> AI-generated videos have developed their own unique look. There's a visual quality that marks them, a subtle wrongness that your brain picks up on even when you can't articulate exactly what's off. It's the new uncanny valley, and I feel an intense revulsion whenever I encounter it.
Yesterday on the comment thread for the "Attention is Bayesian Inference" blog post (which definitely seems to have been written by AI), a couple people were asking "who cares who/what wrote it, if it has good information", and others were struggling to articulate a response. Well, for me this is the response: it has a specific tone, and every time I see it it activates my gag reflex and I back-button straight out of the page. I'm not interested in examining or deconstructing this response either, as far as I'm concerned it's evolutionarily adaptive and I intend to keep it.
That's only relevant until the AI output no longer has a tone that triggers your gag reflex. Then you'll be in the same boat as everyone else, and the "who cares who/what wrote it, if it has good information" question will apply to you as well.
One of the arguments I keep seeing from people churning out AI video is that the tech is enabling people "creative freedom" that's been made possible now even without the technical know how.
However, 99% of the the "creativity" from what I've seen is done by the AI (how it should look, where the cuts need to happen, the tone, color grading, etc). Which is to say, it's taken from other people's (creative) work.
While a big part of being able to create a good video has much to do with storytelling, the craft of shooting and editing a video is a big part of the creative process as well.
AI video isn't "enabling people to be more creative," it is quite literally removing creativity from the process all together.
Not only are they removing creativity but they're eroding the medium with a flood of high quanity/low quality of misinformation or fakes. My youtube feed is innundated in AI generated talks, audiobooks, history etc. The audiobooks may be okay but I can't trust any AI generated content. At this point I wish I could just fiter it out. Just yesterday I saw a video by Yuval Noah Harari and almost towards the end I realize it's AI slop. It has nothing to do with the author except for his syntehtized voice. The content is not even vetted by a human, it's being created faster than users can consume it. For now I can spot these AI channels and attempt to avoid them but am not sure how long this is going to last.
I was involved in a conversation about cheating in video games the other day, and the topic shifted to AI use in music. Someone said, "using AI is like using an aimbot for music." I absolutely love that comparison since it highlights the shortcut past creativity/skill to get a computerized best result while also associating it with blatant cheating.
The "enables creativity" argument is ironic since the root of the word is "create" and AI users are literally removing the "create" step from their production process.
The analogy is great, but it breaks down when we talk about professional use. An aimbot in multiplayer is evil because it ruins the game for others. But an aimbot in game development (e.g., procedural aiming animation) is just a tool. The problem with current AI video is that the model often shoots not where the director wants, but where it's easiest to hit (the template)
Just think of what photography did to portrait painting. The camera and film did all of the color work, the person hitting the button had no need to mix their own pigments anymore. All the creativity was removed.
I agree. This is not about creativity, but about producing something that kinda looks like professional art. People assume that creative output has to pass some bar, so that other people can appreciate it. But you can also produce interesting art with very simple techniques. XKCD is an example that comes to mind.
If you think this for one ai you must think it for all.
I can now instantly visualize anything I think of. That is creative power. The same for code - I can instantly scaffold a frontend I think of in google ai studio. Its not all great and I have to keep the slot machine spinning. But it's empowering.
These "ai kills creativity" arguments are all rather uncreative.
>However, 99% of the the "creativity" from what I've seen is done by the AI (how it should look, where the cuts need to happen, the tone, color grading, etc). Which is to say, it's taken from other people's (creative) work.
I don't get why this is a big deal? Like 99% of the creativity of taking a video of an ocean is also just taken from the nature. Your creativity was actually a small portion of "information" out of all the bits required to make that video.
If you're correct, creative people have nothing to worry about then. Creative people will still stand out and have a role. But the world already has an oversupply of creativity, and a need for a lot more of boring, uncreative things. A plumber, mechanic, or programmer, all need minimal creativity -- 99% perspiration, and 1% inspiration, as they say.
Sorry that isn't true. Lots of smaller communities online, especially around small streamers make small meme style videos. Usually these are mocking someone doing something dumb online, or jokes/memes about the show etc. These are similar to parody videos online that were hugely popular on YouTube back in the early 2010s.
People did do this before AI. Usually cutting people faces out and sticking them on actors faces in existing movies, or subtly doing parody cover, doing clever edits (Cassette Boy is a notable example) or people were performing and recording it like the "Epic Rap Battles of History".
All it done is allow people to create these sort of to a higher creative standard, in other cases it allowed people to create jokey stuff that they wouldn't otherwise be able to create.
People using this technology in this manner is clearly creative. They are using the tech to make something new and unique for a particular audience.
In the 80s/90s you would be complaining about people using tracker software and samples to create music instead of learning to play an instrument. Under your logic someone like FatboySlim isn't a musician.
> 99% of the “creativity” is done by the AI…it's taken from other people's (creative) work.
Yes, and prior to AI, 99% of creativity was also taken from other peoples work. This is how creativity works. Man has no ideas in a vacuum.
As humans we take influences in from what we’ve seen and when we “create” things we’re simply combining things together, resulting in output falling somewhere on the spectrum from a 1:1 copy to somewhat novel.
If you can’t spot the influences in a painting or song or piece of graphic design, it’s simply because you aren’t deeply familiar with prior works in the field at the time.
I've been making music videos with a very satisfying creative process: Use AI to make a ton of images, pick the very best ones, and carefully arrange them in the right order. Example: https://youtu.be/r-_dJNgt3SM
It's both... I think it can cut both ways... I was largely on your side, as most of the AI stuff I've seen is just annoying more than anything. Then the "Star Wars: Beggar's Canyon" video completely changed my mind. The voice generation is clear and inflection appropriate... the cuts, sequence, and overall effects are clearly put together in a consistent way.
I can only imagine how much work that video was, but is really the first thing I've seen as AI video that gives me hope... and obviously a work of passion and a lot of effort to get such a great final result.
It’s almost like people that use the tools to the best of their ability will produce higher quality outputs. This is a new tool like we’ve never seen before with a low skill floor to use it. We don’t know what the skill ceiling is yet.
There is a lot of content whose final value ultimately doesn’t justify the time that would be spent on creating it. Worse, there is sometimes only a specific window of time where a content has entertainment value and it quickly drops off after you leave that window.
For example, someone recently created an entire anime style video of the United States invading Venezuela and capturing Maduro. It came out awesome! But… to make a similar quality video by hand would have easily taken a team of humans several months at best. By then, the news cycle would have moved on, and the delivered video would have even less value being watched so distant from the event that inspired it. People wouldn’t care anymore, and it wouldn’t have as much impact.
AI is the perfect solution for delivering such a video exactly when you need it. To me, this is one of the more acceptable uses of AI, to make fun throwaway content that you just laugh about for a few moments and then move on with your life. Having a human dedicate a large amount of their time to make such videos would be sad, and it’s better if they preserve their effort for more serious artistic works that are timeless in their value.
A lot of technological advancements (or use scare quotes here) that claimed to let “regular people” be creative just end up passifying us and relinquishing our creativity. Why play the guitar, the record player already does it better. Why learn to juggle, some kid in <supposedly whiz-genius country> already does that.
Of course of course of course, people have been inspired by play in bands and to learn to juggle. But the trajectory seems to be to move away from small intimiate creative offerings, like playing music in your living room, to passively recieving Taylor Swift into your earbuds.
Those aspects aren’t the only “creative” aspects of filmmaking.
What if someone used AI to say something important or get across a different perspective?
99% of people in the world do not have the time/money/connections to achieve a film vision. They were locked out of the ability to create their vision - creating an high concept or production-heavy film has been the most privileged position in the world - until now.
It's not like these discussions haven't have quite a lot of precedence. I bought a 4-track tape "portastudio" back in the 90s, but it didn't make me a good producer- years of playing music with dozens of groups did that. And I bought a canon xl1, but that didn't make the movies I was making more compelling than if I'd had budget to shoot on 16mm.
Creativity and good craft come from struggling to communicate an idea; anything that short circuits that will make you a worse artists.
AI is, to me, like claming that because you have access to a forklift going to the gym is no longer useful- you're missing both the aesthetic and material reasons why folks have been doing things.
What's worse- people claim a forklift is better than, say, working out and then say to the folks who were working out "this is what you're doing and it is dumb to spend all that effort doing it". That's nothing new either- I get why folks look at a lot of my early, cracked FruityLoop-based work from 2002 with some disdain-- it misses all of the nuances and trade-craft knowledge that I gained playing in bands over the following 20 years.
And I know how it sounds because I can hear it in the products of other young musicians: I play a lot of jazz with kids in their 20s who have way better chops than I ever will and who have a lot of the real book memorized. That real book allows them to pick up the tunes super fast, but they don't have a lot of the nuance that the 70 year olds I sometimes get to play with have in their playing.
I use AI for the dumbest stuff, like writing contracts for mechanical licensing. It's shit, because I'm just looking to copy some already-written document without paying a lawyer. I can't see see why folks would -want- to use it to make documents that no one else had previously thought would be worth writing and to me it very much seems like it can only produce, by definition, derivative works.
Musicians are terrible about that. I love being in a cover band, because I usually like the people and playing. But absent that, simply being able to "poof here's 'your' cover of Stariway" seems super boring, almost soul crushing.
And paying AI do the creative parts- that's like paying someone to have sex with my partner because I wanted to spend my nights playing video games instead of woodshedding on pedal steel guitar :D
I think there is some tenuous analog between today's AI, and yesteryear's music synthesizers and "modern art". Regarding music, I understand synthesizers were created to mimic real instruments, whereas artists used its capabilities not to mimic, but to create entirely new sounds. My limited understanding of modern art (and without a search I don't distinguish between expressionism, impressionism or any other isms because of my ignorance of the differences) is that it was a reaction to the modern world, and an expression of something that could not be anything other than a creation of the human mind. AI may be helping people to copy existing aesthetics, but I'm hoping/waiting for it to enable people to take it in a completely novel direction that can only be augmented human expression.
Yes, 99% of people are uncreative and use the creative freedom this way. Besides, many truly creative people won't even touch it, some because of the ethics and others because it's not up to their standard yet. Does it really look surprising to you?
>AI video isn't "enabling people to be more creative," it is quite literally removing creativity from the process all together.
That's quite a leap of thought and doesn't follow from the first part at all.
I am currently working on a script with my uncle who is a professional playwright, my grandfather wrote for Benny Hill. So it runs in my family.
Biting the bullet, I have accurately had it generate the scene of the cafe. Generating visual imagery of how I imagine it; a cyber dystopian gothic cafe with wandering goths.
It's capped to ten seconds and that alone would've costed me a fair sum to even get a small render without revisions.
I could now technically pitch this with the AI works and get the funding required to hire actual animators/actors or spend my savings myself giving it to an artist depicting the theme and style; not that I could afford it.
How is creativity lost in that scenario? AI is a tool, it shouldn't be a do-everything, I created the script and it's generated the visual mock.
Personally I wouldn't want it to create the whole show and would personally prefer real folk to do the magic but for my own personal consumption it's given me the excitement and further drive to continue.
> AI video isn't "enabling people to be more creative," it is quite literally removing creativity from the process all together.
This is very much not the case with the Stable Diffusion/Wan/general open diffusion models space.
The amount of effort and creativity that goes on in creating complex workflows, custom LoRAs, fine-tunes etc is genuinely a new area of art imho. Sure a lot of it is going to produce things that most people might not consider "art", but it's unfair to lump the work happening in this space with people randomly prompting Sora and dumping it on TikTok.
I think most people have the ability to express themselves through art, but if their first experience of trying to make something is using AI tools, they will never discover that they can make something genuine that reflects their experience.
> how it should look, where the cuts need to happen, the tone, color grading, etc
That's not creativity. That's skill. A person can think of something (creativity) and not know how to bring it to life without the skills to do so. The creator still needs to have a creative vision, and communicate that vision effectively at the very least.
Of course, there's also people that are just cranking out slop, but I'd argue there's no creativity there, other than finding creative ways to copy others. But that's been a problem since forever.
AI (as it currently exists) will never be "creative" in the sense that it can only imitate or interpolate between past works. At some point we'll recognize that AI generated works are boring and predictable. AI will probably never invent a new musical genre or artform since it can only reproduce or recombine works from the past. I wonder what happens when the internet is so full of AI generated slop that that the only things worth "training" on were made in the time before AI generation became a thing. Will AI generations be full of dated references to a time gone by?
People may say it increases creativity but I see it more as lowering the bar to produce things. The same could be said for a lot of inventions like photography made producing images easier and I'm sure a lot of portrait painters lost their jobs to photographers. I think the danger is that we may see a very rapid erosion of jobs in the creative space that won't be easy to transition into new fields which I feel will have a detrimental effect on our society.
We need to disambiguate what you mean by "AI" here. If you're referring to people typing prompts in a frontier web UI, maybe. But if you're using local models, you can do quite a bit more, ComfyUI workflows can get crazy. Multi-region controlnet guided generations with IP adapters and loras, etc.
The art gatekeepers want to try and paint everyone using AI creatively as monkeys mashing ignorantly on keyboards, taking the unmodified output and parading it around like Picasso. The reality is that is to AI art as bathroom dick graffiti is to manual art.
Or you could argue that it’s simply lifting from other people’s creativity. Like making a song entirely of sampled sounds from other songs with a key change.
The author is right that AI video is being used to propagandize and rot minds at scale in a way sort of unthinkable a couple years ago. Short-form video content wasn't good for us beforehand, but any balance has been fully upturned.
(I think that's all intrinsic to genAI in a way the author doesn't want to confront, but there is a crisis)
Put down the phone, close that youtube, tik tok or whatever tab.
Don't try to explain to your relatives which video is AI and which is not, explain to them that everything on the internet is bullshit and trolling, AI or not it does not matter it's all bullshit to sell ads.
Tell them they are better off watching TV, even Fox news.
AI generated videos are being used in more and more ads. To cut costs I’m sure. The result is that ads that were just annoying are now terrible and jarring.
When I hear talk about AI risks I mostly hear things like runaway super intelligence doing whatever it wants and leaving humanity in the lurch. But what about more realistic concerns, like accelerating the race to the bottom by cheaply and poorly ripping off other people’s work and forcing everyone else to do the same just to keep up?
99% of everything is bad, so that unsurprisingly includes AI videos.
But i've seen several good videos made using AI, including pretty much everything NeuralViz[0] on YouTube makes, but also some that have been posted in older comments here in HN. Igorrr's ADHD music video[1] was also made using AI and fit the music perfectly.
The common aspect with these "good" uses though is that they do not let the AI do 99% of the job (as mentioned in another comment) but they still involve editing, writing/scripting, acting (NeuralViz for example uses his webcam to act both the motion and voice in his videos and uses AI to change them) and to some extent leaning into the "weirdness" that AI videos have instead of ignoring it.
[0] https://www.youtube.com/@NeuralViz (i high recommend watching them in upload order because they all build into the same "universe" and often make references to older videos)
I have an idea for a movie. Its kind of an epic idea - more about the characters, story and the feeling. I want to get the idea across as close as possible to what is in my head.
AI has given me the opportunity to do it. I just have to keep prompting to get the exact feeling and I can share my idea.
I don't like these curmudgeonly takes.. I'm sure this "All AI Videos Are Harmful" take will age just like how "All videos are harmful" and "All cartoons are harmful" have aged.
I can’t seem to figure out a good prompt for these yet. But there are people doing clever things (short clips with first frame previous last frame, instructions on the image of first frame, etc.) and I hope that over time these get to become as good as image models have.
His rationale ignores all the legitimate and harmless use cases for AI. I've seen people make with the help of AI comedy/meme/comedy music videos and put them up on YouTube.
Making a hyperbolic statements stating that it is all harmful, because there are deceptive usages of AI is just asinine.
If you couldn't have made the story into a short film before AI video then AI video isn't going to suddenly make you a talented film maker. Eventually one day you'll be able to put your vague story idea in and get a fully finished coherent well made movie out but currently if you lacked the talent before you're still lacking the talent today.
We know this because actually existing talented film makers have done entertaining work with AI video. But it's silly to use it as justification to write it off just because you're not talented in the field.
I think the concern of the author, that AI video ultimately "enable(s) those who want to manipulate, deceive, and exploit people for engagement, profit, or ideology", cannot be understated. However, the binary claim that the author doubles down on throughout, "All AI Videos Are Harmful", dilutes this messaging.
tilt at this windmill all you like, but the genie is out of the bottle and there's nothing you can do about it but kvetch. you can't ban it, you can't regulate it, no more than the US government could rein in cryptography.
The author has a myopic view on Gen AI videos. They only focus on the most extreme examples and then hand wave cat videos as some crisis about reality. We had these exact same alarmists with previous technologies. Human's have an uncanny ability to adapt as their environment changes. Photoshop and 3D didn't destroy our brains and neither will this. Misinformation and the such have never needed AI to be effective. Those prey upon people's already existing biases and thoughts.
The time where this becomes normal and these alarmists become the fringe crazies ranting about the end of civilization cannot come fast enough.
it always rubs me the wrong way when people infantilize the masses. The "vulnerable" masses already already partake in lots of harmful substances and practices (tobacco, alcohol, drugs, gambling/lotto), AI videos are just another potential pitfall people will need to learn to be wary of.
Not sure how much of this is AI vs other techniques, though it's definitely AI as part of the toolset. But I think the result speaks for itself, it's absolutely impressive the work that has come out. I can only say, that the work that went into this video is definitely real, and I'm happy to see more of this, as opposed to "AI Slop" that we're mostly seeing so far.
This headline does not match the article at all. It claims all AI videos are harmful then immeidately goes on to throw this claim out and say old confused people might be fooled by the videos. This is a very narrow subset of AI videos.
It's not just a pedantic issue. Making these kinds of extremely bold and all covering claims is par for the course for those who want to find and exploit attention from others for profit. If that's not what the author is doing I think the most charitable interpretation I can come up with is that they just have no experience with AI videos and are going off sensational press stories.
They should buy an nvidia 3060 12GB ($200~ second hand) and try out wan2.2 on their home computer. It's really fun to be able to make these kind of videos on a whim. And they'll get some real lived experience with the subject.
> Or it's indirectly harmful by training us to accept a synthetic reality where nothing can be trusted and everything must be questioned.
Good. Its beneficial, I'd think. Lies are not new, the amount of lies, probably not new either.
I don't think we have (yet) reached a level of badness compared to the prior point of tobacco companies telling people around the world that cigarettes have medical benefits.
75 comments
[ 3.0 ms ] story [ 65.8 ms ] threadBigfoot comes to mind.
Yesterday on the comment thread for the "Attention is Bayesian Inference" blog post (which definitely seems to have been written by AI), a couple people were asking "who cares who/what wrote it, if it has good information", and others were struggling to articulate a response. Well, for me this is the response: it has a specific tone, and every time I see it it activates my gag reflex and I back-button straight out of the page. I'm not interested in examining or deconstructing this response either, as far as I'm concerned it's evolutionarily adaptive and I intend to keep it.
However, 99% of the the "creativity" from what I've seen is done by the AI (how it should look, where the cuts need to happen, the tone, color grading, etc). Which is to say, it's taken from other people's (creative) work.
While a big part of being able to create a good video has much to do with storytelling, the craft of shooting and editing a video is a big part of the creative process as well.
AI video isn't "enabling people to be more creative," it is quite literally removing creativity from the process all together.
Is Pixar Incredibles where a computer algorithm made decisions about exact colors worse than Disney Cinderella where it was hand-drawn animation?
The "enables creativity" argument is ironic since the root of the word is "create" and AI users are literally removing the "create" step from their production process.
Just think of what photography did to portrait painting. The camera and film did all of the color work, the person hitting the button had no need to mix their own pigments anymore. All the creativity was removed.
/s
I can now instantly visualize anything I think of. That is creative power. The same for code - I can instantly scaffold a frontend I think of in google ai studio. Its not all great and I have to keep the slot machine spinning. But it's empowering.
These "ai kills creativity" arguments are all rather uncreative.
I don't get why this is a big deal? Like 99% of the creativity of taking a video of an ocean is also just taken from the nature. Your creativity was actually a small portion of "information" out of all the bits required to make that video.
People did do this before AI. Usually cutting people faces out and sticking them on actors faces in existing movies, or subtly doing parody cover, doing clever edits (Cassette Boy is a notable example) or people were performing and recording it like the "Epic Rap Battles of History".
All it done is allow people to create these sort of to a higher creative standard, in other cases it allowed people to create jokey stuff that they wouldn't otherwise be able to create.
People using this technology in this manner is clearly creative. They are using the tech to make something new and unique for a particular audience.
In the 80s/90s you would be complaining about people using tracker software and samples to create music instead of learning to play an instrument. Under your logic someone like FatboySlim isn't a musician.
Almost entirely by studying other peoples works, the rest by doing something and getting feedback on it.
Yes, and prior to AI, 99% of creativity was also taken from other peoples work. This is how creativity works. Man has no ideas in a vacuum.
As humans we take influences in from what we’ve seen and when we “create” things we’re simply combining things together, resulting in output falling somewhere on the spectrum from a 1:1 copy to somewhat novel.
If you can’t spot the influences in a painting or song or piece of graphic design, it’s simply because you aren’t deeply familiar with prior works in the field at the time.
I can only imagine how much work that video was, but is really the first thing I've seen as AI video that gives me hope... and obviously a work of passion and a lot of effort to get such a great final result.
There is a lot of content whose final value ultimately doesn’t justify the time that would be spent on creating it. Worse, there is sometimes only a specific window of time where a content has entertainment value and it quickly drops off after you leave that window.
For example, someone recently created an entire anime style video of the United States invading Venezuela and capturing Maduro. It came out awesome! But… to make a similar quality video by hand would have easily taken a team of humans several months at best. By then, the news cycle would have moved on, and the delivered video would have even less value being watched so distant from the event that inspired it. People wouldn’t care anymore, and it wouldn’t have as much impact.
AI is the perfect solution for delivering such a video exactly when you need it. To me, this is one of the more acceptable uses of AI, to make fun throwaway content that you just laugh about for a few moments and then move on with your life. Having a human dedicate a large amount of their time to make such videos would be sad, and it’s better if they preserve their effort for more serious artistic works that are timeless in their value.
Of course of course of course, people have been inspired by play in bands and to learn to juggle. But the trajectory seems to be to move away from small intimiate creative offerings, like playing music in your living room, to passively recieving Taylor Swift into your earbuds.
What if someone used AI to say something important or get across a different perspective?
99% of people in the world do not have the time/money/connections to achieve a film vision. They were locked out of the ability to create their vision - creating an high concept or production-heavy film has been the most privileged position in the world - until now.
They believe the idea itself is the creative bit: that thinking of the idea of a dog riding a skateboard is in and of itself a creative act.
Creativity and good craft come from struggling to communicate an idea; anything that short circuits that will make you a worse artists.
AI is, to me, like claming that because you have access to a forklift going to the gym is no longer useful- you're missing both the aesthetic and material reasons why folks have been doing things.
What's worse- people claim a forklift is better than, say, working out and then say to the folks who were working out "this is what you're doing and it is dumb to spend all that effort doing it". That's nothing new either- I get why folks look at a lot of my early, cracked FruityLoop-based work from 2002 with some disdain-- it misses all of the nuances and trade-craft knowledge that I gained playing in bands over the following 20 years.
And I know how it sounds because I can hear it in the products of other young musicians: I play a lot of jazz with kids in their 20s who have way better chops than I ever will and who have a lot of the real book memorized. That real book allows them to pick up the tunes super fast, but they don't have a lot of the nuance that the 70 year olds I sometimes get to play with have in their playing.
I use AI for the dumbest stuff, like writing contracts for mechanical licensing. It's shit, because I'm just looking to copy some already-written document without paying a lawyer. I can't see see why folks would -want- to use it to make documents that no one else had previously thought would be worth writing and to me it very much seems like it can only produce, by definition, derivative works.
Musicians are terrible about that. I love being in a cover band, because I usually like the people and playing. But absent that, simply being able to "poof here's 'your' cover of Stariway" seems super boring, almost soul crushing.
And paying AI do the creative parts- that's like paying someone to have sex with my partner because I wanted to spend my nights playing video games instead of woodshedding on pedal steel guitar :D
All that is to say- I agree.
>AI video isn't "enabling people to be more creative," it is quite literally removing creativity from the process all together.
That's quite a leap of thought and doesn't follow from the first part at all.
Biting the bullet, I have accurately had it generate the scene of the cafe. Generating visual imagery of how I imagine it; a cyber dystopian gothic cafe with wandering goths.
It's capped to ten seconds and that alone would've costed me a fair sum to even get a small render without revisions.
I could now technically pitch this with the AI works and get the funding required to hire actual animators/actors or spend my savings myself giving it to an artist depicting the theme and style; not that I could afford it.
How is creativity lost in that scenario? AI is a tool, it shouldn't be a do-everything, I created the script and it's generated the visual mock.
Personally I wouldn't want it to create the whole show and would personally prefer real folk to do the magic but for my own personal consumption it's given me the excitement and further drive to continue.
This is very much not the case with the Stable Diffusion/Wan/general open diffusion models space.
The amount of effort and creativity that goes on in creating complex workflows, custom LoRAs, fine-tunes etc is genuinely a new area of art imho. Sure a lot of it is going to produce things that most people might not consider "art", but it's unfair to lump the work happening in this space with people randomly prompting Sora and dumping it on TikTok.
That's not creativity. That's skill. A person can think of something (creativity) and not know how to bring it to life without the skills to do so. The creator still needs to have a creative vision, and communicate that vision effectively at the very least.
Of course, there's also people that are just cranking out slop, but I'd argue there's no creativity there, other than finding creative ways to copy others. But that's been a problem since forever.
People may say it increases creativity but I see it more as lowering the bar to produce things. The same could be said for a lot of inventions like photography made producing images easier and I'm sure a lot of portrait painters lost their jobs to photographers. I think the danger is that we may see a very rapid erosion of jobs in the creative space that won't be easy to transition into new fields which I feel will have a detrimental effect on our society.
The art gatekeepers want to try and paint everyone using AI creatively as monkeys mashing ignorantly on keyboards, taking the unmodified output and parading it around like Picasso. The reality is that is to AI art as bathroom dick graffiti is to manual art.
When the day comes that the author can fulfill his dream that will not be a good day. And AI videos will be even more dangerous then.
(I think that's all intrinsic to genAI in a way the author doesn't want to confront, but there is a crisis)
Put down the phone, close that youtube, tik tok or whatever tab.
Don't try to explain to your relatives which video is AI and which is not, explain to them that everything on the internet is bullshit and trolling, AI or not it does not matter it's all bullshit to sell ads.
Tell them they are better off watching TV, even Fox news.
When I hear talk about AI risks I mostly hear things like runaway super intelligence doing whatever it wants and leaving humanity in the lurch. But what about more realistic concerns, like accelerating the race to the bottom by cheaply and poorly ripping off other people’s work and forcing everyone else to do the same just to keep up?
But i've seen several good videos made using AI, including pretty much everything NeuralViz[0] on YouTube makes, but also some that have been posted in older comments here in HN. Igorrr's ADHD music video[1] was also made using AI and fit the music perfectly.
The common aspect with these "good" uses though is that they do not let the AI do 99% of the job (as mentioned in another comment) but they still involve editing, writing/scripting, acting (NeuralViz for example uses his webcam to act both the motion and voice in his videos and uses AI to change them) and to some extent leaning into the "weirdness" that AI videos have instead of ignoring it.
[0] https://www.youtube.com/@NeuralViz (i high recommend watching them in upload order because they all build into the same "universe" and often make references to older videos)
[1] https://www.youtube.com/watch?v=TGIvO4eh190
AI has given me the opportunity to do it. I just have to keep prompting to get the exact feeling and I can share my idea.
I don't like these curmudgeonly takes.. I'm sure this "All AI Videos Are Harmful" take will age just like how "All videos are harmful" and "All cartoons are harmful" have aged.
Making a hyperbolic statements stating that it is all harmful, because there are deceptive usages of AI is just asinine.
There has to be a middle ground between insurmountable obstacles and "whip your phone out, push button to generate, and send".
We know this because actually existing talented film makers have done entertaining work with AI video. But it's silly to use it as justification to write it off just because you're not talented in the field.
that's quite the definition of 'harmful'
tilt at this windmill all you like, but the genie is out of the bottle and there's nothing you can do about it but kvetch. you can't ban it, you can't regulate it, no more than the US government could rein in cryptography.
The time where this becomes normal and these alarmists become the fringe crazies ranting about the end of civilization cannot come fast enough.
it always rubs me the wrong way when people infantilize the masses. The "vulnerable" masses already already partake in lots of harmful substances and practices (tobacco, alcohol, drugs, gambling/lotto), AI videos are just another potential pitfall people will need to learn to be wary of.
https://youtu.be/SGJC4Hnz3m0?si=J_8gsXEx4-cL3Mc9
It's not just a pedantic issue. Making these kinds of extremely bold and all covering claims is par for the course for those who want to find and exploit attention from others for profit. If that's not what the author is doing I think the most charitable interpretation I can come up with is that they just have no experience with AI videos and are going off sensational press stories.
They should buy an nvidia 3060 12GB ($200~ second hand) and try out wan2.2 on their home computer. It's really fun to be able to make these kind of videos on a whim. And they'll get some real lived experience with the subject.
Good. Its beneficial, I'd think. Lies are not new, the amount of lies, probably not new either.
I don't think we have (yet) reached a level of badness compared to the prior point of tobacco companies telling people around the world that cigarettes have medical benefits.