Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The Ex-Googler team behind the Rooms for 3D rooms from Things, Inc. has a recent project: an AI editor mixing. The iOS-only app allows anyone to create new AI-generated images with “recipes,” which are like Mad, write-nonsense images on your photos, notes, or drawings.
For example, you can ask mixing Turning your painted chandelier into a beautiful Remaissance painting, or fixing your funny pet. You can use the self-test to see what you would look like with a different hairstyle, or do something else – like review a friend who has changed to Italian.
The program is built on top of google’s nano banana, but the “open” model opens a new way of communication with the model – changing the creative tool in computer games.
“The thing that (Nano banana) did was an amazing example that it could take your picture and save it in a way that was not meaningless,” Jason Toffwhose background includes working on test companies within large third-party companies like Google and Meta, and during which time they manage sales.
But what makes the application even more complex is that the “recipe” of the program – which is created by AI – is important.
“Ai is very powerful, but often you go to these tools and it’s like your text box – what am I writing with?” Toff said, talking about the mistakes his team saw with the existing environment.
“And so, instead of being creative and thinking about what we can do, you see something that works, and you can just fill in the blanks,” he said.
Users create a new mix, they can choose to publish it together with a photo that contains a public feed or just download it for personal use. From the feed, some users can see the image and tap the button next to “try.” This allows everyone on the program to re-use the process of creating a picture using their own picture, words, or puppets. (The latter can be made through a simple, taped section.)
Viewing an image from the side of the path it’s created in can also solve unexpected AI images, the team believes.
“Another problem (and ai) is what we refer to as the ‘Slot Machine’ problem.
By mixing, however, users can see the image and the speed that created it in one place, giving them an idea of ​​what it means. They can also disable the button to see everything before it happens, if the creator can stop this.
In addition, similar to the open source AI software Aividiai Shera, users can upload their own images to mix their own AI models. If you choose to do so, then anyone you follow in the app can create AI avatars with your profile – the profile is called “mixed.”
The company thinks that groups of friends follow each other to use the section, but the Creator’s class can grow on the platform – if they don’t want to call themselves strange ways. (Of course, if you don’t want your picture out there, don’t post it or don’t follow anyone.)
The program also uses Operai’s technology to solve some of the problems expected around AI simulations, but Toff also acknowledges the extensive integration of Google’s controls within the context of banning content or violence.
At launch, integration is guaranteed for iOS 26 but supported on iOS 18 and up. If it goes away, a Web app or Android app can be added later.
Free users receive 100 conversations, which is equivalent to $ 4. Currently, images cost about 4 cm to create. After savings, users can subscribe to different cloud services for 100, 250, or 500 regions per month.
The app launches at midnight on November 21 on the App Store worldwide, but requires a login.
This was updated after the publication of the notes to make the program available on Nov. 21, not Nov. 20 as previously stated.
