Open AI announced last week that the popular image generator, previously available to only a select few from its waitlist, is open to all.
To put this service to the test, the Deseret News sourced prompts from its writers to find out how much the image generator understood about life in the downtown newsroom.
The new Utah state flag selection process is underway, and I was curious to see what the algorithm would come up with. Here are a few AI contenders, for review by the honorable judging panel.
Clearly, the program still has issues with faces, hands and text, but it is just a matter of time before these issues are smoothed over. Additionally, race and gender biases have been seen in previous versions, and the algorithms have gone through modifications to address this, per New Scientist. It could be DALL-E or another image generator that finds the fix to these and other issues.
Open AI’s expanded service has also been improved to let users “continue an image beyond its original borders and create bigger images of any size, and collections,” according to the site. This allows creative users to produce “infinite zoom” videos and animations that would have been unimaginable a year ago.
My submission: “gm”
— cryptoshaynes.eth (@shaynes44580421) August 10, 2022
Created with 10+ dalle stills edited with inpainting.. then infinite zoom animation effect with after effects
Enjoy!!
Bar is raised . pic.twitter.com/f5Ee0OCOl3
Over the course of a few months, these AI-generated images have seen an exponential rise in quality and popularity. It is still unclear where this technology will lead, but its potential is undeniable.