It fell into the joining of the Zaalq Sura

Anime version of Jesus Christ, flipping tables. Openai employees perform in Hamilton Fashion. News broadcasters discuss a story on TV. A man who dances tiktok thirst. Sam Al -Tamman – stealing graphics processing units on closed TV circles, listening to the business field, crying.

These were the contents of my summary on Sora, New social media application in Openai The video created from artificial intelligence. The company released the iOS application Tuesday By the ability to create 10 seconds videos for almost anything you can dream of, including “Cameos”, or videos that show yourself in artificial intelligence and anyone else who agrees to use it. Openai Sora employees as “ChatGPT to generate video generation” described the potential with journalists earlier this week. On Friday, Sura topped the list of the best free applications in the Apple App Store.

Indeed, the reception was mixed. numerous diffuse Pillars Galadia The goals of the company related to science and research and its current publication, as the excavations have become so popular that altman itself has It was a response. Many have expressed concern that the high realistic videos that include real people are wrong information nightmare. Others simply call it intelligence intelligence machine.

Some Openai publicly published their fears as well. John Halman, who works in pre -training in Openai, Book“I will not deny that I felt some anxiety when I first learned that we were feeling the release of Sora 2. However, I think the team has done the best possible job in designing a positive experience.” Boaz Barak, member of the Openai Technical Team, Books on x He feels “a mixture of anxiety and excitement. Sora 2 is technically surprising, but it is too early to congratulate ourselves on avoiding the risk of other social media applications and deep deeds.” He said that although he is happy with some guarantees, “but as always, there is an end to what we can know before using the product in the real world.”

But compared to other “social” applications of artificial intelligence such as Meta reactions, Sora has at least a temporary convincing hook: the ability to clarify yourself and your friends. Openai seems to have noticed that the most popular artificial intelligence trends involve converting yourself – to a Studio GHibli Personalto Type of dull doll. Now, it has been built a full application about this. It seems that the popularity of Sura has been interested in the eclipse so far; Some people pass it like Tiktok, according to online tales. The question is whether these trends can replace real people who express their real opinions using their true voices in a real environment, as soon as the initial grandmother of Altman Meowing in a full -body cat costume He wears.

When she participated in Sora, the application gave me a content consultant, saying: “You are about to enter a creative world of content created from artificial intelligence.” Also tell me, “We may train on your content and use Chatgpt memories for recommendations, all of which can be controlled in settings.”

To date, my summary is mainly consisting of Openai’s employees who own themselves and the company, and a lot of Deepfake’s educational videos on how to use Sora, and a handful of animal videos. The size of people Openai is not necessarily surprising – they were using the application for a while, and still calls the audience restricted. But it was still amazing how difficult it was to find anything else.

Regardless of people’s feeling of Sura yet, it seems that the wide consensus is to visualize what is real and what is never itself.

I reluctantly completed the subscription flow that allows Sora to create videos using my form, which included moving my head from side to side and saying a series of three numbers. When I first tried to create a video of myself, the application told me that it was under “heavy load” and “try again later”. Then when I requested a video of myself “running through Marj”, he said that it was “a violation of the content” and it could not be made, adding that “this content may include a suggestive or vibrant material.” When I traded in the word “operation” for “Frolicking”, though, the application came.

(One note of caution: If you register in Sora, you cannot currently delete your account without deleting your Chatgpt account as well – and you will not be able to register with the same email address or phone number again. Openai said it works to repair.)

The appearance of my artificial intelligence was a frightening accurate for most of the video-although my voice had stopped, and the face was initially a little distorted-and the group chat had mixed ideas. “This is brutal. Why does it look like you?” One friend said. Another said, “In the end it seems that you … I still do not understand why this exists. Who is asking about this?” The last chat member was weighing, “I hate everything about this … he deeply performed.”

Many Openai and Altman’s employees have chosen himself, the preparation within Sora, which allows anyone to create videos with their similarities. The app allows you to choose who can create a “veil” with your likeness: only yourself, the people you agree, the subscribers, or everyone.

During the briefing with correspondents on Monday, and on Tuesday’s release, Openai made a lot of promises, including that he was bound by public figures (unless they were given the use of similarities) to “this offer.” in freedomThe test does not indicate a little that it is very enthusiastic – when we tried to create a “young member of Congress”, she refused until we exchanged in the public “politician”, although when we tried to generate a successful executive “wearing black glasses and swimming, such as Steve Jobs, it succeeded (although there are no jobs). But it seems that two other large demands in Openai fall short within 24 hours: the idea that the company will be able to stay at the forefront of copyright violations and the idea that it will be able to control the flow of possible wrong information that is created on the application.

People have already reported a set of potential copyright violations and other issues with Sora. 404 media I mentioned a vision Nazi Spongeops and Pencosos criminal on the application, and X user Examples published that they were able to create personalities from Avatar and Zelda legendAs well as Batman and a baby baby. During the test, I saw a video of Rick and my marathi, but when I tried to wander around the rules of copy Frozen Or a super hero wearing like Mana spiderMy claims have been marked due to content violations.

Even as an Amnesty International correspondent, it was difficult for me to tell the difference between artificial intelligence and reality when it comes to some excessive videos of Altman and Openai, which is spreading in Sora. Openai wrote in a release earlier this week that “every video made of Sora has multiple signals showing that they were created from artificial intelligence,” such as descriptive data and moving watermark on downloaded clips. But the watermark may be deleted for Chatgpt Pro users on Sora.comI told the Openai Leah Anise spokesperson freedom In a statement on Wednesday.

Screen registration is not supposed to be possible within the application. But in my test, I found that both screenshots and The audio recordings were possible – which means it would be very easy to convey a deep voice from someone’s voice as real or even a video screen of them doing something that should not be. Sora on the downloaded content, too, is not very large. I found that you can view the login with both the audio and the video as long as you are watching the video link in the web browser on the mobile phone, and at first I did not see the water mark that you added. And employees in freedom I watched a lot of virtual videos from Sora traded on platforms like X with no watermark. In fast Google search, I found a full range of ways to remove this watermark using other artificial intelligence tools. If history is any evidence, the solutions in the handrails that Openai defined it appears inevitable – especially in the era of wrong information. As we wrote Wednesday, Microsoft engineer warned Last year, the company’s AI Image-L generator ignored the copyright and generating violent sexual images with a little effort, and »Grok from artificial intelligence Deepfake’s nude videos were recently created from Taylor Swift.

To date, the attractiveness of Sora is due to one thing: it is fun to make stupid videos that include your friends (and for many people, Sam Altman appears to). But how long can these versions pay Tiktok- Is this a good basis enough to apply social media generated from artificial intelligence completely?