Something that doesn't get covered very often is that many clients can not legally accept any deliverables with AI tool usage - including all the way down to the Photoshop Generative Fill. Many broadcast clients require you to sign a "no AI" waiver that puts any legal obligations onto you as the studio/artist.
Yup. But I wonder if it'll end up being kind of like "don't use copyrighted images in the boards." Lots of designers still do it in the shadows, but they're messing with the imagery so much nobody will ever be able to tell where that texture came from, or that little snippet of an image... you could almost claim fair-use but maybe not... if you're using Generative Fill on a 100x100 pixel patch to clean something up on a photograph, is the client gonna know? Does that count? The end of the year pod this year is gonna be fun 🙌
@@joeykorenman Agreed but I don't want to be the owner left holding the bag if someone does an audit and there's a GenFill Smart Object sitting in one of my artists PSD files 👀 I guess that's why we all have production insurance (we all have production insurance, right - RIGHT? 🤣)
Out of interest, are they specific that you can't use generative AI as opposed to a blanket term "AI", as surely the rotobrush on AE would then fall under blanket "AI", likewise, how is upscaling assets perceived?
this industry overhead talk is SUPER helpful, school of motion should do more content like this. it's informative but also this is the type of stuff we've all had on our minds the entire year
This video is fascinating and the conversation is more than needed today. Especially because Joey is talking about the business side of the subject, which most artist tend to not dive deep into because it’s out of their field. I was listening from two perspectives- one of a 3D designer and one of a investor who holds stock in AI (I have a few dollars I’m not rich). So for both perspectives the information was really helpful. I can definitely say that I agree with the general statement that AI is not gonna take our jobs, it’s here to help us. It’s like being an accountant when Excel came out, we still have that job, but you don’t need someone making all of math by hand on a calculator. There are a lot of news that some companies are firing people because they can replace them with ai, which sounds terrible and scary, but if you look in more details, that’s not the case most of the time. The company was most probably gonna cut staff because of financial problems so they just put a label AI on that. Everyone, don’t be scared. Experiment with the tools, they are wanky as hell most of the time, but in some cases they help. They are here to stay.
It's funny that these technologies have disrupted the way a lot people were sustainably making money, leveraging the fruits of those people's labour to create models, just to get in the position where they can lose money on a colossal scale and turn around and tell those people the way *they* are doing things is now unsustainable. It's easy to disrupt when your undercutting of an industry is funded by billions in venture capital and the raw materials for your projects are "free". I say all this as someone deeply engaged in this tech, it's a bit of a messed up situation. Overall, this was a great overview of everything going on in this space. I just wish ethics was left as a footnote. Ethics isn't just doing good for the hell of it. Ethics can inform responsibility and be pragmatic in its own right. It's not competing with pragmatism or practicality.
really refreshing to hear someone finally talk about the actual objective "usefullness" or lack thereof with ai tools. everyone who talks about it is way distracted by the “impressiveness” of the tech and promise of what it could turn into and needs to take off the hype glasses and actually evaluate how much money it’s actually worth to use these things, because for me it’s incredibly low if no zero 99% of the time. I’d like to see some people actually proposing how to actually use this stuff in genuinely useful ways but I rarely see it in ways that aren’t also super niche. completely disagree about the vibe shift, if anything people were more curious about midjourney dalle etc in the beginning and now it seems almost universally hated among artists. ive also gotten way more sick of machine learning tools and the hype overshadowing everything else, making me a lot less interested in exploring them as well. the hype needs to die. i wish we could go back to seeing more excitement and focus on non-machine learning tools that can dramatically improve peoples productivity, updates to existing software, not everything new needs to be AI. feels like a huge distraction (because of the hype). i also miss ai tools that were being explored more before LLMs and diffusion models took off. reinforcement learning for rigs was a really interesting rabbit hole and i'd love to be able to train characters to act and explore in 3d scenes i build, but its been totally overshadowed by chatgpt etc.
Not sure how you managed to discuss AI without mentioning how 'tHiS cHaNgEs eVeRyThInG!!!!' , but you did, and it was well thought out and brilliantly executed.
I was working contract at a company that claimed they had an AI that could generate production ready assets. In reality the clients would upload they're requests and we would work on the project and submit it to them. The only thing that was actually AI was the messaging system 😆
this dude giving us some straight daddy knowledge from the 90s! true say man we got way too much AI marketing hype right now it's gonna burst, open AI C suite's are quitting cuz the companies aren't profitable enough
@@joeykorenman Huh, that’s a good point. I’ve seen the animation industry going to shit since way before the AI hype, post-Covid recession was a bigger culprit.
@@joeykorenman There are SO MANY factors that are in play before we ever get to the AI boogeyman: • End of ZIRP era • Near miss banking crisis • Tech companies getting lean (thanks Elon) • Studios are slowing down episode orders and renewals • Wars, pending elections, political uncertainty, and other global crisis • Publicly traded companies no longer focused on growth but profitability • Absolutely no one knows how to market effectively to new generations of audiences and consumers
Anyone who can "set aside" the legal and ethical concerns is also someone we can "set aside" from working with, supporting or suggesting as a resource. Anyone who still doesn't get it, means they've made their choice and shown what type of person they are. You don't always need to yell at someone, sometimes you just need to ignore them going forward... Like I'm about to do right now.
I agree with this. Joey is advocate of AI because it isn't affecting animation directly, but what about the person who used to make those assets for you? You wont need them anymore. Its all good until it affects you.
Thank you for all of this thinking. Such a benefit to have your take about this subject combining your motion design experience, your industry's knowledge and with the last recent info, while still being cautious about the speculating. Incredible video !
One thing people miss, is that for the last year they specifically have stated multiple times that they won't release the next gen version of GPT until AFTER the election, which means they are holding back and even being transparent about it. I think all this AI winter stuff is overlooking these statements, and disregarding the fact they told us December 2024/January 2025 we will start to see the next iteration of this technology.
The fact that I tried to use AI design once but in the end I decide to ask a real graphic designer because I realize that AI is limited even if AI creates something amazing but human graphic design is the best and amazing as always. You also gave me a reminder that AI can't create without us creating.
First of all, thank you for addressing this important and timely topic that impacts not just creatives and animators but the broader population as well. I completely agree with your analysis of where these AI models currently stand. Personally, I believe the true craftsmanship of creating comes from being hands-on and deeply connected with each frame of your work. These tools, while powerful, don't yet have that essence, and perhaps they shouldn't aim to. As an After Effects user, I’d love to see Adobe develop something akin to Firefly in Photoshop, but for AE. AI could assist with repetitive tasks, optimize workflows with expressions, and even remember your unique working style, becoming a customized collaborator. The idea of AI as a 24/7 assistant, handling tasks like rendering or repetitive actions while you're focused elsewhere, is where I think the technology should be headed. It would also be helpful to have multiple outputs within AE to avoid rendering multiple versions of a comp separately. That said, I feel Adobe jumped on the AI hype too quickly. They need to focus more on the artists who use the software daily and think creatively, rather than simply trying to match competitors like Runway or Midjourney. Firefly is a good start, but Adobe could be more thoughtful about how these tools genuinely support us artists. After Effects is incredible, but it could benefit from deeper involvement and development with its core user base. I hope companies will steer AI development towards truly supporting artists rather than chasing fleeting trends. Just like with NFTs and Web3, we've seen how hype can create bubbles with limited real functionality.
Thanks, that's the most accurate explanation of what's happening with with AI stuff.
3 місяці тому
Currrently I am resistant to use assets created with AI from scratch, but I find AI tools very useful when you bring something to it as a starting point, let's say a image, drawing, design or footage of your own, then you use AI to edit this material like upcaling it, removing backgrounds, tracking objects, cleaning up the audio, interpolating frames, etc... there are tones of boring and technical tasks AI can help which I am comfortable with. But we need to be he masters of the creativity so doesn't matter how fast AI can help us create a new script or write ideas, we need to shake lazyness from ourselves and put our mind at work, imagine, reflect and debate on things, otherwise those things that are create can not be claim as your own creations but AI's. Thanks for the video man.
I'd totally agree on the fact that AI is a tool, and a as a tool I't would never be as creative as the person using it. Ai works very well, not perfect but it is still something you work with, as well as the other tools you use on your workflow, so I think we as designers have to use and put to the test all of these tools because it will also put us to test. I think that natural intelligence will expand over AI, it will make us more inventive and will make us push more and more barriers on creativity and what can be done will the tools at hand.
tbh as an artist, i love to create things. what i actually would want to spend less time on is not creating things, but all the business stuff. i would much rather spend my time making assets than talking to clients, creating and sending invoices, filing taxes, etc.
Problem with technology change is we don’t get to choose what these horrible things do to us. Look at social media and the massive impact it’s had on us.
Nice talk, appreciate the insights. I would only add that it’s probably worth looking at what happened to commercial music houses over the last 15 years if you want to look at the future of motion design + VFX.
I watched the entire video and found it really interesting. Would be nice to have your opinion in the same format on this subject every 6months or every year. Just a question though ... I don't know if it was a bug or if my previous comment was manually removed ... If it was, could I know why ?
@@joeykorenman Ok I'm relieved because my comment was challenging your analysis but in a respectful way so I felt "Holly cow, no contradiction is allowed in here 😅". Glad to know it wasn't intentional.
@@adelacelle I'm all about spirited debate! I'd never remove a comment unless it was spam or just mean. We're all figuring this out at the same time so respectful debate is the best way forward IMO.
Really interesting informed piece Joey, thanks for this. I have to say after a couple years of nightmare fuel of examples of AI it does feel like the gravy train and endless promises of linear development is coming to a crashing halt soon. Some interesting news out there regarding the fact that the training models themselves may literally be pretty much rung dry already, certainly nowhere near large enough to make the kinds of jumps and leaps we have seen early days in these various models. I hope it does settle into good useful time saving tools to avoid the nitty gritty admin stuff we have to do that disrupts the animation and design side of things. That would be cool. The glee at which certain X users had that artists careers were all over seems to have been, like most things on X, completely unfounded nonsense.
Just like broadband was extremely expensive in the late 90's and is much cheaper now, the cost of AI Compute will go down. We already see this happening rapidly.
To what ends? What we have is something that can generate low grade stuff. I think the biggest let down for ai is the shift in attitude that Joey describes. Once the novelty wares off the tools get really boring, really fast. For example, I’ve used ai for basic writing tasks. After a little while my own writing style for these basic tasks improved to the point where I don’t really benefit from the text generation tool. Is this revolution going to be a more efficient and slightly more accurate Grammarly?
I think this is the most level headed/measured response to all of what is happening in this space. I personally have started dabbling in AI for very specific workflow needs that I have. Like for instance a tool called " Depth scanner" on AE scripts works decently for making depth passes on footage. the Rotobrush tool in AE is better than it has ever been thanks to AI. Image/Video upscaling is useful for certain pieces of footage/effects that you are going for, but not everything. I still need to take the depth pass and use it as an asset to manipulate other pieces of footage. Image/video upscaling a lot of the time adds details that you dont want, and will need to be cleaned up in someway to make it usable. The rotobrush tool is very good, but you still have to wait for the results of the brush strokes, and even then it doesnt work on every single piece of footage, which will inevitably means still a bit of rotoscoping. I think these are all realistic uses of AI, in its current form and I personally am sorta ok with their uses in this limited context. When you mention the vibe shift that is happening I think is due to AI people realizing just how not very useful these tools will be and stopping the constant onslaught of "AI will replace animation" posts on every single social media website that exists. Lol for instance this person "robert" in the comments for this video proves my point! He is talking about AI tools replacing the mass majority of artists and that we should still be scared of it. Thats what pisses people off. Sure there was and still is backlash about the ethical stuff that is happening, but i think what really boils peoples blood is people like all of the AI influencers on linkedin who are talking about "oh hey sorry this job market is about to be crippled but thats the way of the world in terms of innovation so deal with it!" That boils our blood because the people who have money to make projects are listening! Investors listen to AI influencers over people who are actually in the trenches willing to work hard to stay in this indsutry, which in term destroys any sort of rational thinking and critical reasoning that can be discussed about AI. Its so damn frustrating when I've applied to over 400 jobs, AND have to deal with the noise that is trying to take over the space that I want to be apart of so badly. Its business, so I don't blame these people for following trends, but I've had enough of them tell me to go fuck off to another, that I am ready for this bubble to explode so they can find some humility. I hope OpenAI gets the axe and we have strict regulation around it so this can move forward into a more symbiotic relationship rather than there being constant antagonizers towards this field. I know its about ignoring and just moving forward but come on, we are all trying to make this work and the only way to move forward is to have civilized discussions like this one. So thank you again for having a level head regarding this subject!
I think you nailed it. Investors and companies are definitely swayed by AI influencers shouting "THIS CHANGES EVERYTHING," but if they listen to on-the-ground artists doing real-world stuff, the results (while impressive) are not nearly as great as hyped. Maybe in a few years that's different.
Useful AI Tools in my motion workflow are depth scanner, (sometimes I use a photoshop action to go through a PNG sequence and apply the photosnop depth neural filter though), Topaz Gigapixel & Video enhance AI (The best upscalers imo - also video enhance has a really good frame interpolation tool), Zoedepth, very useful when you need to quickly add animation to am image that's somewhat spatial, lalala ai for vocal/music separation, adobe podcast, ChatGPT/Dall-E for certain illustration generations, these always need tweaking but there's certain scenarios its useful. Then there's all the standard Adobe ones on top of Adobe podcast eg gen fill/rotobrush/select subject etc I've had other peers blanket refuse my offer to upscale a video because the program I'm using it called "video enhance AI", but will happily use the rotobrush, I think people shoving a blanket hatred to anything "AI" is stupid, although I can appreciate not being a fan of generative AI
Is there a good AI rotoscoping and green screen keying tool yet? Seems like this would be high priority for this industry. Eliminating the most mundane and tedious aspect of our jobs.
I've been saying this for a year at this point. It just doesn't have the capacity to replace an artist and likely never will. Especially without a precise level of control and workfiles.
And enough compute… nuclear fusion would be a prerequisite I think… if we don’t destroy the climate by burning fossil fuels to power these comparatively limited tools.
I feel the bottom line of AI is what we value as artists, as humans , and what clients value from us, Ai may be able to counterfeit something into existence, quickly, and it may even look great, but at what cost, we are the the artisans of our chosen disciplines , our skills should be valued, please dont accept that AI is inevitable in our field, for learning, for live language translation while traveling, personal productivity, all fine, but it has non place in the arts, were humans, and have lost value of living a human experience,
Ai will break away when someone makes an Ai that can use after effects. Not make a rendered video but make an After Effects program file where you can either render as is or make the tweaks you want.
This already exists at a functional level. It's called templates, and some have script-based interfaces. Problem is it's never exactly what you need for a particular client.
This is suuuuper interesting and you make great cases and I learnt a lot. However when you speak about what technological development means for the time you spend on a creative project I have to make my case. @joey You say that studios make more money because they can finish projects quicker and with more quality, right? But how capitalism works is that when one studio decides ‘wait a minute, we can lower our price for high end animation because we input less hours - and then all the business will be ours!’ The market is going to adapt and high quality content will simply become easier to make. So I believe most creators will in the future spend less time to get a project where it needs to be. Simple example: one of my client knows it costs way less time to put captions on a social video, so he’s going to ask for a lower price. Also one other small note: midjourney has a web based interface since a few months and it works great.
I don't think I said companies would make more money with AI (I'll have to re-listen), but that they'd be able to do the same level of work in less time, which tends to just mean that they spend the same amount of time as before but get better / more results. Roto is a great example... I remember having to budget in 2 weeks of roto for certain gigs, but now rotobrush / ML-assisted roto can do 80-90% of that, so in the end you spend more time finessing the comps / design vs. having to do that stuff faster to accommodate manual roto. Globalization has been driving down prices well before AI, and is probably a bigger influence there IMO. I didn't know MJ had released that interface to everyone yet! We didn't have it recently, but it looks like we do now.
The point everyone seems to be missing is that the AI companies stated goals are NOT to be profitable, but to drive market values down, influencing costs. The AI fantasy is that you will invest enough into an AI for the Ai to figure out how to solve unsolvable issues with super intelligence. It goes something like "right now it might seem like it takes an insane amount of energy to run a certain model, but past a certain threshold or even horizon, the AI will be so advanced it will figure out how to power itself efficiently with solutions we can't even think of now." And that's basically the logic behind the Sam Altman's of the world. If you ask me, it's all a Charade. They are doomsday preppers playing god with humanity at stake and they're addicted to the possibility of being the ones to pull the trigger.
I guess I've got a solution for Open AI. Why don't they use customers' GPUs for calculations? Like if you don't have a gpu you need to pay use sort of a render farm, but optionally, you can use your own gpu.
I've been dabbling with AI image generation with StableDiffusion since before version 1.5. Even though I'm no expert in market tendencies or in any of these big techs' vision, while browsing reddit and other dungeon domains of the internet, I still believe the majority of use cases of AI consist of pornography or deep fakes and, sometimes, coupling these two together...
Fantastic job with this video, Joey. I love that you started off with some historical context. And I appreciate your honesty around your speculations. For what it's worth, I am 100% aligned with your takes across the board.
So far still a lot of gimmicks I have tried incorporating it in my workflows but end up giving up. Dont think any of these tools are more than fun features. The technology is incredible but none really is using it to properly help creators in any meaningful way. Let’s hope that after the lawsuit is done, they can get their things together and actually use it in a more meaningful way
The clients will happily accept 70% Quality for 10% of the price, especially because viewing habits change rapidly. We already see a lot of diminishing quality in ads. Further, with AI it much easier to create target audience tailored content in parallel, which is highly in demand.
Great history lesson and commentary. I can see a lot of these technologies becoming things we run locally as GPUs get way better, and so the companies selling it as a service will die off. If Flux gets better and I can run it on my gaming rig, why keep paying my $30 to Midjourney. You're right about the shift in the conversation, but I think it's only happening in VFX and video. I was at a talk by someone high up at a VFX studio, and he was showing off all the tools they are using at the concept stage. And it was just Midjourney and Luma, so it's nice to see that established industry doesn't have anything much more advanced than the average consumer. The comics industry and a lot of artists are still at the frothing at the mouth stage. It maybe isn't quite so bad as two years ago when people were getting threatened with violence, but it's still pretty toxic. I turn to AI when I hit a roadbloack of my own skills. I'm bad at writing emails, so it's become my starting point for correspondence so that I don't get anxiety over putting off the emails. I use it to design characters so that I can redraw them and animate them. I use it for stock photos when I'm doing a boring corporate project that needs 100 specific images I'd take days to find on Envato and Pexels.
It’s a pleasure to listen to you talk about something. I’ve been listening to som podcasts since I’ve completed Ae kickstart back in 2018. Them, and now motion mondays are my source of life insights and news to this day. Thank you for your content!
Ai Generation is at a point where it can generate "A kid's 10th birthday with everyone having fun... Etc etc etc", but what it can't yet do is generate "My Son's 10th birthday as it happened on X date at X location" and I think that "personalization" is a main factor missing from Ai generated design and is where artists can really craft Ai into a tool to help unlock their ability to find solutions for clients.
Imma be honest, I click on these for the wealth of info because I enjoy your content, but also to see what other bands we have in common by way of your t-shirt 🤣
For 3D animation, ray tracing actually simulates how REAL light works. No video generative AI can replace it in this context. AI just morphs pixels, and I personally don’t think it sees that much depth in an image, if we explicitly not optimize it to do so. Actually I doubt that video AI could even get to the point where no morphing at all will be present in a shot. Even with the best AI it is like telling our brains to get a clear and realistic image while dreaming in a sleep, without all the nasty artifacts of our memory recall and imagination. But because morphing between concepts is the nature and the artifact of the system, I find it is really useful in abstract art ☝️ and from the beginning it was always like that, and I was fascinated to see the «out of this world » stuff that AI can spit on you with some good prompting.
❤🔥❤🔥❤ Lots of good stuff to chew on in that one. Important stuff to consider... career-wise I mean. I want to agree, but I also feel pretty amazed and scared by how fast it has improved over the last 18-24 months.
School of Motion Alum here from AE Kickstart. As someone who now works full time in AI (I am the Head of Content Marketing for one of the largest AI Art Resource sites on the web), I keep up with all of these videos and really wish that you guys would dive into whats happening in the open source community sometime. Especially on the ComfyUI side of things. I could show you some things in Comfy by utilizing the skills I acquired in After Effects that would probably give you a lot of new ideas about how the technology could be used. Anyways, great video, great level headed talk. I am still searching for the commercial viability and the path to longevity in all the madness. Onward.
This was one of your best commentary videos yet. You cut through the hype and covered all the problems, roadblocks, aspirations and realities of the situation. AI is now being taught as coursework in art schools, so it's not a question of if but when. As one of the professors at Ringling said to his students: "You won't get replaced by an AI. You'll get replaced by someone who knows how to use AI."
*Can't wait for the burst of Ai Bubble, so you can actually focus all the efforts and resources toward Practical use of AI* Once Ai Bubble is over, grifters will move away from Art, Engineering, Everyday of "dash dash so over cause of this new AI" etc.
Thanks for all these thoughts! This new technology is growing fast and much more faster than the crypto/nft bubble. We certainly integrate them shortly to our pipeline but right now we missed tools to dissect & animate them with the perfection a professional job requires. I was learning graphic design on analog tools back in the days… and I still use some in addition to digital and online new tools. Time flies !
Let's call it what it is: People in the web are blindly pushing and over-inflating E V E R Y new technology that comes out, they don't question it, just because they're afraid if they don't endorse it - they'll be irrelevant on their field. It's like UA-camrs who unbox a product and say it's awesome without testing it for a month at least. Same people who said "Threads" by Meta is the new best thing-are the ones pushing AI without understanding in what way can we (legally) utilize and make profit out of it.... 🤔
I agree with you, we can say we used to do a frame-by-frame on paper after in pc programs now on iPad, it's always will be technology development and we need to go with it because it will make it easier and much better at the end for us.
The stock photography going out of business thing is kinda dumb. Its' actually easier, faster, and about the same cost to use a stock photography site to search for an image than to generate one, and by definition the image will always look real. Also, many stock people sell images of real things and place, that must be real images. There would be no use, for example, for an AI image of Paris in a travel brochure. That would just be fraud.
Am I the only motion designer thinking that this debate is just capitalism selling us, one more time, that eeeeverything is great and not talking about the horrible world that this and other technologies are helping to create?
I mean, can we forget that OpenAi worked with Cambridge Analytics to create hate online and helping Trump? I cannot trust anything of this without feeling ethically a piece of sh*
2 thoughts: Things like the VCR and digital cameras struck fear in to the film and photography industries, but of course were widely adopted. One thing that sets these new generative AI tools apart is how quickly they just seemingly exploded out of nowhere. People had little fore-knowledge and suddenly had to face what looked like totally disruptive technologies. But as you say, now that its had some time to percolate in the collective consciousness we're a bit more level headed about them now and hopefully moving forward. Second - glad to hear you address one thing I've said from the beginning, they can't be accurately directed. If I'm just noodling around at home and make an epic alien space battle, yea that's fun - but when its for a client and they have specific notes on what hero moments need to be hit, framing, story beats, etc - its just not there.
The style transfer thing you're discussing around the 36 minute mark is already very close to being ready for prime time. Great plugins/addons for both 3dsmax and blender that are only gonna get better.
Joey, you've always positioned yourself as a motion person but you've got great analysis on the creative industry at large. I always recommended your manifesto book to many non-motion designer freelancers too.
What a great video, thanks a lot for making it! You shared a lot of good highlights, one thing that makes me curious and that no one is mentioning compared to the Web3 bubble is the energy consumption, the projection is crazy
I understand you sell courses of the tools will be useless with AI, but it is happening right now. It is not that good? Well people have a real low bar theses days.
first half is interesting. So they need to get more customers to be profitable or make a loss and then you say the vibe shifts. so you advise motion designers to get locked in. in the end when no more people are there to lock in, these firms will do some ugly monetization. when the designer unlearned the normal tools and is dependent on some company running a server with a model on it it will be ugly for the designers when that goes away or becomes really expensive (those stockholders want to see a return after all). Also LLAMA is not open source but with open weights. The training data is hidden. Probably still all copyrighted material in there... zuck calls it open source to please the EU AI regulators, but it totally is not! Interesting summary with some weak points. Artists, learn your tools and don't get dependent on generative models. I am totally not against some technology that can perfectly remove backgrounds in videos while preserving every moving hairstrand, but those models trained on copyrighted work are wrong to use!
Great analysis/explanation! Regarding the tide shifting, I still see a lot of blind AI hate pretty much everywhere, but I've seen a few people come around a little bit once they see a project that used AI as a tool and still had a lot of effort involved. But even then, Corridor Digital has gotten SOOOO much flack from attempting to use AI to make their Rock Paper Scissors anime.
This entire video feels like copium. So many of my friends are cover art designers, logo designers and 2d artists.... And they're all doing door dash now because the jobs simply disappeared.
it's funny that nobody cares about stock illustrators and photographers. Everybody just admit that they are doomed. Design community kinda rase the voice but for themselves, essentially.
If a company raises their prices by a factor of 10 and as a result loses 50% of their customers, they're now taking in 5 times as much money with half the overhead. Pretty significant bump in profit.
Something that doesn't get covered very often is that many clients can not legally accept any deliverables with AI tool usage - including all the way down to the Photoshop Generative Fill. Many broadcast clients require you to sign a "no AI" waiver that puts any legal obligations onto you as the studio/artist.
Yup. But I wonder if it'll end up being kind of like "don't use copyrighted images in the boards." Lots of designers still do it in the shadows, but they're messing with the imagery so much nobody will ever be able to tell where that texture came from, or that little snippet of an image... you could almost claim fair-use but maybe not... if you're using Generative Fill on a 100x100 pixel patch to clean something up on a photograph, is the client gonna know? Does that count?
The end of the year pod this year is gonna be fun 🙌
This is the reality. Always has even before ai. I still only use licensed textures etc at work just to be safe. @@joeykorenman
@@joeykorenman Agreed but I don't want to be the owner left holding the bag if someone does an audit and there's a GenFill Smart Object sitting in one of my artists PSD files 👀
I guess that's why we all have production insurance (we all have production insurance, right - RIGHT? 🤣)
@@Oddernod HA! Suuuuuuure we do.
Out of interest, are they specific that you can't use generative AI as opposed to a blanket term "AI", as surely the rotobrush on AE would then fall under blanket "AI", likewise, how is upscaling assets perceived?
this industry overhead talk is SUPER helpful, school of motion should do more content like this. it's informative but also this is the type of stuff we've all had on our minds the entire year
So much this.
This video is fascinating and the conversation is more than needed today. Especially because Joey is talking about the business side of the subject, which most artist tend to not dive deep into because it’s out of their field.
I was listening from two perspectives- one of a 3D designer and one of a investor who holds stock in AI (I have a few dollars I’m not rich). So for both perspectives the information was really helpful.
I can definitely say that I agree with the general statement that AI is not gonna take our jobs, it’s here to help us. It’s like being an accountant when Excel came out, we still have that job, but you don’t need someone making all of math by hand on a calculator.
There are a lot of news that some companies are firing people because they can replace them with ai, which sounds terrible and scary, but if you look in more details, that’s not the case most of the time. The company was most probably gonna cut staff because of financial problems so they just put a label AI on that.
Everyone, don’t be scared. Experiment with the tools, they are wanky as hell most of the time, but in some cases they help. They are here to stay.
It's funny that these technologies have disrupted the way a lot people were sustainably making money, leveraging the fruits of those people's labour to create models, just to get in the position where they can lose money on a colossal scale and turn around and tell those people the way *they* are doing things is now unsustainable. It's easy to disrupt when your undercutting of an industry is funded by billions in venture capital and the raw materials for your projects are "free".
I say all this as someone deeply engaged in this tech, it's a bit of a messed up situation.
Overall, this was a great overview of everything going on in this space. I just wish ethics was left as a footnote. Ethics isn't just doing good for the hell of it. Ethics can inform responsibility and be pragmatic in its own right. It's not competing with pragmatism or practicality.
really refreshing to hear someone finally talk about the actual objective "usefullness" or lack thereof with ai tools. everyone who talks about it is way distracted by the “impressiveness” of the tech and promise of what it could turn into and needs to take off the hype glasses and actually evaluate how much money it’s actually worth to use these things, because for me it’s incredibly low if no zero 99% of the time. I’d like to see some people actually proposing how to actually use this stuff in genuinely useful ways but I rarely see it in ways that aren’t also super niche.
completely disagree about the vibe shift, if anything people were more curious about midjourney dalle etc in the beginning and now it seems almost universally hated among artists. ive also gotten way more sick of machine learning tools and the hype overshadowing everything else, making me a lot less interested in exploring them as well. the hype needs to die. i wish we could go back to seeing more excitement and focus on non-machine learning tools that can dramatically improve peoples productivity, updates to existing software, not everything new needs to be AI. feels like a huge distraction (because of the hype). i also miss ai tools that were being explored more before LLMs and diffusion models took off. reinforcement learning for rigs was a really interesting rabbit hole and i'd love to be able to train characters to act and explore in 3d scenes i build, but its been totally overshadowed by chatgpt etc.
bro the second paragraph has my heart ❤❤❤
Not sure how you managed to discuss AI without mentioning how 'tHiS cHaNgEs eVeRyThInG!!!!' , but you did, and it was well thought out and brilliantly executed.
R.I.P. Video Editors!!!
Cuz he's a pro and can see the limitations
I was working contract at a company that claimed they had an AI that could generate production ready assets. In reality the clients would upload they're requests and we would work on the project and submit it to them. The only thing that was actually AI was the messaging system 😆
this dude giving us some straight daddy knowledge from the 90s! true say man we got way too much AI marketing hype right now it's gonna burst, open AI C suite's are quitting cuz the companies aren't profitable enough
I have more and more friends without work. At least in social media Ads, generative Ai is already affecting creatives.
Do you think AI is the primary culprit? My sense is that the economic slowdown is more responsible for this than AI (at this point).
@@joeykorenman Huh, that’s a good point. I’ve seen the animation industry going to shit since way before the AI hype, post-Covid recession was a bigger culprit.
@@joeykorenman There are SO MANY factors that are in play before we ever get to the AI boogeyman:
• End of ZIRP era
• Near miss banking crisis
• Tech companies getting lean (thanks Elon)
• Studios are slowing down episode orders and renewals
• Wars, pending elections, political uncertainty, and other global crisis
• Publicly traded companies no longer focused on growth but profitability
• Absolutely no one knows how to market effectively to new generations of audiences and consumers
@@Oddernod Agree 100%
@@joeykorenman I think so.
Anyone who can "set aside" the legal and ethical concerns is also someone we can "set aside" from working with, supporting or suggesting as a resource. Anyone who still doesn't get it, means they've made their choice and shown what type of person they are. You don't always need to yell at someone, sometimes you just need to ignore them going forward... Like I'm about to do right now.
At least in social media Ads, generative Ai is already affecting creatives.
And all those stock photos of cog wheels, illustration, music, rotoscoping and all that AI can do - it was someone else's work too
I agree with this. Joey is advocate of AI because it isn't affecting animation directly, but what about the person who used to make those assets for you? You wont need them anymore. Its all good until it affects you.
Thank you for all of this thinking. Such a benefit to have your take about this subject combining your motion design experience, your industry's knowledge and with the last recent info, while still being cautious about the speculating. Incredible video !
Joey, thanks for making this info so accessible to understand. You have your way to make it fool proof
Realistic take on AI. I do have almost the same thoughts about AI for designers.
One thing people miss, is that for the last year they specifically have stated multiple times that they won't release the next gen version of GPT until AFTER the election, which means they are holding back and even being transparent about it. I think all this AI winter stuff is overlooking these statements, and disregarding the fact they told us December 2024/January 2025 we will start to see the next iteration of this technology.
Damn Joey, that was a good breakdown.
The fact that I tried to use AI design once but in the end I decide to ask a real graphic designer because I realize that AI is limited even if AI creates something amazing but human graphic design is the best and amazing as always. You also gave me a reminder that AI can't create without us creating.
First of all, thank you for addressing this important and timely topic that impacts not just creatives and animators but the broader population as well.
I completely agree with your analysis of where these AI models currently stand. Personally, I believe the true craftsmanship of creating comes from being hands-on and deeply connected with each frame of your work. These tools, while powerful, don't yet have that essence, and perhaps they shouldn't aim to.
As an After Effects user, I’d love to see Adobe develop something akin to Firefly in Photoshop, but for AE. AI could assist with repetitive tasks, optimize workflows with expressions, and even remember your unique working style, becoming a customized collaborator. The idea of AI as a 24/7 assistant, handling tasks like rendering or repetitive actions while you're focused elsewhere, is where I think the technology should be headed. It would also be helpful to have multiple outputs within AE to avoid rendering multiple versions of a comp separately.
That said, I feel Adobe jumped on the AI hype too quickly. They need to focus more on the artists who use the software daily and think creatively, rather than simply trying to match competitors like Runway or Midjourney. Firefly is a good start, but Adobe could be more thoughtful about how these tools genuinely support us artists. After Effects is incredible, but it could benefit from deeper involvement and development with its core user base.
I hope companies will steer AI development towards truly supporting artists rather than chasing fleeting trends. Just like with NFTs and Web3, we've seen how hype can create bubbles with limited real functionality.
Thanks, that's the most accurate explanation of what's happening with with AI stuff.
Currrently I am resistant to use assets created with AI from scratch, but I find AI tools very useful when you bring something to it as a starting point, let's say a image, drawing, design or footage of your own, then you use AI to edit this material like upcaling it, removing backgrounds, tracking objects, cleaning up the audio, interpolating frames, etc... there are tones of boring and technical tasks AI can help which I am comfortable with. But we need to be he masters of the creativity so doesn't matter how fast AI can help us create a new script or write ideas, we need to shake lazyness from ourselves and put our mind at work, imagine, reflect and debate on things, otherwise those things that are create can not be claim as your own creations but AI's.
Thanks for the video man.
I'd totally agree on the fact that AI is a tool, and a as a tool I't would never be as creative as the person using it. Ai works very well, not perfect but it is still something you work with, as well as the other tools you use on your workflow, so I think we as designers have to use and put to the test all of these tools because it will also put us to test. I think that natural intelligence will expand over AI, it will make us more inventive and will make us push more and more barriers on creativity and what can be done will the tools at hand.
I love the amount of trust you have in your render output Jo!
tbh as an artist, i love to create things. what i actually would want to spend less time on is not creating things, but all the business stuff. i would much rather spend my time making assets than talking to clients, creating and sending invoices, filing taxes, etc.
Problem with technology change is we don’t get to choose what these horrible things do to us. Look at social media and the massive impact it’s had on us.
Dang Joey! Thanks for the Brevidy shoutout 🫡
Great episode. Very interesting and insightful. Thanks.
Maybe the companies will need to ask Ai how to keep on being profitable.
Nice talk, appreciate the insights. I would only add that it’s probably worth looking at what happened to commercial music houses over the last 15 years if you want to look at the future of motion design + VFX.
Best explanation of the current AI world I've seen. At least for our industry. Nice work.
I watched the entire video and found it really interesting. Would be nice to have your opinion in the same format on this subject every 6months or every year. Just a question though ... I don't know if it was a bug or if my previous comment was manually removed ... If it was, could I know why ?
AFAIK we haven't removed any comments, but YT auto-moderates and it can be heavy-handed.
@@joeykorenman Ok I'm relieved because my comment was challenging your analysis but in a respectful way so I felt "Holly cow, no contradiction is allowed in here 😅". Glad to know it wasn't intentional.
@@adelacelle I'm all about spirited debate! I'd never remove a comment unless it was spam or just mean. We're all figuring this out at the same time so respectful debate is the best way forward IMO.
Really interesting informed piece Joey, thanks for this. I have to say after a couple years of nightmare fuel of examples of AI it does feel like the gravy train and endless promises of linear development is coming to a crashing halt soon. Some interesting news out there regarding the fact that the training models themselves may literally be pretty much rung dry already, certainly nowhere near large enough to make the kinds of jumps and leaps we have seen early days in these various models. I hope it does settle into good useful time saving tools to avoid the nitty gritty admin stuff we have to do that disrupts the animation and design side of things. That would be cool. The glee at which certain X users had that artists careers were all over seems to have been, like most things on X, completely unfounded nonsense.
Just like broadband was extremely expensive in the late 90's and is much cheaper now, the cost of AI Compute will go down. We already see this happening rapidly.
To what ends? What we have is something that can generate low grade stuff. I think the biggest let down for ai is the shift in attitude that Joey describes. Once the novelty wares off the tools get really boring, really fast. For example, I’ve used ai for basic writing tasks. After a little while my own writing style for these basic tasks improved to the point where I don’t really benefit from the text generation tool. Is this revolution going to be a more efficient and slightly more accurate Grammarly?
Nice talk about the future. It's refreshing to hear a more positive view. And yes I remember the steam days! Greetings from back in Boston - Brian
I think this is the most level headed/measured response to all of what is happening in this space. I personally have started dabbling in AI for very specific workflow needs that I have. Like for instance a tool called " Depth scanner" on AE scripts works decently for making depth passes on footage. the Rotobrush tool in AE is better than it has ever been thanks to AI. Image/Video upscaling is useful for certain pieces of footage/effects that you are going for, but not everything. I still need to take the depth pass and use it as an asset to manipulate other pieces of footage. Image/video upscaling a lot of the time adds details that you dont want, and will need to be cleaned up in someway to make it usable. The rotobrush tool is very good, but you still have to wait for the results of the brush strokes, and even then it doesnt work on every single piece of footage, which will inevitably means still a bit of rotoscoping. I think these are all realistic uses of AI, in its current form and I personally am sorta ok with their uses in this limited context.
When you mention the vibe shift that is happening I think is due to AI people realizing just how not very useful these tools will be and stopping the constant onslaught of "AI will replace animation" posts on every single social media website that exists. Lol for instance this person "robert" in the comments for this video proves my point! He is talking about AI tools replacing the mass majority of artists and that we should still be scared of it. Thats what pisses people off. Sure there was and still is backlash about the ethical stuff that is happening, but i think what really boils peoples blood is people like all of the AI influencers on linkedin who are talking about "oh hey sorry this job market is about to be crippled but thats the way of the world in terms of innovation so deal with it!" That boils our blood because the people who have money to make projects are listening! Investors listen to AI influencers over people who are actually in the trenches willing to work hard to stay in this indsutry, which in term destroys any sort of rational thinking and critical reasoning that can be discussed about AI. Its so damn frustrating when I've applied to over 400 jobs, AND have to deal with the noise that is trying to take over the space that I want to be apart of so badly. Its business, so I don't blame these people for following trends, but I've had enough of them tell me to go fuck off to another, that I am ready for this bubble to explode so they can find some humility. I hope OpenAI gets the axe and we have strict regulation around it so this can move forward into a more symbiotic relationship rather than there being constant antagonizers towards this field. I know its about ignoring and just moving forward but come on, we are all trying to make this work and the only way to move forward is to have civilized discussions like this one.
So thank you again for having a level head regarding this subject!
I think you nailed it. Investors and companies are definitely swayed by AI influencers shouting "THIS CHANGES EVERYTHING," but if they listen to on-the-ground artists doing real-world stuff, the results (while impressive) are not nearly as great as hyped. Maybe in a few years that's different.
Useful AI Tools in my motion workflow are depth scanner, (sometimes I use a photoshop action to go through a PNG sequence and apply the photosnop depth neural filter though), Topaz Gigapixel & Video enhance AI (The best upscalers imo - also video enhance has a really good frame interpolation tool), Zoedepth, very useful when you need to quickly add animation to am image that's somewhat spatial, lalala ai for vocal/music separation, adobe podcast, ChatGPT/Dall-E for certain illustration generations, these always need tweaking but there's certain scenarios its useful.
Then there's all the standard Adobe ones on top of Adobe podcast eg gen fill/rotobrush/select subject etc
I've had other peers blanket refuse my offer to upscale a video because the program I'm using it called "video enhance AI", but will happily use the rotobrush, I think people shoving a blanket hatred to anything "AI" is stupid, although I can appreciate not being a fan of generative AI
Bro, so many golden, chewy nuggets here!! Much respect. Thanks for this.
I think you are spot on!
Is there a good AI rotoscoping and green screen keying tool yet? Seems like this would be high priority for this industry. Eliminating the most mundane and tedious aspect of our jobs.
I've been saying this for a year at this point. It just doesn't have the capacity to replace an artist and likely never will. Especially without a precise level of control and workfiles.
And enough compute… nuclear fusion would be a prerequisite I think… if we don’t destroy the climate by burning fossil fuels to power these comparatively limited tools.
I feel the bottom line of AI is what we value as artists, as humans , and what clients value from us, Ai may be able to counterfeit something into existence, quickly, and it may even look great, but at what cost, we are the the artisans of our chosen disciplines , our skills should be valued, please dont accept that AI is inevitable in our field, for learning, for live language translation while traveling, personal productivity, all fine, but it has non place in the arts, were humans, and have lost value of living a human experience,
Ai will break away when someone makes an Ai that can use after effects. Not make a rendered video but make an After Effects program file where you can either render as is or make the tweaks you want.
This already exists at a functional level. It's called templates, and some have script-based interfaces. Problem is it's never exactly what you need for a particular client.
Feeling less scared of AI after watching this - hope youre right sir!
This is suuuuper interesting and you make great cases and I learnt a lot.
However when you speak about what technological development means for the time you spend on a creative project I have to make my case. @joey You say that studios make more money because they can finish projects quicker and with more quality, right? But how capitalism works is that when one studio decides ‘wait a minute, we can lower our price for high end animation because we input less hours - and then all the business will be ours!’ The market is going to adapt and high quality content will simply become easier to make. So I believe most creators will in the future spend less time to get a project where it needs to be. Simple example: one of my client knows it costs way less time to put captions on a social video, so he’s going to ask for a lower price.
Also one other small note: midjourney has a web based interface since a few months and it works great.
I don't think I said companies would make more money with AI (I'll have to re-listen), but that they'd be able to do the same level of work in less time, which tends to just mean that they spend the same amount of time as before but get better / more results. Roto is a great example... I remember having to budget in 2 weeks of roto for certain gigs, but now rotobrush / ML-assisted roto can do 80-90% of that, so in the end you spend more time finessing the comps / design vs. having to do that stuff faster to accommodate manual roto.
Globalization has been driving down prices well before AI, and is probably a bigger influence there IMO.
I didn't know MJ had released that interface to everyone yet! We didn't have it recently, but it looks like we do now.
The point everyone seems to be missing is that the AI companies stated goals are NOT to be profitable, but to drive market values down, influencing costs. The AI fantasy is that you will invest enough into an AI for the Ai to figure out how to solve unsolvable issues with super intelligence. It goes something like "right now it might seem like it takes an insane amount of energy to run a certain model, but past a certain threshold or even horizon, the AI will be so advanced it will figure out how to power itself efficiently with solutions we can't even think of now." And that's basically the logic behind the Sam Altman's of the world.
If you ask me, it's all a Charade. They are doomsday preppers playing god with humanity at stake and they're addicted to the possibility of being the ones to pull the trigger.
I guess I've got a solution for Open AI. Why don't they use customers' GPUs for calculations? Like if you don't have a gpu you need to pay use sort of a render farm, but optionally, you can use your own gpu.
I've been dabbling with AI image generation with StableDiffusion since before version 1.5. Even though I'm no expert in market tendencies or in any of these big techs' vision, while browsing reddit and other dungeon domains of the internet, I still believe the majority of use cases of AI consist of pornography or deep fakes and, sometimes, coupling these two together...
Fantastic job with this video, Joey. I love that you started off with some historical context. And I appreciate your honesty around your speculations. For what it's worth, I am 100% aligned with your takes across the board.
So far still a lot of gimmicks I have tried incorporating it in my workflows but end up giving up. Dont think any of these tools are more than fun features. The technology is incredible but none really is using it to properly help creators in any meaningful way. Let’s hope that after the lawsuit is done, they can get their things together and actually use it in a more meaningful way
Thank you
have you seen the article where if you feed generative imagery to the generative model it gets spits out horrifying results very fast
The clients will happily accept 70% Quality for 10% of the price, especially because viewing habits change rapidly. We already see a lot of diminishing quality in ads. Further, with AI it much easier to create target audience tailored content in parallel, which is highly in demand.
Some clients will, for sure. Some won't, though. I think the bar gets raised for what an artist is expected to do.
Great history lesson and commentary. I can see a lot of these technologies becoming things we run locally as GPUs get way better, and so the companies selling it as a service will die off. If Flux gets better and I can run it on my gaming rig, why keep paying my $30 to Midjourney.
You're right about the shift in the conversation, but I think it's only happening in VFX and video. I was at a talk by someone high up at a VFX studio, and he was showing off all the tools they are using at the concept stage. And it was just Midjourney and Luma, so it's nice to see that established industry doesn't have anything much more advanced than the average consumer. The comics industry and a lot of artists are still at the frothing at the mouth stage. It maybe isn't quite so bad as two years ago when people were getting threatened with violence, but it's still pretty toxic.
I turn to AI when I hit a roadbloack of my own skills. I'm bad at writing emails, so it's become my starting point for correspondence so that I don't get anxiety over putting off the emails. I use it to design characters so that I can redraw them and animate them. I use it for stock photos when I'm doing a boring corporate project that needs 100 specific images I'd take days to find on Envato and Pexels.
It’s a pleasure to listen to you talk about something. I’ve been listening to som podcasts since I’ve completed Ae kickstart back in 2018. Them, and now motion mondays are my source of life insights and news to this day. Thank you for your content!
the first mover isnt always the winner
Ai Generation is at a point where it can generate "A kid's 10th birthday with everyone having fun... Etc etc etc", but what it can't yet do is generate "My Son's 10th birthday as it happened on X date at X location" and I think that "personalization" is a main factor missing from Ai generated design and is where artists can really craft Ai into a tool to help unlock their ability to find solutions for clients.
I appreciate you taking the time to do such a thoughtful deep dive. Keep up the great work SOM.
Imma be honest, I click on these for the wealth of info because I enjoy your content, but also to see what other bands we have in common by way of your t-shirt 🤣
I really want AI to become sentient, I've always wanted a robot friend! I hope we get to that point before funding stops lol
Seems like it's over for voice over artists as well
For 3D animation, ray tracing actually simulates how REAL light works. No video generative AI can replace it in this context.
AI just morphs pixels, and I personally don’t think it sees that much depth in an image, if we explicitly not optimize it to do so.
Actually I doubt that video AI could even get to the point where no morphing at all will be present in a shot. Even with the best AI it is like telling our brains to get a clear and realistic image while dreaming in a sleep, without all the nasty artifacts of our memory recall and imagination.
But because morphing between concepts is the nature and the artifact of the system, I find it is really useful in abstract art ☝️ and from the beginning it was always like that, and I was fascinated to see the «out of this world » stuff that AI can spit on you with some good prompting.
❤🔥❤🔥❤ Lots of good stuff to chew on in that one. Important stuff to consider... career-wise I mean. I want to agree, but I also feel pretty amazed and scared by how fast it has improved over the last 18-24 months.
School of Motion Alum here from AE Kickstart. As someone who now works full time in AI (I am the Head of Content Marketing for one of the largest AI Art Resource sites on the web), I keep up with all of these videos and really wish that you guys would dive into whats happening in the open source community sometime. Especially on the ComfyUI side of things. I could show you some things in Comfy by utilizing the skills I acquired in After Effects that would probably give you a lot of new ideas about how the technology could be used.
Anyways, great video, great level headed talk. I am still searching for the commercial viability and the path to longevity in all the madness. Onward.
Also, what you mentioned about applying the style frame to animation you create......has been possible in ComfyUI for at least the last 6 months :)
This was one of your best commentary videos yet. You cut through the hype and covered all the problems, roadblocks, aspirations and realities of the situation. AI is now being taught as coursework in art schools, so it's not a question of if but when. As one of the professors at Ringling said to his students: "You won't get replaced by an AI. You'll get replaced by someone who knows how to use AI."
Corn on the bob for coffee steam, huh? Who would've thought?! 🤷♂ #TheMoreYouKnow
36:45 sounds like a fancier version of just ebsynth which is a pretty nice ai tool that existed before all this stuff
Someone's been listening to Ed Zitron
Off-topic but is Joey looking younger ? What are you eating Joey …did you change your diet ? Inquiring minds want to know your secret lol 😂
New haircut.
*Can't wait for the burst of Ai Bubble, so you can actually focus all the efforts and resources toward Practical use of AI*
Once Ai Bubble is over, grifters will move away from Art, Engineering, Everyday of "dash dash so over cause of this new AI" etc.
Good stuff Joey, thanks for sharing!
Thanks for all these thoughts! This new technology is growing fast and much more faster than the crypto/nft bubble. We certainly integrate them shortly to our pipeline but right now we missed tools to dissect & animate them with the perfection a professional job requires. I was learning graphic design on analog tools back in the days… and I still use some in addition to digital and online new tools. Time flies !
Let's call it what it is:
People in the web are blindly pushing and over-inflating E V E R Y new technology that comes out, they don't question it, just because they're afraid if they don't endorse it - they'll be irrelevant on their field. It's like UA-camrs who unbox a product and say it's awesome without testing it for a month at least. Same people who said "Threads" by Meta is the new best thing-are the ones pushing AI without understanding in what way can we (legally) utilize and make profit out of it.... 🤔
IMO you can never have too many teeth :BBB
I agree with you, we can say we used to do a frame-by-frame on paper after in pc programs now on iPad, it's always will be technology development and we need to go with it because it will make it easier and much better at the end for us.
The stock photography going out of business thing is kinda dumb. Its' actually easier, faster, and about the same cost to use a stock photography site to search for an image than to generate one, and by definition the image will always look real. Also, many stock people sell images of real things and place, that must be real images. There would be no use, for example, for an AI image of Paris in a travel brochure. That would just be fraud.
Am I the only motion designer thinking that this debate is just capitalism selling us, one more time, that eeeeverything is great and not talking about the horrible world that this and other technologies are helping to create?
I mean, can we forget that OpenAi worked with Cambridge Analytics to create hate online and helping Trump? I cannot trust anything of this without feeling ethically a piece of sh*
From curiosity, why do you think this technology is creating an horrible world?
@@itayozari6252 apart from our uses, It is being used to kill people in Gaza, for example.
Such a spot on video. Reflects my own personal view almost 100% love it!❤
Great video!, Thank you for sharing your point of view in a very structured, clean and professional way. Keep it up!
Fantastic breakdown and insightful analysis of the current situation. Joey, your delivery is incredibly smooth!
2 thoughts: Things like the VCR and digital cameras struck fear in to the film and photography industries, but of course were widely adopted. One thing that sets these new generative AI tools apart is how quickly they just seemingly exploded out of nowhere. People had little fore-knowledge and suddenly had to face what looked like totally disruptive technologies. But as you say, now that its had some time to percolate in the collective consciousness we're a bit more level headed about them now and hopefully moving forward. Second - glad to hear you address one thing I've said from the beginning, they can't be accurately directed. If I'm just noodling around at home and make an epic alien space battle, yea that's fun - but when its for a client and they have specific notes on what hero moments need to be hit, framing, story beats, etc - its just not there.
The style transfer thing you're discussing around the 36 minute mark is already very close to being ready for prime time. Great plugins/addons for both 3dsmax and blender that are only gonna get better.
I think this will unlock a lot of cool possibilities.
Joey, you've always positioned yourself as a motion person but you've got great analysis on the creative industry at large. I always recommended your manifesto book to many non-motion designer freelancers too.
3D modelers have been moving away from Maya because it's $300 a month. Lol.
everyone should see this, even out of graphic industry
What is the source? 5B loss per year?
Tons of articles about it. Yahoo news, The Information, etc… I’ll try to post a link in another reply. Sometimes UA-cam deletes links.
deadline.com/2024/07/openai-chatgpt-home-may-lose-5b-this-year-1236024704/
@@joeykorenman Thx! Great video btw.
Is this channel about motion design or AI?
What a great video, thanks a lot for making it! You shared a lot of good highlights, one thing that makes me curious and that no one is mentioning compared to the Web3 bubble is the energy consumption, the projection is crazy
I understand you sell courses of the tools will be useless with AI, but it is happening right now. It is not that good? Well people have a real low bar theses days.
first half is interesting. So they need to get more customers to be profitable or make a loss and then you say the vibe shifts. so you advise motion designers to get locked in. in the end when no more people are there to lock in, these firms will do some ugly monetization. when the designer unlearned the normal tools and is dependent on some company running a server with a model on it it will be ugly for the designers when that goes away or becomes really expensive (those stockholders want to see a return after all).
Also LLAMA is not open source but with open weights. The training data is hidden. Probably still all copyrighted material in there... zuck calls it open source to please the EU AI regulators, but it totally is not!
Interesting summary with some weak points. Artists, learn your tools and don't get dependent on generative models. I am totally not against some technology that can perfectly remove backgrounds in videos while preserving every moving hairstrand, but those models trained on copyrighted work are wrong to use!
Ai training on intellectual property without consent of the owner is theft. Right! RIGHT?
Is this a finance channel now?? I don't care! Ai is here to stay. Stay on top of it, that's all you can do.
So what percentage of this talk was written by AI then?
The LLM in my brain wrote this.
x2 Speed type video
Hell yeah, brother.
Great analysis/explanation! Regarding the tide shifting, I still see a lot of blind AI hate pretty much everywhere, but I've seen a few people come around a little bit once they see a project that used AI as a tool and still had a lot of effort involved. But even then, Corridor Digital has gotten SOOOO much flack from attempting to use AI to make their Rock Paper Scissors anime.
+1000 aura
what are hallucinations?
Let's. (Great T-shirt design)
Johnny Booth are so good.
This entire video feels like copium.
So many of my friends are cover art designers, logo designers and 2d artists.... And they're all doing door dash now because the jobs simply disappeared.
Drone work and abstract vfx are already suffering.
it's funny that nobody cares about stock illustrators and photographers. Everybody just admit that they are doomed. Design community kinda rase the voice but for themselves, essentially.
I stopped caring because AI sucks. It is a clown show. and if anyone uses it they are cut out. It just simple that way. And AI sucks still.
If a company raises their prices by a factor of 10 and as a result loses 50% of their customers, they're now taking in 5 times as much money with half the overhead. Pretty significant bump in profit.
Could be, we'll have to wait and see.
progress is moving faster than your realize
Ai is your employee and you are Walt Disney. Thats all. It will always need your love.