OPENAI Live Stream – Could we see a major Chatppt upgrade?

Back to live demos, Openai shows that we can now chat with Chatgpt more visually. This means that you can ask for requests for images in a line, and it will remember the context.

In this example, a photo of a room was sent, then the team asked Chatgpt to make it transparent, among other requests.

Since improving the generation of images is already available – or at least deployment – the editor -in -chief of Techradar, launches Ulanoff, has already tested the functionality.

Lance took a selfie and downloaded it on Chatgpt via the iPhone application. He then asked that he was transformed into an anime style. The first time, it gave him full hair, but then corrected when he asked for it to be bald.

Sam Altman has also teased that the native images generation model in 4o is designed to be a bit offensive in time if this is what you direct. The key sentence is “in reasonable reasons, and many users will undoubtedly put this to the test.

Now, the second demo requires a colorful image describing the theory of relativity, with a little additional humor. Altman also noted that the image generation model is a little slower but that the result is much higher in quality.

In the second demo, the Openai team took a selfie and then asked Chatgpt to make an “anime style”. It took several seconds, but he has indeed generated what was requested. You can see it above.

Sam Altman was then quick to note that the improvement in the image generation is starting to deploy now in Chatgpt and Sora for professional users, and it will also be available for free users.

We also see the process of the generation model of native images in the 4O model, transforming which generated a selfie into an “meme act”.

Openai launches the generation of native images in the 4O model

And we leave for races – Sam Altman calls this one of the most fun progress, and it is the generation of native images in the 4O model. He quickly noted “it is a huge step forward” and something that Optai was excited to deploy for a while, for a whole multitude of people.

Altman notes that the best way to explain it is to show it, so we are already in a demo. In just a few seconds after the invite, Openai showed an image with what the team said to “a perfect text”. Apparently showing a jump in terms of understanding the prompt and creation of the image with clear text, and a unique point of view effect.

The live flow of Openai began, and at the imprint at the GMT start time of 2 p.m. and / 11 a.m. Some of them overlap, but it refreshes every few seconds and shows all the different styles.

The description of the live flow notes that we will hear Sam Altman, Gabriel Goh, Prafulla Dhariwal, Lu Liu, Allan Jabri and Mengchao Zhong discussing the generation of 4o images.

Well, the title Livestream is much lighter to look at that we can expect … Much more than the intimate image tanned. It is entitled “Generation of 4o images in Chatgpt et Sora”, which means that we are probably obtaining improvements to create images in Chatgpt and Sora.

The mention of the latter could also mean more general improvements for the generation of video texts.

Less than 15 minutes to leave now!