Skip to content

Introducing ChatGPT Images 2.0



Gabriel Goh, Kenji Hata, Kiwhan Song, Alex Yu, Boyuan Chen, and Nithanth Kudige with host Sam Altman introduce and demo …

Tags:

41 thoughts on “Introducing ChatGPT Images 2.0”

  1. Before the test was to see if this image models can generate proper text. The new test is to ask them to generate an image with a provided chess position on a nice chess board. Even this new model can't do it…even though its supposedly thinking.

  2. But What are you doing to HELP People KNOW THESE are NOT a REAL Images. People make things up to sway people opinions. Where are the Safety Features? How do they talk to Social Media? The Web?

  3. With this it'll get harder to distinguish between whats real and whats fake. Real history and facts will get harder to discover, fakes will be everywhere. Make a robot to do my dishes and laundry! We dont want this

  4. Chinese people everywhere 😂😂😂 We have a lot of smart engineers from America and Europa. Why did you chose asian engineers ? Now I understand why chat gtp answers very poor and slow.

  5. Hey everyone, and yes, I liked this update, but I want to talk about something else: the fact that after Sora 2 was shut down, the developers focused on developing ChatGPT. I'm not saying that's bad, it's good, but the fact that the company shut down its most popular project because of its own mistakes seems stupid to me. Remember what they said: "We can't control content." We don't have enough budget, but I bet if I were the head of OpenAI, I would have prevented both problems: the fact that they couldn't control content is their own fault. They didn't introduce any rules other than the use of people unless they were characters on their platform, nothing else. There was no talk of censorship yet. They talked about how people had learned to remove watermarks and that this problem was also due to the development of artificial intelligence technologies, and that this can no longer be eliminated, because watermarks can be removed from far more than just Sora2 videos. That was the first problem.
    And now about the second problem, the budget, which is why OpenAI terminated its contract with the Disney Company. I don't remember the exact amount discussed there, but the fact that the decision was made very quickly. Well, about the budget, they said that they had invested a ton of money in this platform, but excuse me, if you give each user 30 generations a day and through… Of course, you'll lose 50% of your company's profits for a small period of time. That's figuratively speaking. Even image generation in ChatGPT is much more restrictive, with only three generations, depending on usage. I'm not saying it's right, but it's very low. But the point is, where does OpenAI's generosity come from? And in general, Sora 2 seems different in some ways, as if it were made by a different company, but that's just my opinion. And the final problem is that they admitted it's not their calling, that they're focused on higher interests. But let's not, we're already adults, and we all understand that there are many competitors in the market for creating videos using artificial intelligence: Nano Banana and others, I won't even list them, and if you couldn't keep your project in your own hands, it's not a problem, it's in your interests that's why I wanted to write about it, so even the Russian neural network Alice AI is already catching up, okay, after all this I want to say, I'm not trying to insult anyone, I'm not trying to offend anyone or point a finger, I was just expressing my opinion, because the technologies that we have thanks to OpenAI are incredibly powerful – this is the creation of images, videos, music, texts, etc. Let's do this, if you really liked my opinion about all this, you will appreciate and spread it, because for many, the ability to create videos so easily, without a budget and other means served as an opportunity to bring their creativity to life, including for me, let's vote on the relaunch of the Sora2 model or will it be Sora3, I don't know if OpenAI has such thoughts, but I hope I was able to influence at least something.

  6. Похоже замеры IQ в США не врут – самые умные – евреи-ашкеназы, потом жители юго-восточной Азии (Япония, Корея , Китай), потом белые европейцы, потом латиноамериканцы , и потом пришельцы из Африки. Ребята-азиаты на первых ролях что у Альтмана, что у Маска.

  7. How many of these engineers will get the offers and leave in a week or two? Maybe it’s better to learn from the guys like Steve Jobs, who was always made sure that the products are presented by the people who are compensated so much that they can never leave, and have very senior level position

  8. Soon people will be creating 2hr long movies in minutes off prompts. Custom movies thought up by you, not someone else. Actors, directors, costume designers, who needs them?

  9. "Images 2.0" implies there was a satisfying 1.0 — and we all remember what that looked like (six fingers, nightmare hands, text that spelled nothing in any known language).

  10. who idiot made this shit? I am asking it to make separately this dumb shit make multi image in one, fire the idiot who made it like this, you made a shit which doesnt give a fuck to user prompt

Comments are closed.