Google text-to-image artificial intelligence model Imagen debuts publicly for the first time
CTOnews.com, November 3 (Xinhua)-- Google is extremely cautious about releasing its text-to-image artificial intelligence system. Although the company's Imagen model produces output quality comparable to OpenAI's DALL-E 2 or Stability AI's Stable Diffusion, Google has not previously made the system available to the public.
Today, however, the search giant announced that it will add Imagen to its AI Test Kitchen app in a very limited form as a way to gather early feedback on the technology.
AI Test Kitchen was launched earlier this year to test Google's various AI systems. Currently, the app offers different ways to interact with Google's text model LaMDA, and the company will soon add similar restricted Imagen requests as part of what it calls the app's "season 2" update. In short, there will be two ways to interact with Imagen: urban dreamers and Wobble.
In Urban dreamer, users can ask the model to generate urban elements designed around the theme of their choice, such as pumpkins and denim. Imagen created sample buildings and plots (city squares, apartment buildings, airports, and so on), all of which appeared in isometric models similar to those seen in the simulated City.
In Wobble, you can create a little monster. You can choose its material (clay, felt, almond, rubber) and put it on the clothes of your choice. The model will generate your monster, give it a name, and then you can poke it and make it "dance".
CTOnews.com learned that compared to other text-to-image patterns, these interactions are so limited that users can't just create what they like. However, this is also what Google intends to do. Josh Woodward, senior director of product management at Google, explains that the whole point of AI Test Kitchen is to: a) get public feedback on these AI systems; b) test which behaviors cause the system to crash.
But the big question is, will Google want to bring these models to the wider public, and in what form? At a time when competitors OpenAI and Stability AI are rushing to commercialize text-image models, will Google feel its system is secure enough to provide it to users outside AI Test Kitchen?
