Thoughts on creating with AI so far

My primary question coming into this class was whether I could create animations in my established style, but faster. That's the whole ballgame for me - can I do what I have been doing the past several years, getting the same results, but faster? Thus far the answer has been no.

This is not a condemnation nor rejection of AI production, it's just been a shift in my thinking and my hopes for the class. Based on the results I'm getting thus far - I have not been able to replicate the established style of my animations using these new tools. I can't even conceive of how to do it, which has been frustrating for me. I think, in order to get what I'm asking for, a project itself needs to be conceived based on the strengths and weaknesses of the AI tools. I have to write a script/series specifically for AI production, not trying to imitate what I've done previously.

Again - not a condemnation - I want to be very clear about this. My current style is based on the strengths and weaknesses of Cartoon Animator 5. I conceive and plan my animations based on what I know I can (and cannot) do with that program, and as a result, I am able to produce a new animated music video or story in two weeks. That's the minimal pace I need to maintain to keep my channel growing and producing revenue. The scope of a production is dictated by the schedule and the tools.

For example, when my characters need to overcome an obstacle, having them hop over a lava flow from rock to rock is quicker to animate in Cartoon Animator than having them swing overhead on vines - so that's how that creative choice is made. Boom - lava flow. That's how many of my choices are made. I can see it in my mind and I know how to quickly create it in Cartoon Animator.

But with AI, it's very hit or miss. The shots may be stunningly detailed, but each shot is a piece of a puzzle that must fit together. So many details (look, action, timing, etc.) have to match from shot to shot, and I'm finding that it takes much longer to do these things in AI, if it's even possible. In short, speed of production is a higher priority than Pixar-level animations for my channel.

How I think I can begin to apply the skills we're learning in this class is to develop ideas specifically for AI production rather than trying to imitate my established style. I feel like it has to be a separate thing, developed natively for AI with that look and feel and AI vibe baked in. Yesterday I went back and looked at a couple AI generated videos that inspired me to join this class - the infamous McDonald's Christmas ad, and a couple of demo trailers created by Garry Pye.

In all of those examples, you can see that they were created based on the strengths and weaknesses of AI. The McDonald's ad is made up of clips that are self-contained scenes. There is no cutting from shot to shot within each "scene". Each clip is a separate location with different people doing different things.

Garry Pye does manage to do a bit of "real time" cutting from shot to shot within a scene, but you see that it was created knowing what AI can handle. Garry's trailers look AMAZING to me, and that's the level of work that I'm aspiring to through this class. I'm hoping with patience and persistence I can develop that level of skill.

5 comments