Believe it or not: This image was generated by an AI program using three text prompts
The latest version of the artificial intelligence (AI)- powered image generator, Midjourney can create highly realistic photos that could compete with some of the top professionals in the field. This is a massive improvement over the previous version which was still a little rough and messed up the number of fingers on a human hand.
Created by a San Francisco-based company that goes by the same name, Midjourney, is one of the few companies that can deliver some amazing artwork with just text prompts. Launched only last year, the service has already seen multiple version releases in the past eight months and the latest version tops everything the AI has done so far.
AI-generated photographs are more real than before
A beta version of the Midjourney v5 was released last week for its paying customers and has found quite a few admirers. A designer who goes by the name Nick St. Pierre on Twitter has shared some images on the social media platform, comparing the differences in outputs between v5 and v4, when prompted with the same text
1990s era, photo of a young man, sitting, mechanical keyboard, working, desk, 1990s CRT monitor, monstera, 1 glass of oj, juice, wide shot, side view, natural lighting, home, polaroid, 4k --ar 16:9
— Nick St. Pierre (@nickfloats) March 15, 2023
v4 (left) v5 (right) pic.twitter.com/iMvG5z81mU
It is not just humans that the AI can realistically recreate in its images, even images of food like doughnuts and french fries, appear so tempting that one would want to grab a bite straight away, as was seen in images shared on PetaPixel.
Shared by graphic designer Julie Wieland, the images leave the viewer spellbound and it is hard to believe that they have been generated by a computer program that is barely a few months old and hasn't seen life in its truest colors, the way we have.
Naturally, the next question that arises after seeing AI at work, is whether this spells the end of photography as a career. Wieland disagrees. While calling Midjourney's v5 "fascinating" and "unparalleled", Wieland says that this is the beginning of "synthography" and adds a new tool for artists and photographers
Founder of Midjourney, David Holz has previously said that artists are the intended customers for the product as they can use it to rapidly prototype concepts before they even begin work.
While companies and businesses have rushed to embrace AI-based tools like DALL-E and Stable Diffusion, Midjourney's new version also demands that users write longer and more explicit prompts to get the best out of the AI.
Guess that's the new avenue opening up for modern-day artists who get the AI to do better work.