Science & Technology

OpenAI Launches Text-to-Video Model

OpenAI, which became world-famous after launching its artificial intelligence-based chatbot called ChatGPT, presented a new Sora product.

OpenAI Launches Text-to-Video Model

The new machine intelligence configuration can generate videos that last up to one minute. The AI model copes with these tasks based on text prompts from the user. The relevant information was published on the company’s blog on the occasion of Sora’s debut last Thursday, February 15.

OpenAI is currently focused on teaching artificial intelligence to understand and simulate the physical world in motion. The purpose of the company’s relevant activities is to create training configurations of machine intelligence that will help consumers solve problems related to interaction processes in the space of objective material reality.

So far, it is difficult to characterize the scale of Sora’s capabilities in terms of practical results and development potential, but it is already obvious that this is an important step by OpenAI toward making artificial intelligence, whose digital consciousness accommodates the surrounding world in all its dimensions and manifestations.

The new AI model from the ChatGPT developer can generate complex scenes in which the process of interaction of several characters is carried out. In this case, the actors have such a characteristic as the accuracy of movements. Moreover, artificial intelligence generates precise details of the subject and background. The company notes that Sora not only understands the essence of the user request but also has a preliminary understanding of how things exist in the space of the physical world.

Also, the new artificial intelligence model can generate videos based on a single image. Moreover, in the list of functions Sora, there are extending existing visual content and filling in missing frames.

OpenAI at the same time notes that the current functional state of its new product is not ideal. The company warns that within the framework of using Sora, it may be difficult to synthesize the physics of a complicated scene or to comprehend the causal relationship. For example, a character in a video may bite out a cookie, but after that, there may be no bite mark of the corresponding action on the cookie.

OpenAI reported that Sora will be tested by experts on issues such as disinformation, bias, and content containing narratives that provoke the spread of hate ideas. Moreover, the company develops tools to help detect misleading information. One of these security products may be a classifier that identifies videos generated by Sora.

Also this week, OpenAI announced a new memory feature for ChatGPT. In this case, it means solutions that allow the chatbot to remember conversations with the user. Consumers will be able to ask the virtual assistant to remember only specific data. The chatbot will also have the ability to independently determine which information needs to be stored in digital memory.

Serhii Mikhailov

2107 Posts 0 Comments

Serhii’s track record of study and work spans six years at the Faculty of Philology and eight years in the media, during which he has developed a deep understanding of various aspects of the industry and honed his writing skills; his areas of expertise include fintech, payments, cryptocurrency, and financial services, and he is constantly keeping a close eye on the latest developments and innovations in these fields, as he believes that they will have a significant impact on the future direction of the economy as a whole.