Sam Altman, the founder of OpenAI, the developer behind the ChatGPT artificial intelligence model, revealed that the research team at the platform is working on adding a feature for video content generation. This enhancement will allow the platform to create various types of content, including text, images, and also comprehend audio clips.
During a conversation with Microsoft co-founder Bill Gates, Altman mentioned that ChatGPT is undergoing significant advancements in handling content and understanding the context of information provided by users.
Altman explained in the Unconfuse Me podcast that the next generation of artificial intelligence models developed by OpenAI will offer more advanced capabilities in dealing with different forms of information. These models will be better at understanding user-provided information, especially concerning spatial descriptions and material details of scenes. They will be capable of designing diverse environments in the form of images and videos, enriched with effects and sounds.
OpenAI started training its upcoming artificial intelligence model, GPT-5, last year, and reports suggest it may perform at a similar level to Google’s Gemini Ultra model.
Altman surprised Gates by stating, “These models in 5 to 10 years from now will be the dumbest AI models.”
During another panel discussion at the W24 conference for Y Combinator founders, one participant on the “Ask the Investors” platform noted that Altman advised against creating companies relying on business ideas related to solving problems with the GPT-4 model. Altman suggested this because the new model will arrive soon, followed by the era of Artificial General Intelligence (AGI). AGI refers to a level of artificial intelligence that is generally capable across various tasks.
Leave a Reply