In the realm of 2D animation, there is a need for an AI tool that can assist animators by automatically generating in-between frames and coloring the frames for them.
The futur of animation is ANIMATORS and people, not AI. We need to stop to put ai everywhere or things we can do by ourself, you want animate? Learn to do it par yourself not using AI or waiting to Ai to animate for you. This is insane.
Thanks for commenting. As mentioned throughout the video, I see AI being an assistant in a lot of cases. But it does already enable individuals keen to tell their stories (or create fun clips or low cost marketing efforts) without the time, budget or inclination to learn complex animation techniques to be able to generate content quickly. I think AI tools (for professional work) like all tech advancements will be used to assist an animator, not replace them, (though AI video generation i.e. Pika, Runway muddy the waters here). The skill sets that clients/studios look for in vast number of cases will evolve. Ultimately a client/audience may not care if we find the process fulfilling as long as they get what they want. In a similar way that digital photography has enabled efficiencies for stop motion, digital software improvements sped up motion graphic and animation design, GPU and software improvements have improved 3D rendering quality and speed for decades. Green screen, motion capture, digital media, broadband, easy stock media.. etc.. it's always been changing. But 100% the artistry of giving exact emotion and control to a scene, object or character regardless of how it's done will take skill. The joy of bringing an image to life through your own time and effort will still have a value. We recently produced a hand drawn animation for a client at our studio, it took over 6 months, it looks good and we enjoyed the process. But hell. I dunno... I'm just trying to figure this out myself too.
@@zemptai theres a guy on RU-vid named Dong Chang and he's currently recruiting and working on a program with his private community that does inbetweens.
It would be nice to have an AI that knows the ins and outs of animation programs. Like where you create a 2d character with individual limbs from a few perspectives, and then you just tell the AI to animate it a certain way and it does it based on the model and how that program animates things.
Yeah something like that would be great, where AI is assisting in the animation of a character you've setup. You can do this with 3D characters via a few steps (retargeting bones etc) but this with 2D character rig animation will be great.... Hmmmm.... that idea gets the cogs turning for some future ideas around my AE plugin.
Thank you Jon for your continued support! Exciting news - Mootion is gearing up to deliver even more astonishing features and improved services in 2024. Your ongoing support means the world to us! Happy New Year!🥂
Really good summary of the industry so far. I see the various tools from the various industries being combined with the use of AI in the next year. Gaming and AR/VR will be optimized in the next few year to the point where we will be blown away. New careers will be created in the entertainment and visualization industries.
Thanks Jon. A really excellent round-up and so good to find someone talking in detail about how professional roles will be affected by AI - without repeating the ‘it’s only a tool’ cliches! 🥱 What are your thoughts on the recently announced Getty / Runway partnership, offering media companies and studios the chance to train AI models on their own IP. I think that could really shake things up, especially in animation when studios can create content featuring their own characters using text prompts, in the knowledge that the models are legally safe and within their sole control. Copyright law will also need to change though before this can be fully adopted. There’s a lot of legal teams holding things back right now until the law catches up, not to mention the likely backlash from audiences when AI is seen as replacing jobs.
Thank you for a wonderful recap and future looking summary. I'm personally most excited about your own AE plugin. That looks totally awesome! Congrats. I look forward to hearing more about it and purchasing it. Merry Christmas and Happy New Year!
Great video! I haven't done much with the animation, video and 3D work yet but I'm pretty deep in image generation. The problem with AI (I think) is that although it can generate amazing visuals, there's practically zero control over layout, subject matter and such. And as we know, clients love making changes and I have found so far that I had to keep answering back with "sorry, AI won't let me do that". One time I spent 2 days in Midjourney trying to get the right image as a reference for a photoshoot and ended up roughing it in 3D which only took me a couple of hours. I just don't see how turning the creative process over to a computer can work, but that's what clients want for the sake of saving money. They may as well say "please write a prompt that will generate money in my bank account"!
Yeah definitely agree. Consistency and control seem like the next big hurdles to overcome to make image gen ALOT more useful for client work. Things like in painting in MidJourney, photoshop and now Pika go some way to helping here. The release of controlnets (and similar things), give some other good ways for a bit of extra control too.
I think script writers and creative storytellers are the safest of all, at this point in time. They also hold the largest margin of advantage moving forward as their creative words become both images and audio that pass the "eyes and ears" tests... there's no substitute for the "human logic" test that comes through the prose and how it is presented. This is an abstract concept, not easily picked up by all people or more difficult for some than others.
Excellent breakdown. As an actor, 3d artist, filmmaker and audio editor, I am overwhelmed and super stoked learning and leveraging all these tools. I’m spoiled for choice. I’m not really scared of job security, I’m just trying all of this stuff and seeing what sticks. I’m tail end GenX so it’s all crazy, scary and cool. I definitely think the solo artist/produced content is going to be insane. And then, long term, the AI Agent solo work is the real game changer.
Hey AIamination! I'm super thrilled for Artificial Intelligence publicly realising out next year by 2024. Yeah it may be lots by 2030, but at least next year because i can't stand a second & life will be easier. I've been learning AI is fortunately realising; hinting that A.I cover is pubicly famous & overall the internet. I'M HYPED FOR ROBOTS! 🤖
Good question. I guess it depends what areas you're most interested in as well as where the demand will be. But right now I think keeping an eye on the emerging tools and understanding where you can utilise AI is big plus when approaching employers. Being familiar with existing and upcoming AI tools for rigging, motion-capture ( DeepMotion, Plex, Mootion, Move.ai and more) . Improving animation with things like Cacadeur, plus check out what's happening with basic 3D model generation tools with things like CSM and Meshy.ai. Plus texturing tools for Meshy and Adobe (substance). Plus advances in character creation (i.e. Meta human) plus other emerging tools around lipsyncing (both 3D, and applied to existing video content with things like SyncLabs).
Can you make a list in order from best to worst for moonvalley, runway ml gv2, pika labs 1.0, Google text to video. And I don’t need a full video. A comment really is fine because I’m trying to use image prompts
Am I the only one that looks at AAA media created in the last decade and asks myself, "Is human control really desirable, if the average quality of AI content exceeds the average quality of human-directed content ?"
I’m struggling at getting a consistent character for each image prompt… the other steps are possible … watched many videos but no luck … any tips or tricks¿
I'm just a keen part time amateur with all this AI stuff, but it has made it easier for people like myself to get their ideas out of their heads, I remember trying animation back in the days of adobe flash and failing miserably, ha ha, it was just too hard to learn without spending tonnes of time on it, so I have massive respect for you guys who do it for a living as I don't believe it's something any tom dick or harry can do, you've gotta put the effort & time into it, like anything else I suppose? Cheers for the tips dude, although I'm thinking back to all my favorite movies and we've already been warned about AI, so we kind of know where it ends, I just hope John Conner is still around to save us in time ha ha
i really think that so many people are delusional if they think that they can build a career out of ai animation, or animation overall. If it's gonna be available for everyone, it means that everyone can do it, which means there would be overload of animation on the internet to the point that nobody would be interested in watching your content if they can make it on their own i think it's gonna deminish views drasticly. also if there's gonna be a point in the timeline (i think soon) where animation would be made without human imput, kinda like a factory of animaiton popping up every few seconds. it's gonna be wild. i guess people will stop admireing other's work. I think most people's AI animations would be watch by themselfs only. Also im worryint that unlimited animation content or not even animation but overall AI movies might be another tiktok moment and people might get addicted to constant new things, until we become wall-e human babies, cuz also robots are soon taking over the physical works: like cooking, cleaning is already there, building houses, amazon bots, so much more. I also see that some of the people say that future might give us opportunity to work for free just because many people like challenges and not everyone is stuck in a chair type of person, i think it also won't work since if robot work gets more afficient by the time goes on in every department, slow, weak low attention spam typical human of the future would be to slow to adjust themself for the robot work. it's gonna be a future where maybe they'll stop selling or giving away art accessories, papers etc. because it would be just minority using these where majority would watch movies, live in the simulation forever until they die.
Considering how things have changed in 2023, I can safely say that if this continues to be the case, by May 2024 we will definitely be using excellent models for AnimatheDiff. On the newer models with a 10GB video card, you should be able to animate 1-2 minutes of video in 1 hour with no problem. And the videos will be much more stable than this video.: ru-vid.comPxcp3yzRGS4 Happy coming-crazy new year 2024 everyone!
They’re far from perfect. But both have got so much better compared to the results when first launched… but 100% getting something useful for your production still takes trial and error. The addition of camera controls in both, plus in painting in pika, and the motion brush in runway ML have been great steps forward.