Wednesday, 31 December 2025

AI VIDEOS -- 2025


AI VIDEOS 2025

Perhaps the biggest overall Artificial Intelligence push in art during 2025 was in Developing AI VIDEOS

I started the year by generating short AI Videos in HAIPER, until they shut down on February 1, 2025 (though I managed to generate some free videos with their next incarnation; ultimately they became Video Studio and/or Imagine.ART).  I also used PIKAVIDU, VIGGLE, HEDRA, LUMA DREAM MACHINE and WAN 2.1 (both in a Docker Container and in KREA).


MIDJOURNEY VIDEO PLAYLIST



I animated my raw line drawings with Midjourney:


  • I also made animated GIFs of the animated figures in the above video

VEO2 and MIDJOURNEY

I remixed a video I made with VEO2 -- AI ROPE DESERT -- with Midjourney to make a longer animation that was more coherent -- Midjourney Continuity:

As Midjourney allows the user to make an AI Video with a starting frame and and ending frame, I exploited that to make a longer AI Video with more continuity.  I simply used the ending frame of the first video segment as the starting frame of the second segment...and a continued that approach

WAN 2.1
DOCKER CONTAINER
 
Holly sent us a Docker Container with WAN 2.1 (28 Gigs), and we "pulled" it and generated AI Videos with it at Quelab with a 10 Gig Video Card:
Not that the videos were great....


MAXIUM AI FIREPOWER

The Most AI firepower generated the worst AI Video of 2025 --  I even heard it from the comments section on YouTube -- "Why..?" and "Booooo":

No comments:

Post a Comment

Note: only a member of this blog may post a comment.