Different from any AI I've seen: Why this 3D modeling program is suitable for anyone

I've tested a lot of AI-powered creative software, and when it comes to image (now video) generators, there seem to be two options: you enter a prompt, you cross your fingers and want the AI ​​to know all the details and directions you want. Or, you don't use AI at all, you have to do everything manually in Photoshop or other programs. I can hardly see a good middle ground… until the invisible AI.

Invisible is one of many startups that utilize newly generated AI technology. Its flagship tool is a 3D modeling program that allows you to type text prompts and create actionable 3D scenes that can turn them into still images and videos. You can get that kind of control, like you rotate the camera on a collection (or virtual set) without manually creating every element in the scene. All of this is browser-based, which means you don't need to download the software. This is different from any AI image or video generator I've seen before.

At first glance, you may not immediately think of the times when you need to use 3D models in your daily life. However, this may be useful for those who are visual learners or those who are struggling in spatial intelligence and trying to imagine and move through physical space. Invisibly, it is not only designed for animators, but also used to visualize advertising ideas or to plan physical event spaces such as wedding venues.

“If you're making a Hollywood big production or triple game, you won't stop intangible. But for many people, the quality you get from intangible assets will be far enough for web applications, presentations and tone materials,” the intangible Charles Migos told me in an interview.

The invisible idea started with Migos being vice president of product design at Unity, a global gaming software company. But when it was obvious that Unity could not launch the product, Migos left in 2024 and was invisible with co-founder Bharat Vasan. Since then, intangible assets have received $4 million from VCS and other Angel Investors. Currently in a private beta. Invisibly, Beta users include designers working on CGI-intensive films, including Star Wars and Shrek.

Migos confirmed to CNET that the intangible assets will be released publicly in June this year.

Invisible primary product designer Philip Metschan gave me an early exclusive study on how the program works. At the text prompt, he creates a city full of buildings, reminiscent of a 3D, AI-ED New York City. He takes the character from a batch of pre-created elements and makes her a runner. He personalized her appearance with text prompts—the same as you do other edits—and adjusted her posture to get her running. He can change the entire viewing angle of the shot, bringing her into the foreground and adjusting the exact angle.

Metschan spent nearly 20 years at Pixar, producing much-loved movies like Wall-E and Incredibles 2. He is familiar with detailed work of detail, as familiar as a high-quality film, including a high-quality animated film – including how long it takes to get a group of leaders to reach a consensus on the creative direction. AI tools like invisible are one way to help speed up the process, as they can manually laugh at a lot of ideas for a fraction of the time. Even if AI images or videos have never been used in the final product, AI can help bring artistic visions into life and help people sell people in general aesthetics or directions.

“Decision is the most expensive thing in the production pipeline,” Metshan said. “Any tool that can speed up the process or make the iterative process more efficient and collaborative is always great.”

Metschan showed me an example of how invisible can help speed up the process. He created a 49-second AI video in an invisible way to imitate TV commercials. From start to finish, it took him only 3 to 4 hours to produce, greatly reducing the typical production process of storyboards, position reconnaissance, casting and shooting. He can also make significant changes at any time during the creation process, such as changing the weather of the scene. The program can save the way you characterize elements with prompts, and you can use those pre-existing elements on other canvases. So, like Metschan showed me, he could walk through the red jeep on the forest trail in snow, rain and sun.

In the invisible interior, there are two different types of models that work: one can identify and process your prompts when you ask to create things while the other creates these elements. Invisible use of Chatgpt and Grok to help run things, you will be able to choose between steady diffusion, flux and Clin to visualize the scene. You will have any images and videos that are invisible, but so far there are no visible watermarks, so you need to disclose using AI.

Invisible leaders know that not all creators use generative AI. Creators have expressed many concerns, from work safety to potential copyright infringement during model training to the flood of “AI Slop”, online is becoming increasingly common. Migos hopes that technology like invisible can expand creativity, one of many invisible human qualities that AI will never replace, and therefore the company's name.

“[Creators] “They are masters of storytelling. They understand the audience's ideas and the truth and beauty of art forms. These are the people we want to empower.”