Sam Altman made several important announcements about what’s coming in the future from ChatGPT.
Despite the proliferation of chatbots in 2023, it is still ChatGPT that attracts attention. OpenAI’s AI is still at the forefront of possibilities in the sector, powered by the powerful GPT-4 language model. A model which should experience a small revolution with the following version, GPT-5, thanks to which users will be able to generate videos!
GPT-5, the “fully multimodal” language model
When we questioned him last summer about GPT-5, Sam Altman tended to want to calm the public’s enthusiasm, explaining that we would have to be patient, and that a lot of work was still to be done.
The change in tone was palpable during the appearance of the OpenAI boss in the podcast Unconfused me with Bill Gates. Because Sam Altman took this opportunity to tell us more about GPT-5. According to him, the language model will be “ fully multimodal “, And ” will support speech, image, code and video. »
A release planned for 2024
This announcement means that GPT-5 will be able to analyze videos, but also, ultimately, like what Dall-E already does for ChatGPT, directly generate videos using a prompt!
According to comments made during the podcast, the new GPT-5 language model should be made official during the year 2024. As for the actual video generation functionality, the function which will surely be the most anticipated, we do not know exactly when it could be offered.
Note that other players are working on this possibility, which should revolutionize the relationship we have with the video format, with an ever greater increase in coming across a deepfake. What will the Internet look like in a few years with all these freely available tools?
Source : Mashable
3