Have you been wondering where Google is amidst this year’s wave of AI text-to-image generators? Wonder no more. It just beta-launched its own AI model called Imagen.
This is in answer to the many AI imaging technologies that came out this year such as Stable Diffusion, MidJourney, and DALL-E.
But in contrast to similar technologies by Open AI, Microsoft, and Meta, that of Google is taking a cautious stance. Users will get only a limited taste through the internet giant’s AI Test Kitchen.
The purpose is to gather community reactions which will be used to further fine-tune the AI tools and get them ready before full release.
To access the tools, prospective users will have to fill out a form and be put on a waiting list. Once given access to Imagen, users will only be able to play with City Dreamer and Wooble.
Like other AI text-to-image generators, you can ask the tool in City Dreamer to generate urban elements based on your descriptions and a selected theme and it will bring one to life.
In Wooble, you can design and dress up a monster based on your text prompts and have it dance when you poke it.
These may be simple actions but the results are what AI Text Kitchen is aiming for – to get public feedback on the AI systems and see which behaviors will produce system issues.
The AI Text Kitchen is available on the iOS App Stores and Google Play.
