Midjourney
California
United States
Overview
With all the recent excitement about text-generating AI like ChatGPT, it’s easy to momentarily forget another advance in the world of generative artificial intelligence in the form of machine-created artwork. Like the large language models that generate text, these transformer-based generative tools let users enter a written prompt (think “Elon Musk and Mark Zuckerberg do bare-chested battle in a cage, rendered in the pop art style of Roy Lichtenstein”) and allow the AI to take a crack at creating a finished product.
Various startups and tools have already popped up in the generative art space, with DALL-E and Stable Diffusion being two notable examples. However, AI research lab Midjourney is helping to lead the pack with its impressively detailed and realistic image-generation platform, initially entering open beta back in July 2022. Since then, Midjourney has continued to iterate and develop its offering, releasing frequent version updates to optimize its algorithmic creation engine.
Any discussion of artwork quality necessarily enters subjective territory. But the fact that an image generated using Midjourney recently won first place at the Colorado State Fair’s fine art competition, generating reams of publicity in the process, suggests that some creative Rubicon may have been crossed.
Unsurprisingly, AI-generated artwork has proven a thorny topic. From broader concerns that AI will replace human artists to more focused complaints like copyright infringement, due to the scraping of web images for use as AI training data, Midjourney and its compatriots aren’t—to put it mildly—always viewed benevolently by the artistic community. Whether this could potentially disrupt its business model remains to be seen. On the flip side of the coin, high-quality generative AI makes it easier for anyone to create bespoke images and can offer a creative aid by assisting with the brainstorming of ideas. The art of “prompt engineering” (the writing of a well-crafted description of an image to create unique, exciting images) is certainly an intriguing new development; calling for human creativity to help clarify and explain exactly the kind of image the AI should conjure up.
FIND OUT MORE
Various startups and tools have already popped up in the generative art space, with DALL-E and Stable Diffusion being two notable examples. However, AI research lab Midjourney is helping to lead the pack with its impressively detailed and realistic image-generation platform, initially entering open beta back in July 2022. Since then, Midjourney has continued to iterate and develop its offering, releasing frequent version updates to optimize its algorithmic creation engine.
Any discussion of artwork quality necessarily enters subjective territory. But the fact that an image generated using Midjourney recently won first place at the Colorado State Fair’s fine art competition, generating reams of publicity in the process, suggests that some creative Rubicon may have been crossed.
Unsurprisingly, AI-generated artwork has proven a thorny topic. From broader concerns that AI will replace human artists to more focused complaints like copyright infringement, due to the scraping of web images for use as AI training data, Midjourney and its compatriots aren’t—to put it mildly—always viewed benevolently by the artistic community. Whether this could potentially disrupt its business model remains to be seen. On the flip side of the coin, high-quality generative AI makes it easier for anyone to create bespoke images and can offer a creative aid by assisting with the brainstorming of ideas. The art of “prompt engineering” (the writing of a well-crafted description of an image to create unique, exciting images) is certainly an intriguing new development; calling for human creativity to help clarify and explain exactly the kind of image the AI should conjure up.
