Now that Blender will have an AI rendering function (and also a film that just won an Oscar), I expect many other 3d packages to follow. I just completed an award winning doc feature using extensive AI, and it’s clear to me that this is going revolutionize animation.
I would bet money that Reallusion will introduce an AI render of some sort in v9, as they always seem to be in step with the times when it comes to 3d animation. If so, just take my money now.
If I could create a rough scene in Iclone, generate consistent characters and camera angles with the quality and speed of AI, I think it would be a game changer.
I am eagerly waiting for AI to be able to re-style animation frames with perfect fidelity.
The issue of needing a “high quality” render engine will be moot.
Just render your animation with the default.
and have the AI convert the frames to the style of your choice:
Pixar, unreal engine, Anime, 90’s cartoon etc.
I don’t want to have to buy Iclone 9
I would like this as a third party service where I could just buy “conversion sessions” for a reasonable price the same way people use render farms today.
or buy a local install and run to conversions myself.
How about an AI that fixes RL bugs ? I am not kidding, with Grok (Twitter) i solved quite some very interesting technical issues in the past. An AI like Grok with knowledge of the RL code would be able to fix Iclone ect. if only a user describes a bug close enough. Iclone 8, not 9 or 10 !
An AI like Grok with knowledge of the RL code would be able to fix Iclone ect. if only a user describes a bug close enough. Iclone 8, not 9 or 10 !
Also I am not a programmer but I do not think the end user can fix bugs with AI and have the fixes implemented without recompiling a new build of the software.
Speaking of AI
Cascadeur’s AI will Fill in the animation of your character based on where you moved him, totally avoiding any “Root motion” issues
Nono, the idea was to implement an AI (like Grok or Chtgpt etc etc, not using Grok itself) into Iclone or a Subapp that comes delivered with iclone for example. I just liked the idea of it.
Hmm, a full sized LLM like Grok or ChatGPT probably needs about 500-1000 GB VRAM. So that wouldn’t be possible without another subscription to an online service. I’m not really interested in that.
So that wouldn’t be possible without another subscription to an online service. I’m not really interested in that.