter of time before the AI community gets an update on the release of GPT-5.
