Are you as full of loathing for all the HN people saying “just integrate the AI into your practice” as I am? I like drawing stuff and I love that this pays my bills and I really have zero interest in becoming a “prompt engineer” instead.
I think a strategy of creating an image of AI art as cheap and tacky is useful. For how long, I don’t know. If we’re lucky then it’ll turn out that getting these things out of the domain of creepy claw hands is a lot harder than anyone thinks it will be, and there will be a lot of obvious tells for a long time.
It may also be useful to try and get your professional associations to bring some suits against these things for playing fast and loose with “fair use”. The legality of these things is debatable, and so’s where “fair use” should fall - personally I feel like nobody posting work on the internet anticipated their stuff being scraped and fed into a giant neural net designed to take their job. Make this shit a lot pricier by demanding that it be built on art explicitly licensed for machine training.
Figuring out where the intersection of “art you like to make” and “art the AI sucks at” lies is worthwhile. If your passion is realistic painterly work then you’re fucked, what do you enjoy doing that isn’t that?
> I like drawing stuff and I love that this pays my bills and I really have zero interest in becoming a “prompt engineer” instead.
I'm not sure how to say this in a non-dickish way, but most of us have to do stuff we don't particularly enjoy in order to make a living. I get that it sucks from your vantage point, but like... I'm sure there were stablehands who loved horses and had no interest in becoming mechanics. Shit happens.
I'm a writer and equally in danger of automated replacement, but so it goes. Maybe I'll end up curating and editing GPT7 output instead of drafting everything myself — I can live with that. It's not like my paid work was written for fun in the first place.
The art I make for my own satisfaction, I expect to continue crafting word by word, and I think people will continue to appreciate that painstaking expression of the human spirit. Just not in a commercial context.
The problem with your example is that it assumes there is a place for him to go to. Stable hands becoming mechanics. That is the key concern with automation replacing jobs is that it is a zero sum game.
The danger is less with automation and more with the speed of automation.
It's hard to retrain. When the change happens between generations there's a natural progression with parent: stable hand, child: mechanic.
When it happens 1-2-5 times in a person's career, then those are some really painful resets.
Sure, some will swim. But the social contract with democracy isn't that "some" swim, it's that "almost everyone swims" (which we generally call "everybody", but that's a separate discussion).
Long term and across society automation is absolutely positive sum. But short-to-long term for those on the short end of the stick it's clearly not.
> That is the key concern with automation replacing jobs is that it is a zero sum game.
it's not zero sum, because the resources that used to be spent paying you now could be spent on something else. This increase in efficiency means more goods/services could be produced!
Of course, that newly saved money would be spent on someone else, instead of the person being made redundant by automation. It is thus a societal responsibility to retrain/reskill that person, and the training perhaps also partially be paid for by the entity benefiting directly from the automation.
I hear this a lot, that “automation creates jobs” chestnut.
Whenever people try to support this argument, they either fall into the “not thinking quantitatively” fallacy by pointing out that “sure, 25 writers got laid off, but the company used to have no programmers, and now they have 2!” Or they fall into the “not explaining the mechanism” fallacy by giving examples from the past where “stablehands retrained as mechanics” without examining why that happened, and what aspects make 20th century mechanization and 21st century automation completely different in kind.
> without examining why that happened, and what aspects make 20th century mechanization and 21st century automation completely different in kind.
So, what aspects are completely different in your opinion? I don't see any fundamental differences. People's wants are unlimited, so the workers no longer necessary due to automation can move to fields where automation can't be applied. There will always be plenty of those because, again, unlimited wants. I mean, in Japan, people are already paying for artificial friends, who will go out for dinner with them. I think we'll have much more work in human contact and companionship in the future. Hopefully, much more doctors per patients and much more teachers per student than today. Etc.
Part of the reason I think 21st century automation is different from 20th century mechanization is that when you mechanize jobs, that requires physical machines, and the number of factory workers and mechanics needed to build/maintain those machines scales linear-ish-ly with the number of machines needed. And so mechanized farming got rid of a lot of farming jobs, but also allowed more farms to be built, which required more and more machines, and thus created manufacturing/mechanic jobs.
Software jobs don’t scale like manufacturing jobs. If you automate a particular profession’s work, they’ll lose their jobs, but meanwhile maybe now 10x more customers can afford the service they provide. But now, scaling the automating software to serve 10x the customers won’t require 10x as many programmers, maybe just a couple good devops people to tend the flock.
You might then argue that customers will still find things to want from the newly freed up labor, but I also think we’re going to start being hard-pressed for categories of work where “the human touch” makes a difference, and also where there is room to employ large sectors of the workforce. Add in that customers “new demands” won’t necessarily correlate with the need for human workers, and it becomes easier to see a world where people keep coming up with new wants, and automation keeps swooping in and immediately satisfying the demand.
How many people work on farms as a percentage of the population now as they did 100 years ago? We still eat, but now almost no one farms. Most everyone is still employed. We’ll figure something out.
Ok but that's exactly the problem, the situation we've "figured out" is kinda trash. A bunch of rich people own all the farms and we gave everyone bullshit admin jobs because we can't admit to ourselves that almost nobody farms and we still all eat and maybe that means a good portion of the population could just chill out.
What? I'm not comparing the two. The point is that we don't need to do the admin jobs, and we need fewer people to farm the same amount of food, so we can just split the extra free time and maybe some people don't even need to have a job.
> that means a good portion of the population could just chill out.
so you're asking those people who farm (using huge machines and automation to do so extremely efficiently) to give you free food? Why would they want to do that?
The people producing the food get paid.
A third party might pay for the feeding of those who cannot pay.
Or money accumulated might get redistributed to enable everyone to afford it.
This doesn't always happen mind you. But it shows that your (and my) economic instincts are piss poor. We can't just assume what the future holds.
-------
Another few examples:
3D animation has killed a lot of tweening artists (the people who smoothed out animations for Disney Films like 101 Dalmations or Sleeping Beauty). But 3d animation created the need of modelers, texture artists, riggers, and more.
Automatic Drum machines didn't kill drummers either, but allowed for more music to be made in general.
The only thing that automation "killed" recently was Lawyers, as online webpages that auto-generated common forms removed a ton of jobs that Lawyers used to do. IIRC, Lawyers are in somewhat of a decline because of this.
So it just goes to show that no one really can predict these things.
> 3D animation has killed a lot of tweening artists (the people who smoothed out animations for Disney Films like 101 Dalmations or Sleeping Beauty). But 3d animation created the need of modelers, texture artists, riggers, and more.
together with the previous comment on mechanics and stablehands, i wonder how many people in history grew up training and wanting to be a mechanic, only to be told, sorry we just don't have that many automobiles in the world, we need more stablehands. i wonder how many people went to do fine art in school only to be told 'look you have good skills and all but we just need more tweening artists'.
This really resonates with me. I used to enjoy working as a 35mm film projectionist. Now that everything is digital the opportunities for making money doing it have basically disappeared. Am I supposed to be excited that now I can just push a button and focus more of my attention on selling popcorn?
Hey, I did that too in high school and college! I think I must have been one of the very last ones (late aughts). What a cool and fascinating piece of machinery the "modern" (now obsolete) three platter tree and "brain" system is, isn't it?
I still remember splicing together There Will Be Blood and enjoying looking at some of the beatiful still frames.
My boss at the time spliced together Quentin Tarantino's Grindhouse, and somehow fucked up and spliced an entire reel out of order... and nobody ever noticed (because Grindhouse was full of intentional fake projectionist mistakes, being an homage to an old era). He ended up noticing on his own on the second day of showings and fixing it overnight!
What's the difference between a human artist walking around in the world, seeing images and videos (many of which are under copyright) and then using those images as inspiration, versus an AI being fed millions of images and videos as part of a large training set?
I am not a lawyer but I don't really see the "it's not fair use" argument.
> What's the difference between a human artist walking around in the world, seeing images and videos (many of which are under copyright) and then using those images as inspiration, versus an AI being fed millions of images and videos as part of a large training set?
What's the difference between retelling someone's story orally, and using a printing press to make an exact replica?
What's the difference between writing down a conversation from memory, and recording that conversation?
What's the difference between making a nude painting of someone, and taking a nude photograph?
What's the difference between going out in public among human beings, and going out in public where there's a facial recognition camera on every corner feeding everyone's movements into a centralized database?
What's the difference between the grandma who knows everyone in the village, vs the social media company that knows everyone in the world?
Social customs evolved in a context of fundamental, sharp limits to human cognition and skill. When technology smashes through those limits, those social customs don't work anymore. Copyright law didn't exist in a world where you couldn't copy things mechanically. If you apply the old rules naively, you end up with a nasty world that nobody wants. So we have to invent new rules to limit how people use the new technology, otherwise people get exploited and life becomes intolerable.
"What does copyright law say about this" doesn't even make sense as a question. Copyright was invented before AI was. The question should be, "what would a society someone would want to live in, say about this?"
> What's the difference between retelling someone's story orally, and using a printing press to make an exact replica?
But that's not what's happening. You can't "crack open" a model to find the bitmap of a piece of training data. It's not there any more than a painting you've seen is "in" your brain. A model sometimes creates things that look similar to existing pieces if it got a ton of copies of the same image, for the same reason that most people's artistic rendition of a tree is going to be more accurate than their rendition of an anteater.
The entire AI art debate is a symptom of the fact that society is structured such that labor-saving technological advancements can harm more people than they help, and it's baffling to see otherwise intelligent and ideologically similar people fixating on this single, relatively minor technology rather than the much more consequential, broader issue of the fact that automation always moves wealth into the hands of a smaller and smaller group of people.
That was a fact during the industrial revolution and it’s quite possible it will happen now with the digital revolution.
The average height of Englishmen actually went down during the industrial rev. presumably due to malnutrition. New histories of the period think of it as potentially as one of the few times we know where there was an enormous economic boom alongside a general collapse of quality of life.
>But that's not what's happening. You can't "crack open" a model to find the bitmap of a piece of training data. It's not there any more than a painting you've seen is "in" your brain. A model sometimes creates things that look similar to existing pieces if it got a ton of copies of the same image, for the same reason that most people's artistic rendition of a tree is going to be more accurate than their rendition of an anteater.
In other words, a technology that has never existed before and doesn't follow any of the categories and schemas we dreamt up to organize the world. So we need to make new rules.
For humans the inspiration can be something else than the medium to use. Watching reality can provide inspiration for creating a video. And I'm pretty sure that's wher most inspiration comes from. It cannot for AI. Of course in a world where everything exists AI doesn't need inspiration from other sources.
Then what was it saying? It said that humans can take inspiration from outside art, from the real world. You can do the same thing with AI by training it on pictures from the real world.
Always back to the same debate then.
Why aren't human minds glorified functions, that perhaps deserve their glory a bit more than current state of AI?
Obviously they're not currently equivalent. But obviously there are similarities.
If I can take copyrighted images and feed them into an AI, why can’t I take any code I find online and ignore the license terms and feed it into a compiler? How do you justify the existence of something like the GPL?
Is it ok to scan copyrighted works. I think Google Books' win at the Supreme Court shows that yes it is.
Is it ok to process them down to the rawest information?
Is it ok for people to generate content from the raw information (and ok for them to charge for it)?
The last two I don't have an answer for, but I understand the fear of artists and the anger that their hard work is being shovelled into the monster that is going affect their work. But also more broadly the make-up industry, the lighting industry, studio spaces, lens makers, paint manufacturers, etc etc etc.
> why can’t I take any code I find online and ignore the license terms and feed it into a compiler?
a compiler is not transformative. You translating a book into another language doesn't make it a new work.
But an AI that takes billions of images, and uses it to synthesize something different and new, is fairly transformative under my eyes, and deserves new copyright. Unless the AI generated image is largely composed of a small number of works, i don't see why anyone should have copyright ownership of such an output!
its actually exactly the same thing, the ai doesnt "synthetize" anything, its following its programming and data based on training sets
if anything I would argue that human taking code from multiple places and making it compile into something more useful is much more synthetiz-ing than AI...
What's the difference between a programmer looking at lots of open source code and github copilot looking at open source code? (And I'm talking MIT licensed code here).
this reads a lot like what some painters said when photography hit, and they weren't interested in "painting with the light" as so many photographer would love calling it nowadays. Technology disrupts and it is neutral to taste. Seems like those who knew the way should probably try and instil better taste for the future generation; otherwise we'd be reinventing wheels with different tools for the foreseeable future.
Questioning legality of the A.I. models won't save you, and is likely to backfire (I'm not saying this is good or fair, but rather prepare to be screwed either way.)
In the worst case corporations will just rebuild the A.I. models from data licensed from stock photo sites, getty, etc. All you can get from this is a one-time bulk-discount payment, and then you will be paying them for use of the model in the next Photoshop.
Plus there are copyright-abusing corporations that jump on any excuse to kill off fair use and scraping, so any attempt to legislate scraping may instead end up creating more Disney copyright and link-tax laws.
If that's how you feel, then I assume you won't be adapting at all (as evidenced by your other comment [0]), so your 100k job will go too to the ones who do adapt. I'm not saying this to be mean, you asked the question, people answered and it seems you don't like the answer. What are we supposed to do about how you feel?
It’s not about how I feel - it’s about the potential obliteration of my way of life. The people who will make money out of this transition will be AÍ engineers, not image makers. I will bet you 1 ETH and place it under a smart contract bet if you’d like.
Then perhaps you'll have to do what the vast, vast majority of humanity has done for thousands of years, work jobs they don't like to put food on the table. Working in art is a luxury reserved for only the highest echelon of people, like Michaelangelo or Leonardo. The rest of us are not so lucky.
Lol. So I work hard and scrounge for 10 years and get to work for Apple eventually after fighting for life and now “I’ll have to work jobs to put food on the table”
You’re crazy if you think many artists I know don’t outwork 99% of the population. Condescending and inappropriate. Many artists I know are the most hard working people I’ve ever met - because making it in this area is fucking hard. And a lot of the laziest people I know are well paid data scientists. I don’t blame either of them. But your comment is just a lazy ridiculous preposterous idea. Read the biography of Patti Smith and tell me she didn’t work hard. Such a presumptuous comment.
I don't see anything about laziness in OP's comment. Any one who knows anything about "creative" fields knows that people in them usually work very hard and often for not that much money, compared to people in other high-skill fields.
BTW I'm sorry you might be getting disrupted. It truly sucks.
You worked hard, so what? That doesn't mean anyone has to pay you. I could roll a boulder up a hill all day, that doesn't mean anyone would find value in what I'm doing.
And if you think the vast majority of the human population in history, subsistence farmers, didn't work harder than virtually every artist in existence, I'm not sure what to tell you. Being an artist for a living is absolutely a privileged position for 99.999% of the world's population.
I beg to differ. Art isnt just fancy paintings for pseudo intellectuals to show off. Art is a good looking app, a nice looking car, a cozy feeing house design or kitchen furniture, an ad that communicates vision, and so on. Art is emotion which ai cant create. Sure it can randomise visuals within the constraints of what it has “learned” but it definitely cant create human emotion. No amount of micro-dosing can convince the world it is going to completely replace artists or indeed programmers.
nor does the art that humans create - the emotion comes from the human viewing/experiencing a work. The "source" that work is perhaps irrelevant, if the human viewing it is being evoked.
Personally i find ai created content utter shite. Even copilot’s autocomplete feels like a toy, let alone that it needs to be fed with new content all the time in order to stay up to date. If not where will it steal its dataset from? Itself?
Moving somewhere less expensive than London and doing your current work remotely might be an option, everything’s become less focused on going into the studio to work after the ‘rona. Lower your expenses and start playing with personal work, your client history can probably get you some interesting gallery opportunities to give you a decent chance at selling Authentically Human-Made work to rich folks at an absurd price.
I ditched your kind of career path a while back, I’m using the drawing skills learnt chasing the Hollywood animation dream to draw indy comics and furry porn. Makes less money but I can eke out a living in a much cheaper town than Los Angeles, and I pretty much draw what I want.
Have you seen the demo of Stable Diffusion integrated into Photoshop[1]? It seems like a very useful tool for an artist to have in their toolkit. And I think professional skills will still be in demand to blend everything together and create portions of the image from scratch - for many years to come.
I think this is what people have in mind when they talk about integration. It doesn't seem ridiculous to me.
I work in CGI/design. We have had machine learning denoisers for some time but other than that ai has not changed cgi much. VR another hype, merely a distraction- the serious work is not VR. Now things like stable diffusion- lots of people toying around- serious tool for artists- not really. Do not worry. Develop your artistic skills and enjoy the journey.
For each technological advancement we had in the past 200 years had made it easier to do things at a certain level. Which enabled people to do more things and more complex things. Which just pushed up the average level of what we create higher.
Can't AI be used to handle what the artists had to spent more effort to create before, therefore enabling artists to create more complex works of art and therefore push the creative level up?
Are you as full of loathing for all the HN people saying “just integrate the AI into your practice” as I am? I like drawing stuff and I love that this pays my bills and I really have zero interest in becoming a “prompt engineer” instead.
I've been using these ai tools to help imagine some paintings. Previously I'd use image search.
If you are drawing completely new images then I'm not sure how competitive these tools are. They're just serving a mash up of existing works. Sure they can produce something interesting, sometimes stunning, but usually not much better than what you could just find already.
As for producing a physical painting, that has its own constraints of material and ability.
I guess the issue is that for lot of commercial work, "good enough, cheaper" is more attractive than "better than most, at the right price". Of course that's not a rule always, everywhere, but at scale it tends to be.
I agree there'll be a market for the ai stuff. It'll have to improve though. For example, say you get an illustration done and you say "yeah it's good but can you just modify it a bit and then I want 5 variations doing different actions with the same characters". The AI tools can't do that, they'll come up with completely different images for very minor prompt changes. And if the input set doesn't contain the desired action then they cannot generate it at all.
For now I think it'll only replace some low level commercial stuff and perhaps add images where there were none before.
> It may also be useful to try and get your professional associations to bring some suits against these things for playing fast and loose with “fair use”.
I understand your livelihood is being threatened, and that’s very scary. Can we please avoid using the legal system to kill nascent competitors? Surely there are better responses than this.
Sure, I’ll just politely roll over rather than use the legal system. I don’t need to put food on my table anyway, it’s just a luxury, I have infinite time and money available to go learn to do something that pays more and makes me absolutely miserable. I’m not already constantly fighting the way all the social media sites have stolen the ability to serve ads alongside my work by building systems that suppress any post with an outgoing link, as well as posts with words like “commission” or “Patreon”. I can just bend over and take it a little harder when all these image generators based on megacorps scraping the entire internet for a use nobody thought of when defining “free use” make it harder for me to earn a living, and more importantly when they make it a lot harder for a beginner artist to get to that happy place where they get to spend the majority of their workday drawing, and enjoying the feedback loop of getting a little better and faster at drawing a thing every time they do it.
>Are you as full of loathing for all the HN people saying “just integrate the AI into your practice” as I am? I like drawing stuff and I love that this pays my bills and I really have zero interest in becoming a “prompt engineer” instead.
I wish it were that simple. As a person who genuinely enjoys being a data scientist, and practices all those skills in a project I create on my own: the industry says otherwise. Integrating a technology into your practice that takes nuance out of it, just really means that managers and c-suites are going to expect you to give more nuance to your choices (which they deadlocked you out of). ML DevOps for the last two years are just me integrating more services.
Are you as full of loathing for all the HN people saying “just integrate the AI into your practice” as I am? I like drawing stuff and I love that this pays my bills and I really have zero interest in becoming a “prompt engineer” instead.
I think a strategy of creating an image of AI art as cheap and tacky is useful. For how long, I don’t know. If we’re lucky then it’ll turn out that getting these things out of the domain of creepy claw hands is a lot harder than anyone thinks it will be, and there will be a lot of obvious tells for a long time.
It may also be useful to try and get your professional associations to bring some suits against these things for playing fast and loose with “fair use”. The legality of these things is debatable, and so’s where “fair use” should fall - personally I feel like nobody posting work on the internet anticipated their stuff being scraped and fed into a giant neural net designed to take their job. Make this shit a lot pricier by demanding that it be built on art explicitly licensed for machine training.
Figuring out where the intersection of “art you like to make” and “art the AI sucks at” lies is worthwhile. If your passion is realistic painterly work then you’re fucked, what do you enjoy doing that isn’t that?