Midjourney Monday Machine Learning, June 16th 2025
goodMorning ➝
░▒▓ ▓▒░ ░▒▓
THIS Midjourney Monday Machine Learning
Midjourney video generations ➝ first looks! hotly anticipated, Midjourney’s own video model is beginning to bring images to life on the platform. we’re now in a phase of seeing (&& rating) these early glimpses of what will be. we’re helping refine what this nonhuman system becomes
what sets this new feature apart from other generative video platforms is how Midjourney leans into being excellent at pushing the ‘weird’ && wonderful aspects of Generative AI Art as Art. Midjourney’s /imagine creativity has always intentionally brought artistic sensibilities to AI image-generation. while giving us groundBreaking flexibility, Midjourney is committed to this core competency, this ability to enable Digital Art with its special idiosyncratic ideas && aesthetics. the Midjourney AI is famously a poststyle metamachine, a ‘vibe’-image-making moodEngine, a painterly probability space where we can get lost in liminal corners && creations
those of us who love Midjourney love its haunted artifacts && Glitch Art curiosities so much that now the company is concerned we are ranking these kinds of outputs too positively! we have already started to push the system deeper into that frontier… as they say:
“The video rankings data we’ve been getting has been a bit weird. Some people clearly like artifacts too much, so using the data is sometimes increasing the error rates in our video model.”
which is a moment for the unintended consequences that constitute Glitch Art, the Art of Surprise. all those wandering glitches, ghostcodes, and dustlore getting inscribed in the system itself!
░▒▓ ▓▒░ ░▒▓
here is one of my favorite outputs from the testing so far, which illustrates what i mean about Midjourney’s AI ML as it finds its voice in video — jonCates, Midjourney Monday Machine Learning, June 16th 2025
an open educational resource from the
《 glitch.school 》of my Glitch Art Gallery
in 台北，台灣 (Taipei, Taiwan) && online
