If your "art" is actually threatened by AI, I question whether you are doing art at all. Art is not going anywhere. AI will be another tool in the toolbox, but artistry and artistic sensibility is here to stay. AI tools can create pictures, but actually creating art is still a human endeavor.
From orchestras to video games, AI is going to change the way that art is done (and it already has), but the artists are not going away.
I think it is more accurate to say that AI is threatening a lot of the ways artists use their skills to create economic value. Making corporate logos, creating visuals for products, etc.
Art won’t go away, just a lot of the ways artists support themselves.
I personally highly doubt that any company worth $100 million or more will have an AI-generated logo in the next 30 years.
Product visual mockups also aren't art. The product design itself is art, but I sort of suspect that a human designer will be the one driving the AI to make the visual.
The artists at the firms making the logos are absolutely going to be using AI to come up with ideas to pitch to their creative director (with some human tune up).
The trouble being that replacing drafts people creating options for the Art Director with AI creating options instead is that it removes a central path by which artists gained skill and experience in their industries (and the same is true across all disciplines). All will go well so long as people who came up the old ways are the ones using AI. Eventually they’ll be gone and all the experience and discernment will go with them.
To put it in more immediate terms: the goal of business is profit. Cutting costs is a great way to do that immediately. It’s only over a longer span that things like Boeing reveal the issues. The goal of artists and the art industry is to create art that moves other people, at acceptable costs. Cutting costs with AI is a great way to do that immediately…
You see where I’m going. Sure maybe we’ll have figured out an entirely new economic paradigm by the time this bill comes due. But I wouldn’t count on it.
And hey, maybe you think art isn’t a career path worth any sort of consideration on this front, and that society will keep humming along just fine when AI takes over in that regard. I sure hope so, because that seems like the obvious conclusion of all current incentive structures.
An awful lot of the economy is made up of companies much smaller than $100M. If I walk down the street, I see all sorts of small shops & restaurants. Cars and trucks go by with logos advertising local plumbing services, nursing home care and catering. Yesterday I took a tour of a brass foundry which forges brass sculptures that hang on the walls of churches. How big is the company? Its one guy. One guy and a small warehouse of equipment.
I think a lot of incidental work is already increasingly being handed over to AIs. Business cards. Images for the wall of a cafeteria. The logo for a bakery. Whatever. Photoshop is full of AI assistant tools. So is Divinci Resolve now.
AI is slowly nibbling away at creative work from every angle it can find, to reduce cost and automate. I think we should expect this process to continue for quite some time.
The problem is that AI needs good training data to work well. However, everyone knows that training AI on AI generated content has poor results. Therefore, the more market share AI generated art commands, the worse it will perform. It is inherently non-sustainable.
If 'the artists are not going away' involves either paying the artists less or reducing the number of artists or the meaning of being an artist being redefined I would not immediately agree. ... but the third one is doing a lot of lifting.
I think you will find that neither will be the case. There have always been a vanishingly small number of artists who can make money on art, and that pool (and their value) will not be decreasing.
Much of that pool is doing live performance art, like dancers and musicians, and when Napster and Spotify came to "destroy" the musician by giving everyone infinite reach, what actually happened was that more people succeeded as professionals.
The photograph was also a tremendously good thing for visual artists, bringing in the impressionist era and opening up a whole new field of art in itself.
As to the concept of changing the nature of the work of the "artist" - I can't think of an artist whose work wasn't changed by the technology of the time. From Prussian blue to the electric guitar, artists are usually made great by showing the new possibilities of the new technology of the age.
There's some nuance here. Industries such as stock photography and audiobook narration are indeed artistic forms but can mostly be replaced with AI equivalents.
I have not yet met an AI system that can get close to what an audiobook narrator does. I assume that more audiobooks will get lower-quality narration, but narration of audiobooks is often done either (1) by the author or (2) by someone who is very good at doing a dramatic reading. The former won't be replaced because the author's own reading of their work provides emphasis and nuance that shows their intentions, and the latter is still far beyond the capabilities of current text-to-speech.
AI is very good at giving you an average-quality output. Art, by its very nature, is almost always a superlative expression of something.
> Human creativity and the good it offers to society cannot be replicated by a machine. No matter how advanced AI systems become, it is not likely that they will ever be able to think like a person.
The story of computers since their inception has been a series of statements of this form being proven wrong.
Up until not very long ago telling a random person you worked in AI reliably produced a very skeptic response or a look of disbelief. (In my experience.)
We are naturally human centric in our thinking. But the rest of this century is going to require we face & work with reality, not against it.
Making every advance in AI as usable for as many people as possible, and in service of their own initiatives, instead of replacing and channeling people is important.
There is no reason “post-human” needs to mean the end of humans, as apposed to the future of humans.
Unless we keep settling for tech that dehumanizes us. In which case, we bifurcate and the human race suffers from a very “hard landing”.
Ethics, positive sums, transparency, on ramps, that all means more now than it ever has.
If you think about how we've handled climate change, then we're fucked, because what I'm reading from you is it's going to take.
1) Co-ordination on a massive scale.
2) Honesty and integrity on all political fronts, including between hostile nations.
3) Uncompromising forward thinking and planning.
Yes I know I'm negative, but people who expect some type of solution to come in the form of organization, are deluded IMO.
The only saving grace for me is that we lack a lot of imagination about what the future will actually look like and so the AI takeover scenarios, and worthless humans hypothesis are not really what happens, we don't really know what will happen.
No. If things get really bad, pretty much everyone will want to fix them. And as far as temperatures go, it's not even that expensive to fix - block/reflect some of the sun at scale. Cover some deserts in cheap mirror material is probably the cheapest way to do this. USA alone can afford to do that.
The expensive part is taking CO2 out of the atmosphere. But if the temperatures are back to normal (whatever "normal" we choose), then CO2 is less of an issue.
No. If things get really bad, pretty much everyone will want to fix them. And as far as temperatures go, it's not even that expensive to fix - block/reflect some of the sun at scale. Cover some deserts in cheap mirror material is probably the cheapest way to do this. USA alone can afford to do that.
Things are really bad for some people already, does the things you describe start to happen when they get bad for you specifically ?
How quickly do you think we can cool the Earth off to stop crop failure issues from propagating even faster, assuming we've not already trigger some pretty decent feedback loops?
> The makers of these tools promise they will boost efficiency and productivity by making workflows easier and faster. The problem with this message is that art is not about efficiency, productivity, or smooth workflows. Art is not about “executing” a task. Art is certainly not about the metrics of quantity, low cost, and speed that characterize industrialism.
This here is really the crux of the issue.
Making art as an expression of yourself isn't about these things.
Making assets for games, etc is more about about the metrics of quantity, low cost, and speed that characterize industrialism
Having the systems for the second, doesn't remove the first from existence.
Disagree. Unless you're going for absolute realism, making assets for a game is 100% under the normal realm of self expression based art. The piece being a part of a larger whole doesn't change that any more than an art piece being a part of a gallery does.
He is referring to the difference between having corporate motivators for art vs. pure expressionism, not saying that there is zero artistic expression in corporate settings.
Why yes, my self expression was to remove these exact 5000 triangles on my model so that it has the exact amount of fps that I expected in this cut scene on a computer with the following specs…
Said no one ever. Most game assets are not “self expression based art”
Give an artist unlimited freedom and they will often produce drivel and sometimes produce masterpieces, even if with a few constraints they could have produced interesting work
Give a corporation unlimited freedom and they will inevitably create dystopia, even if with a few constraints they could be a great boon to society
A counterpoint to offer to this position beyond the legitimate concerns a new technology like this is having... if it was decided to re-imagine shooting the first 3 Star Wars films created, we probably would use today's sci-fi abilities.
Of course, this doesn't cover the gamut of artistic skills and abilities.
I'm hoping this means if it makes average people into a designer or programmer, a designer or programmer with AI will be 10-20x as effective.
Learning, reskilling, upskilling, may be a reality in more areas.
The question does move towards agency, and who is driving the content creation - experienced artists, designers, etc seem best positioned to describe creative descriptions and explore artistic imagination in these new ways.
Human created art may be worth even more too.
English is the new programming language, and those people who spent many years learning to describe and discuss nuanced things like art, and the creation of it, are probably better positioned than most.
I find the premise that AI “cheapens art” to be spurious at best. There have been generations upon generations of crummy, unimaginative art. There have been decades of labor undercutting domestic artists.
AI makes it easier for companies to be cheap, but that’s not new. They’ve been fighting to be cheap forever. How many artists of all stripes have been baited by “do it for the exposure”
It doesn’t cheapen imagination. More people can express their ideas. I can’t help but think any company worth its salt is going to have artists on staff anyway - if for no other reason than to provide useful context for the art being generated.
In Germany, there is the term "Gebrauchsgrafik." It refers to visualizations of products or services. These works typically lack artistic value and are often only visually aligned with the zeitgeist. Think of your bog-standard low effort advertising or IKEA instructions, generative AI is more than suitable for that. Not everything is or has to be considered art.
Low effort advertising sure. Diagrams I've had no luck with personally. With current models it's been a very frustrating process to even try. I can generate stuff with maybe one or two key ideas, but instructions are ignored as you try to incorporate more details. There also aren't really any affordances yet for dimensions. Perhaps styling / editing from sketches fixes this, I would be interested to hear from anyone who has managed to get a useful, Ikea style diagram out of gen ai.
To me it feels we are a long way from AI being able to produce useful minimalist (i.e, omitting un-necessary detail) diagrams at better than a surface level of stylistic mimicry.
Producing useful functional diagrams entails some understanding of the physical components and assembly process, along with some model of how people think and what information they need at what step to minimise confusion.
Naturally a lot of this information is embedded in the design of the assembly itself and perhaps the big picture can all be "prompted in" but it's not a domain where "bullshitting it" works well.
We work with flux, custom loras and controlnet and it's working exceptionally well. Using only prompts as conditioning is also not the way to go for higher complex tasks. It's day and night compared to dall-e,imagen, sd1.5 or sdxl.
There is a real concern rapidly developing for artists just like there is for all types of human labor. That is valid. But the idea that AI necessarily reduces the level of art produced by artists seems to be a false conclusion.
AI art DOES make it harder for artists to rest on pretension. But if the actual product of the artist matters, then it should make it easier to create works that are both more aesthetically pleasing AND more conceptual and relevant to society.
The more substance there is to the art, the easier this is to see. The AI can help flesh out the details while the artist focuses on the high level concepts and experience they are trying to communicate.
Like all automation, it does make classist viewpoints weaker. This article seems to be coming from that standpoint.
Sadly for many artists the AI is/will be better than the average artist. I compare it to elite football players. "FIFA estimated that there were 123,694 professional soccer players worldwide." [0]. How many are really elite football players? Probably less than 2200 [22 (11 field + 11 outfield players) * 100 (top countries)]. Top 10% artists probably will get very well paid job but the rest will either get fired or salaries reduced or stagnant.
I'm aware the difference between those two things is mucky, but 'real' art is exclusively a human-to-human communication. You can use AI to augment your work for sure, but even then it won't carry as much emotional weight (read: value) as Guernica or the Ecstasy of Santa Teresa.
> As the photographic industry was the refuge of every would-be painter, every painter too ill-endowed or too lazy to complete his studies, this universal infatuation bore not only the mark of a blindness, an imbecility, but had also the air of a vengeance. I do not believe, or at least I do not wish to believe, in the absolute success of such a brutish conspiracy, in which, as in all others, one finds both fools and knaves; but I am convinced that the ill-applied developments of photography, like all other purely material developments of progress, have contributed much to the impoverishment of the French artistic genius, which is already so scarce. In vain may our modern Fatuity roar, belch forth all the rumbling wind of its rotund stomach, spew out all the undigested sophisms with which recent philosophy has stuffed it from top to bottom; it is nonetheless obvious that this industry, by invading the territories of art, has become art’s most mortal enemy, and that the confusion of their several functions prevents any of them from being properly fulfilled. Poetry and progress are like two ambitious men who hate one another with an instinctive hatred, and when they meet upon the same road, one of them has to give place. If photography is allowed to supplement art in some of its functions, it will soon have supplanted or corrupted it altogether, thanks to the stupidity of the multitude which is its natural ally.
It cheapens humans, period. In the coming years, AI will be good enough to replace X% of a human's productivity. It will be as cheap as duplicating a process in a computer somewhere. It may never get to 100% but it will get to a high level. How many human equivalents will we be able to buy for a buck? AI is not the human savior people wish it would be.
I think this line of thinking is what happens when economies are allowed to calcify by consolidation of power in the hands of too few. I think the degree to which current systems are actually meaningfully transformative is greatly overstated by the tech industry. However, usually a transformative technology changes how economic distribution works. For example, even very wealthy households no longer employ as many human servants, because many of the tasks they once used human labor for are now done by household appliances. If the effect of "AI" is to cheapen human labor in its entirety, then a working economy will simply no longer center around labor, because the purpose of a working economy is to support the activities of humans by coordinating the distribution of resources. The likely alternative to the economy rearranging itself this way isn't most humans simply accepting being deprived of resources, but instead a breakdown of the cooperation necessary for most technology we have to continue functioning, as even if all the necessary infrastructure could run itself, it would also have to defend itself against defectors. Maybe it'll be effective at doing so, which probably means wiping out most humans. Of course, global-scale conflict could also do that, so it seems preferable to redistribute resources
One problem is how fast change is coming. We can adapt to change but we can't at the speed that it is happening. People are talking about giving everyone free money to help with the transition. OK. But that leaves millions of people with more free time that they can healthily use. The quote "an idle mind is the devil's workshop", is absolutely right. Look at people that have lots of free time. See how hard they work towards killing themselves, through dangerous activities. Now imagine a whole society where the majority of members are bored because they have exhausted their entertainment. Yikes! Or just as bad, a population that is plugged into constant entertainment. Neither is good.
So your position is that people who have free time mostly do awful things and your support for this position is like... a misquoted aphorism from the bible ("idle hands are the devil's workshop")? I don't find that particularly compelling at all. People find things to do with their time when they don't have external pressure on them. Most of those things are "useless" but also pretty harmless, and make them happy. Often people spend their free time building community with other people, which can often be valuable to others as well. Sometimes they do stuff like discover laws of physics or write novels. To claim that there is some great evil that comes from merely giving people more leisure time is an extraordinary claim that to me seems to lack even a modicum of evidence. Sure, some people do harmful things and we should stop them from doing those things. To attribute this to people being able to pursue their own interests is not even wrong, it's completely incoherent
There is no absolute proof of how people will behave. People are just too inconsistent from moment to moment. But we can get an idea from past activities. The COVID lockdown was a small view into people's behavior when they have too much idle time. I remember riots, protests, and crimes, to name a few, and that was a small amount of time. Also, when do people start to get into trouble? When they have too much time. It does have to be everyone, not even the majority, for it to be a huge problem.
Even you talk about it as a "useless" use of time. It may be great for the individual at the given moment. A dopamine hit always feels good. But it's not great for society. How is that a plus for the future? Thousands of years of societal evolution just to have people figure out what to stream next. Boy, AI will lead to a great future. Happy and fruitful times are on their way. :-(
AI's future is all speculation. There's no way to know until it happens but I don't see it as a plus.
If we purport that human labor has no economic value, people being useless in their idleness has no bearing on the future. Also, the COVID pandemic was not people having free time, it was people being ordered to stay at home, many people being excluded from this due to being "essential workers", society doing damn near nothing to adjust to this state of affairs and thus keeping people in the "normal", stressful state of having to labor to live, and also in fear of a deadly disease, and often socially isolated and prevented from pursuing many of their interests. This seems like a poor comparison. If your experience of the global pandemic was just "I had a bunch more free time" it seems like you're already of the subset of the population experiencing something closer to the post-scarcity reality that all possible valuable human labor being cheaper to automate could be, and I assume you weren't doing crimes or whatever, right?
On top of that I doubt every country can afford to give everyone free money. US maybe can for their own citizens but will they give money also to e.g. Malaysians? (they currently do a lot of hollywood vfx and postproduction). I highly doubt that and there is only US, China and maybe France in AI race game right now.
Money is an abstraction. If AI can't produce enough value to support humans, then clearly there's stuff humans value that AI can't do. If AI can produce enough value to obsolete human labor, then human labor is unnecessary and money need not represent it
Money will survive. We humans don't just use it for physical goods. Think about all the monetary exchanges where no physical or virtual goods are exchanged. In Japan, you can rent a friend. An example of an out-of-the-ordinary transaction. Yea, you may be able to have an AI but I'm sure there will be stuff it won't be able to fulfill.
I agree that this is a more likely path for the human economy to take, and one that would disproportionately benefit from people being able to operate independently, as if stuff like social companionship become the kinds of things humans value in each other, and money is exchanged on that basis, the existence of a class of people who can be compelled by desperation to that kind of service seems kind of rife for abuse to say the least, and not really fruitful to let people hoard wealth to an indefinite degree.
That said, any future economy in which humans value services from each other (in a monetary sense) is one in which humans have not been outcompeted by AI in all possible value-generating endeavors, so by some people's standards it might negate the premise that "AI has replaced all human labor"
Either way, if people really believe something like "AI is obviating the contributions of humans in a way that will eventually increase to 100%" I think it's kind of hard to justify an equilibrium where no one does anything an AI can't do better but we still have societal losers who can't do anything to better their station and don't have autonomy and security. I think no loser in that scenario is likely to accept it, so we're talking about either killing or enslaving a whole lot of people or doing some "wealth" (whatever that means in a post-scarcity world) redistribution
Our quality of life more than doubled compared to ~50 years ago.
We have powerful computers with amazing screens, constantly connected to a global network that allows you to talk to pretty much anyone on the planet, with video. That allows you to search for all kinds of information near-instantly that previously could take weeks or months of digging through libraries or newspaper archives. Computers that translate in near real time.
We have medicine that solved so many diseases, and is so far ahead of what we had before.
We have robots that do all kinds of boring and annoying and dangerous tasks. I have two robots cleaning and mopping my floors at home. We have whole factories that are mostly run by robot labor.
We have freaking GPS!
Capitalism is not the issue. It's infinitely better than any other system, which most of the time steals labor, forces labor. I grew up in USSR, I know.
I disagree, as I’ve seen lots of genuinely artistic content that was at least partially generated by AI, that gave me certain emotions or made me see the artist’s vision. I do agree it doesn’t “feel” like it should be called “art” without a lot of effort put in or because the person generating it does not have to have the same technical skill. But where do we draw the line? Modern brushes and paints and papers make art easier - should we devalue that art because the artist did not create their own paints from scratch like they did hundreds of years ago?
Humanity is going to have such a delightful crisis when live performance and physical presence become the only finite resource left we have to value. It can't scale, and like a big crunch at the end of heat death, a new universe will scatter.
I doubt I'll live to see it but it's my most optimistic fantasy for the future. If it's delusional, let me at least keep it.
After almost half a century of technology pushing a linear-scaled drip of dopamine-tied-neophilia, the recent bloom of ML has started to chew the flavor bare of this zebra gum and my jaw is concerned there is nothing left to experience if my effort remains at zero.
This is why people who spend their whole weekend on their phone end up with anxiety, our lives are sand timers, we can see the end coming in the future, and when you spend your valuable time consuming bullshit, you look up at the clock, it says 6pm, and you wondered why you squandered your life, and think about what you might've done instead. This thought becomes too much to handle, then you look back down at the device so not to have to face the truth and on it goes until one day, reality bites.
I'm so tired of this argument, people think art is a zero sum game, its not. It's like saying photographs cheapens the art of paintings... they are just different.
Our concern for cost so eclipses our concern for quality that the horrible truth is we'll defer to infinite and ok to finite but great. I am not an artist and 1000 free wedding photos from my friends sounds much better than 30 wedding photos from a professional photographer. AI wrote my vows, my friends took the photos, and AI touched them up. I saved 2500 bucks and that photographer was free to upskill on their next industry.
I have no idea if I am being facetious. I hate all of this, as much as I love it.
The trade-off was that you had to burden your friends to take the photos, spend your own time collecting the images from them after your wedding, research AI services, and oversee how AI touches up your photos.
You traded money for your and others time. Nothing wrong with that, but AI hasn't changed anything except the crowdsourced images look a little better, maybe?
I would rather pay to get the 30 great photos ready to go - save me the hassle. Plus, a photographer will coordinate the key wedding photos with the various family and friend groups and will help find the ideal scene for shots.
Of all jobs, wedding photographers are relatively safe from generative AI.
It's totally zero sum. AI art is far better. Where artists had starving wages they'll soon get nothing. Same for actors and musicians and well everyone else. But the starving artists are somewhat unique. They're vocal because they've had a preview of a profession where you either are the best at what you do or broke for life. That's our future. I want my meals cooked by cheap hygienic ultrafast robots with no wage. Human made food can be a once a month novelty, if that. Every product, service and job will fall to AI in the next decade. History suggests this type of abrupt disruptive change quickly leads to a war. That's why those billionaires are digging bunkers.
I don't see much difference in art created by an artist using an AI, art created by an artist using photoshop, and art created by an artist using red ochre. They are all using tools to express something.
Many artists in fact do exactly this -- they do not make the art themselves, but instead imagine and manage the 'art project' and delegate tasks to other artists that work for them. Leonardo (Da Vinci) in fact himself did this with much of the work required for his paintings, although the most crucial parts he left for himself.
Was your only input to ask some guy? When I play with SDXL, I can spend an hour or more on a single image, between prompting and inpainting, and playing with dozens of variables. Each final work is a collage made by generating each detail of the image separately many times. And then I would usually process that final image in some other software to get the final result I want.
Photography and cinema are art forms, even if your phone has a camera. So is painting, even if a baby can fingerpaint. You can ask ChatGPT to give you an image of a cat, but you can also use AI as a tool to express yourself in personal and unique ways.
I honestly don't want to be rude or attack anyone's opinions, but I really think one can only believe generative images can't have artistic value due to ignorance.
Yea, it's pretty silly that this is the case in our culture, and is a mark against using the paradigm of property rights to manage acknowledging the contributions of creatives
From orchestras to video games, AI is going to change the way that art is done (and it already has), but the artists are not going away.