[ad_1]
The World Government Summit just happened in Dubai. Speaking at the summit was OpenAI CEO Sam Alton. He dropped some news about the next-generation GPT model the company is working on. According to reports, Sam Altman talked about GPT-5, and how it’s going to be “better at everything.”
Just recently, OpenAI announced Sora. This is the company’s AI text-to-video generator. It was able to create some extremely realistic videos that could fool most people. Currently, it can produce up to 1-minute long clips based on text input. Right now, the company is only testing this internally, so people in the general public cannot use it.
Sam Altman talks about GPT-5
The World Government Summit just happened, and we got a little bit of insight into what OpenAI is planning on doing next. The company already has GPT-4 and GPT-4 Turbo. These are the most powerful AI models that are available to the public. Under those, there’s GPT 3.5, which users can access with a free ChatGPT account.
Right now, all these models are extremely powerful and capable of deep understanding and reasoning. So, it’s hard to imagine that the company could push things even further. However, according to Sam Altman, it’s “going to be better at everything across the board.” That’s to be expected, but we’re all wondering how much better it will be. Obviously, every Leap forward makes AI more comparable to human intelligence. An AI that’s as smart as a human is called AGI (artificial general intelligence).
As for how much more powerful GPT-5 will be compared to GPT-4, when it’s coming out, where it’ll be available, and whether it will give human beings a run for their money is yet to be seen. We’re certain that the company is doing some very early testing with this model. One bit of information that we have about its capabilities is that, via Microsoft, GPT-5 can decipher a 3,000-year-old language that no person in today’s society can understand.
As for speculation, we’re pretty certain that will include image generation via DALL-E, video generation via Sora, and possibly audio generation. However, that remains to be seen. We know that it’s going to be multimodal, so it’s possible that it’s going to be true.
And “Gobi” is its name-o
Currently under the code name “Gobi”. We don’t have an exact release date, but the company does plan on releasing it sometime this year. Seeing as chatGPT originally launched in November of 2022, maybe the company will stick to releases around that time frame. That, along with everything else about gp5 remains to be seen.
[ad_2]
Source link