AI-generated images appear to be taking over the internet lately and are really causing a stir and debate among photographers, artists, and creative communities. Alarm bells are ringing as people worry about job security, copyright, deepfakes, safeguarding of the internet, and even a potential robot uprising – yes, really.
Should we be worried? Google doesn't think so. Its latest offering to the field of AI image generators takes the form of Imagen, which we saw several months ago in the research stages. The service is almost ready for unlimited public use, available currently as a beta model through the AI Test Kitchen app.
Imagen is a powerful AI text-to-image art generator created by Google, first announced back in May, that can turn written text prompts into very realistic and convincing images. Though, Imagen had not previously been available to the public for fears of potential misuse and data-coded biases that exist in the software.
Imagen is now entering the next stage of its testing, through what Google calls its AI Test Kitchen, an app launched earlier this year that is designed to allow a select few users to try out Google's AI offerings and provide valuable feedback. The app can be downloaded now but you won't be able to log in until Google has emailed you, confirming access after users register through the waitlist.
As The Verge reports, the AI Test Kitchen app will offer Imagen requests and fun features rolled out as part of a season two-style update to the app, with two new functions called "City Dreamer" and "Wobble". These features will allow testers to interact with Google’s text model, LaMDA, and create a city made from virtually anything you choose, but with Wobble, users can design a little monster!
If by some miracle you've avoided the chaos and aren't familiar with the booming trend of AI-generated images, then you have some catching up to do. The most popular and recent AI image software widely available to the public is the DALL·E 2 developed by Open AI, though confusingly there's another software known as the Dall-E Mini (Now renamed as Craiyon) that isn't associated with the DALL•E.
There are also AI generators such as Midjourney which works via the social media platform Discord as a chatbot, and also Stable Diffusion which runs directly from your PC via a graphics card - and soon as a photoshop plug-in, and then there's DreamStudio which is in currently in a public Beta testing mode.
We haven't yet been granted access to try out these new Imagen AI features from Google, so we can't comment on exactly how they work, but the idea of designing a cute little wobbly monster through our own idea prompts has us sold already.
Google isn't stupid, though, and it knows that people will undoubtedly attempt to break or misuse the software. As Josh Woodward, senior director of product management at Google explained to The Verge.
“Places mean different things to different people at different times in histories, so we’ve seen some quite creative ways that people have tried to put a certain place into the system and see what it generates,” he shared.
While it might not be as open-sourced and able to generate wildly free-for-all content in the same ways as Stable Diffusion or the Craiyon, there's no doubt that this Imagen AI generator will be capable of great things while heavily fine-tuned and controlled to be more PG-rated.
Google has previously expressed how it hopes that AI development will eventually be able to address issues such as the ongoing climate crisis and even help fight against certain cancers. There's so much that we don't yet know or understand, but maybe AI isn't so scary after all? What do you think as a visual creative? Let us know!
You may also want to take a look at how a Photographer used AI to reimagine the appearances of dead famous people, as well as learn why Getty and Shutterstock have banned AI-generated images.