Midjourney has become even better at generating photos with its latest version, but the company seems to have more surprises. According to recent reports, you’ll also be able to generate videos through Midjourney in only a few months.
According to Decrypt, Midjourney will begin training its video model in January 2024. In other words –that’s only a few days from now. The final result will reportedly be released in a few months. According to this source, Midjourney CEO David Holz announced during an “Office Hour” Discord session on Tuesday. However, neither Holz nor the Midjourney team shared more information about the upcoming capability of the AI generator.
This is a logical step for Midjourney, considering that many platforms have already launched AI video capabilities. For example, Stable Diffusion is working on an AI video generator, and Adobe has introduced generative AI to video with Premiere Pro and After Effects. AI-generated videos have already been largely (mis)used for different purposes – from animating memes to spreading fake news and using celebrities’ likenesses without their permission.
I’ll remind you that Midjourney recently came under fire after a massive list of artist names used to train its image generator was revealed. This latest development in the ongoing debate about AI art creation comes after artists sued Midjourney and Stable Diffusion in October 2023. It adds fuel to the fire with specific examples like photography giants Ansel Adams and Henri Cartier-Bresson potentially seeing their styles copied. While some people argue that artistic inspiration and borrowing is intrinsic to cultural progress, the legality and ethics of AI training on copyrighted work remain hotly contested.
[via YM Cinema]
