Goodbye cruel world, it’s been a ride. The time has come for me to give up and settle into a comfy non-existence, devoting my remaining time on this godforsaken rock to playing Gran Turismo 7 all day without getting dressed and tipping fatty snacks into my snark hole. Another tortured genius crushed by the inexorable march of progress. Creative skills have traditionally been earned and learned the hard way, but unfortunately for Silicon Valley, they’ve been the purview of us troublesome meat sacks with their demands for secure employment, healthcare and fair wages.
Instead of that: “Can’t we just get A.I. to do it?” croaked Elon through a haze of sweet-smelling smoke, probably.
Yep, it’s time for another rant. But before we get into the specifics of how A.I. will or won’t affect car design let’s take a look at the broader trends of what’s going on. The tech industry, flush with cash for years and hungry for The Next Big Thing, didn’t look at a single industry they thought couldn’t be improved by their disruption and insight.
Sometimes it’s succeeded. Digitization of information has revolutionized how we consume popular media; no more Friday night trips to Blockbuster Video. We all carry a personalized, powerful and configurable personal assistant and communication device with us all the time. In others, they’ve failed. Self-driving cars. Battery-swappable EVs. The Internet of Things (dammit, refrigerator, I keep telling you I only drink San Pellegrino!)
I’m not talking about specific formats but broader ideas and the arrogance that they could take something and make it better by the brute-force application of technology and turning it into a subscription-based service. Juicero is my prosecutorial star witness in how to spunk $120 million of other people’s money in the pursuit of solving a non-problem.
Since the Industrial Revolution, advances in technology have improved the processes used to build things. Line workers today can reasonably expect not to clock off by being mangled inside a huge piece of industrial equipment. Robots now carry out repetitive and dangerous tasks to a high degree of quality, which is what machines excel at.
But there is a limit to how much assembly of a car can be automated, as Tesla has found out— for some tasks you still need humans. This means it’s hard to save more money by automating further. For most products (with a few significant exceptions) the profit is made at the beginning of the process in the design, and at the end in the marketing and sales. The bits that require creative minds in other words. Where there’s money to be made, there’s money to be saved, so no wonder this is where the tech companies are setting their sights next.
Society in general has always had trouble seeing the value in the creative industries, even though the output is what makes life interesting—and gives choice to us consumers. I always tell my students never to work for free; an internship should be paid, and if a company values your work enough to use it, they should compensate you for it. Exposure doesn’t pay the bills.
This is why the design of a lot of automotive start-ups and custom or accessory suppliers is sometimes shocking. They don’t want to pay the going rate of contract designers and modelers because of the expense. Like good lawyers, we bill by the hour, and you won’t always like what we have to say. If the design process can be distilled down to an algorithm, you’re only paying for a few hours on a CPU farm. Much cheaper, and it should do what your outsized CEO ego tells it to do.
The race towards autonomous driving has driven a lot of the machine learning that underlies AI image creation. The two most common generative art AIs, Midjourney and Stable Diffusion have been trained on the open source LAION-5B dataset, an index of nearly 6 billion image/descriptive text data pairs on the internet, with all the copyright and privacy issues such a massive image fishing trip entails. The AI recompiles existing images into something new based on words you provided as influences. The buzz goes anyone with no drawing skills whatsoever should, with the correct sequence of inputs be able to produce a compelling studio-quality render.
So where does this net out in the car world? This week, to celebrate the 30th anniversary of the Twingo, Renault announced it would build a physical show car based on AI-created images submitted by the community.
If I was on the exterior team at Renault, you’d better believe I would be pissed about this. There would be some flouncing out of the studio and perhaps some door slamming. I’m going to quote from the press release directly, because it says a lot:
“Once solely available for the elites in tech, artificial intelligence is now open to everyone in just three clicks. This newly accessible technology enables everyone to unlock the power of their imagination. We wanted to pay homage to the iconic Twingo by giving everyone the opportunity to take part in the creative process. We are going to present the first show car produced using artificial and human intelligence, directly generated from the designs that we will receive,” – Arnaud Belloni, Global Chief Marketing Officer. (Emphasis mine)
Silicon Valley will save us, because fuck me! Without an app, us Morlocks couldn’t dress ourselves or make coffee in the morning. Remind me again why I spent years studying for a profession that only the most talented get accepted into?
Automotive AI is an Instagram account we’ve covered recently that regularly posts some fun thought exercises: What if Walter Gropius designed a car in the 1930s? Or, imagine if the Smart had actually been in production for 70 years.
While these are a fun diversion, they rely on taking the familiar and transposing it into something else. Remixing X & Y and crowbarring them together in an unholy bastardization that’s the automotive equivalent of “What if all the Harry Potter characters were in Star Wars.”
It’s the type of low-effort concept shit you can find all over DeviantArt, except that at least was created by actual humans putting effort into what they do. There’s nothing particularly wrong with this, but let’s not lose sight of what we’re looking at. This is not design, because it doesn’t have any context or external drivers beyond wish fulfillment. Car design is an iterative and time-consuming process, not just bashing out a few fancy images.
My social media feeds have been stuffed to bursting with AI hot takes recently. So, with David threatening to repeatedly run me over with a BMW iX press car I figured it was high time I entered the room under his frozen garage floor and fired up the Autopian MU-TH-UR 6000 mainframe to give Midjourney and Stable Diffusion a whirl.
Because this is highly scientific, journalistically copper-bottomed and not thrown together by the seat of my pants, it struck me the best way to illustrate what a designer actually does compared to what an AI does would be to use one of my own projects as an example. I knew how my inspirations and influences were woven into a final design that emerged organically through trial, and error and a lot of sketching.
I had a few different ideas for my final (graduation) project when I was at university. Because it forms basically the entire last year of study and at the end, you have to produce a roughly fifth scale model to display at the degree show. It’s a lot of work and expense.
I wanted to punch people in the eyes with something that would stand out and be as individual as me. This is something that AI cannot do. It cannot spontaneously have an initial idea. I came up with an electric hot rod because I thought it would be cool to design and there was a lot of scope for moving the look forward, and it fits in with my self-imposed brief to reflect my personality and style.
Here are the actual mood boards and sketches from my final presentation, 10 years ago. I’m not sure what possessed me to use that font, but hey, I wasn’t the epitome of exquisite taste then that I am now. You can see how the initial sketches evolved from something you could probably see at SEMA into something totally different that captures a lot better what I wanted.
Here’s the final design rendered from the digital model.
I chose Midjourney for this experiment because the beta is a little less restrictive in terms of how many images you can create. It works in a Discord server. You type in your prompts and a few seconds later it replies with 4 images.
Here’s what I started off with. I didn’t really give it enough to go on, as these look like bad merch images. Notice how it has trouble getting wheels in perspective and one of the rods has no front suspension.
Second try. I mentioned Braun and Daniel Simon to try and give it a bit more direction. The text in the second image almost says RuPaul, which suggests it can’t tell the difference between drag racing and Drag Race.
This time, I gave it as many prompts as I thought feasible. I mentioned some fashion designers hoping it would reference their work and incorporate those visual signatures into the design. Gareth Pugh designed the jacket on the mood board (I adore his work and he’s an RCA grad like me). Maybe the word hot rod was too restrictive. By taking it out I hoped it would not look at traditional ’32 Ford-style cars and interpret these instructions into something closer to what I had in mind.
I didn’t say rejected Tumbler Batmobile proposals. (On an unrelated note, Matt Reeves: Call me?) Version 4 is somewhat closer. Enhance!
You can spend all day doing this, trying to come up with the correct series of words to coerce the AI into giving you what you want. It’s a dice roll every time, the very opposite of considered and careful refinement. It’s not efficient, and in the design studio, time is everything. There are ways of giving the bot more sophisticated instructions by making it reference specific images but that’s getting more into the realms of programming, and my artsy fartsy head brain doesn’t work that way (I’m autistic, and guess what—not all of us are number prodigies.)
As I mentioned earlier, my social feeds are ablaze with AI mentions at the moment. A fellow pro designer and founder of Car Design News, Brett Paterson posted this design for a self-driving taxi on LinkedIn.
This was the result of about 80 iteration loops and some Photoshop touching-up. I asked him what his final prompts were so I could try them for myself.
Different but much closer to the desired result than I managed with the hot rod. And a lot more cohesive and believable.
Further iterations. These are all the raw images produced by Midjourney. I’ve not done any editing on these at all. So what have we learned from all this? Basically, with the correct programming, AI could be used to generate alternative design iterations pretty quickly.
But it’s a scattergun approach—you never know what you’re going to get. In a worst-case “let’s sack all the designers” scenario you could imagine the correct prompts could become a closely guarded corporate secret like the formula to Coca-Cola, but it’s not likely because the results are too variable, too subject to the unseen whims of lines of code masquerading as something it isn’t.
One argument I’ve seen online coming from sponge-brained knucklecocks is that designers will have to adapt and like it. It’s about as well thought out an argument as “learn to code or become homeless” or “get a better job.”
The design process has been adapting to integrate new tools ever since Harley Earl introduced full-size clay models into the studio in the 1930s. We’ve had the advent of ever more sophisticated CAD/CAM, 5-axis milling machines, Photoshop replacing traditional pencil and marker rendering, 3D scanning and additive manufacturing, desktop polygonal modeling and now virtual reality. All of these have found their appropriate place in the designers’ workflow.
At the risk of sounding like a bleeding-heart socialist European, I believe in equality of opportunity, not equality of outcome. Some of the tools you need to learn have already been democratized. A basic Photoshop subscription costs $20 per month, and there are open-source and one-time license alternative programs available. Polygonal modeling software Blender is free for individuals. Shit, I’ve even started doing sketch tutorials on this very site that you can use for nothing.
But all these things are simply tools. You need the vision and ability to understand what you want them to do in the first place. Midjourney can only create within certain limited parameters. It cannot refine, divine, add emotion or context or come up with an idea on its own. It can only reformulate something that already exists in a manner it’s programmed to, nothing else. It can’t look around and write a design brief or see a problem to be solved. A chess program can’t suddenly think up an idea for a new tabletop RPG.
I can go to Guitar Center and pick up a starter Squier Strat and practice amp for less than $250. Given enough time and practice I might be able to murder the intro to “She Sells Sanctuary.” But I would never write something like that because that’s not where my creative gifts lie. I’m not Billy Duffy (and yes, I know he plays a white Gretsch Falcon.) The same will almost always be true of AI. AI should do the menial stuff so humans are free to create art, not AI creating art so humans are stuck doing the menial stuff.
None of this will stop the world’s tech bros from using AI to replace as many humans as they can, until this technology inevitably fizzles out just like crypto and self-driving cars. But for the rest of you—those that aren’t in the tech-investor class—consider what you’re about to lose in the process.
And think hard about whether that’s really a world you want to live in or not.
There’s another aspect that’s missing from this parasitic brotech: collective energy. When you get more than one creative people together, their energy can combine like Voltron to be more than the sum of their parts. The output is most obvious to most of us with great bands that break up with the resulting solo careers of the artists being more mediocre work than before. It doesn’t necessarily have to be collaborative and it’s something much more than mere inspiration, a feeling like being “in the zone”, but elevated by the presence of someone else on a similar creative wavelength. A machine will not be able to do that.
I believe the intellectual name for it is the third mind, which implies if there are two people in a group there is a third, in the form of the consensus. But yes, car design these days is never the work of one person, no matter what the PR department would have you believe.
I’ve been trying to get a job in car design for 5 years now, and once I started to see AI designs get more sophisticated, I got nervous that by the time I managed to break into the industry, that career would be phasing out. This gives me hope that maybe someday I can still have that career.
Feel free to reach out on Twitter/Instagram/email (adrian.clarke@outlook.com). I can’t get you a job, but I may be able to give you some advice.
Paint by numbers. 1 is red, 2 is blue, etc. At the end you have The Mona Lisa…with a mustache. That bit of whimsy makes one an artist uninhibited by convention. Ta-daaa!
Gotta love it! As a person who lost a job due to computer efficiency, newspaper circulation, who also said they can’t do what I can, i love to see the next planet in line for Galactus. Just a heads up they Couldn’t do what I did. But they could cut so many people that at half the circulation they made more money. As far as design I agree most AIs cant design anything. But have you noticed most car companies are leaving sedans, coupes, ICE, and more? They will accept the most boring Suv Design ang just make it longer as long as it is profitable. Hey Toyota succeeded with exceptional performing boring cars. Noone wants the sportscar market. Just expensive SUVs and Pickups. Hell atesla hasnt redesigned any of their cars and they are doing great. Most of America and Americans are average and that is what they expect from cars average. Hell i live in the Pittsburgh area, what in the 70s was called the city of champions because in 1978 our teams won the SB, Stanley Cup, and the World Series. Now the Steelers havent been in the playoffs in 6 years, the Penguins usually get in the playoffs and lose in the 1st round, and the Pirates? Well every other team loves to play the Pirates because they suck. So yeah everyone is settling on cars like Pittsburgh is settling in sports for losers.
Unfortunately a lot of customers do like SUVs because the elevated entry height makes it easier to deal with kids. Plus, the handling and economy compromises are not what they were twenty years ago.
Car building is a very capital intensive and low margin business (10% if you’re lucky for the mass market) so they have to build what sells.
But these things are cyclical. It went from wagons to minivans to SUVs and with increasing and environmental headwinds it may swing back again.
Just one thing… the modern Twingo is, behind the scene a Smart. They share the platform. (1)
Now if you used an original Twingo, it’s something else as the platform was totaly original and Renault specific.
(1) and trust me, for having driven and been passenger in the Twingo V1 ( despite a few mechanical issues best left untold ) and having done the same in the modern one… the modern Twingo s**ks.
First it’s much heavier, second it went from FWD to RWD and the engine didn’t follow the weight gain [ since it’s the Smart engine ], third the space inside shrank ( you could comfortably house 3 adults in the old Twingo for intercity trips in France, no way it can be done in the modern one… the poor guy behind will be cramped ), and I’m not even talking about the trunk… because it would get bloody.
Yes but at least the gen 3 Twingo is weird. The second gen was just an cheapoid hatch.
Great writing as always Adrian, reminds me of some of the best columns that Wheels Magazine used to have here decades past.
The more I think about it, how far has all this really moved on from the ‘Mechanical Turk’?
People still need to guide and feed the algorithm and the final results still aren’t quite there, same as so much other ‘AI’ technology and getting that last few percent is so difficult it may as well be an impossibility regardless of the billions thrown about.
The problem with AI drawing is that they’re taking a fun toy and trying to make it something you can do work with.
I very much enjoy putting absolutely absurd prompts into these generators to see what happens. But to even attempt to get something usable real humans need to give it a few passes anyway. So why use the computer?
It’s the same with every creative process. I write, the grown ups here edit, tweak and suggest. You go write a screenplay, someone else reads it and gives you notes. In the studio you put your sketches up and your peers and the managers give you direction.
No creative person is an island, and you’re right Midjourney isn’t solving that non-problem either.
Related, the Final Cut Pro people are insistent that Apple’s greatest mistake with the program right now is that it doesn’t enable collaboration. Post-production work, like editing, color grading, etc, needs to be done at the same time by a lot of people in a lot of different places. If design-like tools don’t enable that today then they’re not as useful as they should be.
Yeah, that’s another reason why starting with a real person would actually make it easier than starting with AI. People can explain themselves, so if I question why you put that line there you could tell me. A computer can’t do any of the at.
Using it for a commercial product, you’re going to be more likely to just outright toss anything that the AI comes up, because it doesn’t have reasons for what it’s doing. I almost wonder if it would actually make design more difficult at the end of the day – how do you collaborate with such a thing?
Midjourney has at least one homerun:
Per a Reddit post, the input was:
“Glamorous photo of a rugged Bob Ross wearing an intricate barbarian armor, muscular, holding paintbrush, vibrant and colorful.”
And the result was glorious:
Meet Bob-Barian Ross:
https://www.reddit.com/r/funny/comments/x4jd00/bobbarian_ross_the_trees_are_unhappy/
I think it does better at that sort of thing, presumably because it has much more material to reference.
But holy fucking emergency hams, that Bob Ross image….
Play some Bathory to go with it and you complete the aesthetic.
“Sponge-brained knucklecocks”!?!
Poetry, sir. Sheer poetry.
Your servant sir.
I don’t see AI like crypto or NFTs at all. I think it has “potential”, but more for aesthetic appreciation (more images to look at!) than for utility, at least so far.
Not pretending I can even approach the fustercluck of “well it scraped my image for training data and I don’t want it to”, but otherwise I don’t see it as a huge problem *when labeled as AI art*.
My friends and I got a lot of laughs with DALL-E 2. And not in the “haha this is bad” kind of way. Absurd stuff like “The Minions go to Burning Man”. And some of my attempts at vehicle ideas looked surprisingly realistic.
I did a lot of reading to research this article. I think like all emerging tech, once the buzz and hype wears off it will find some uses, and they won’t be what we expect.
In the field of car design, I can see cloud based solutions for patch layout when surface modelling maybe, but I don’t think it would be able to build a G5 continuity model from scratch. But it might be useful for suggesting solutions to solving certain modelling problems quickly, that can then be rebuilt by a human.
(Midjourney) “cannot refine, divine, add emotion or context or come up with an idea on its own.”
Yet.
As a college professor with 27 years of teaching experience, I’ve been hearing variations on this theme as it pertains to teaching. So far, no computerized tool can replace a human teacher.
Yet.
I do take comfort in the fact that I will still be employable for the rest of my life (probably). But the tools are getting better and better and I firmly believe that the teaching robots from Asimov’s fertile imagination (as seen in Foundation and in other stories) will arrive one day. I don’t think it will be in my lifetime, or even my children’s, but I do believe we will get closer in the next few decades. Currently, we are observing what might be an actual intelligence (T.B.D.) making its first crawling, mewling, puking attempts at imitating us. I, for one, intend to treat this infant intelligence with caution and begrudging respect until I see how it turns out. I just hope it has decent parents who guide it into becoming a productive citizen of the world, if it becomes anything at all.
I would put money on it happening no later than your children’s professional lifetime. It’s the same reason that I don’t think young people ought to become long-haul truck drivers, because I think that career will begin going away in about a decade. They’ll run up against 40, with no saleable skills.
Things happen very, very slowly, and then very fast.
I love this article, for all the reasons outlined in it.
AI “art” isn’t really art. It is completely soulless. For cars, it won’t quite ever get it right as the tech exists today, and I personally don’t think it ever will. I had an AI crap these images out for a car description I input:
https://i.imgur.com/O8j6TSj.jpg
https://i.imgur.com/yXIuCNX.jpg
However, this technology is great for making nightmarish grotesque abominations, because the uncanny valley hasn’t been crossed yet, allowing for the generated images to look all the more unsettling and strange. I’m using it to generate concept images for a pixel-based dark sci-fi jRPG I’ve been working on. Some examples:
Nite nite, don’t let the bled blugs bite:
https://i.imgur.com/zIXGFdk.jpg
They get plump when they feed:
https://i.imgur.com/ehALdwl.jpg
Duckbill Platyhelms:
https://i.imgur.com/4iacRV3.jpg
Mighty Mite:
https://i.imgur.com/NVOip0D.jpg
I’m drawing my own pixel art based upon these images. They make good templates with which to consult for aesthetic guidance. But they will never substitute for an actual artist’s vision.
Alright, so I’m also a designer, but industrial design, not transportation design, so my take is a little different on it. That said…. this shit is scary especially if it affects your field, and watching it evolve over the past 12 months has been absolutely frightening.
At first, the results were not very good, but it could capture illustrative styles okay. I would say 6 months ago, I started seeing it output incredibly detailed concept art. My peers that graduated with a degree in illustration were seething; their bread and butter was doing creating art based on commissions, and things like paintings for Magic/Gathering cards, stuff like that. In almost an instant…. their entire livelyhoods seemed vaporized. I started seeing them post about AI created art winning art contests, where the winner had no technical skills whatsoever, and was just touching up output from these AIs.
I would say 3-4 months ago, I started seeing the crossover to industrial design, and SO FAR, I’m not too worried. Almost all the results are fairly ‘stupid’ if you start looking at them closely, have major issues, would not be usable, have elements where they shouldn’t be, poor ergonomics, etc etc etc. I could see this being a useful tool in the ideation phase when you want quantity over quality; I even tried it for a program I was working on, unsuccessfully.
That said, the time period it went from “haha amusing chipmunk with afro” to “wow that is an impressive piece of concept art” was less than 6 months. Soon, I feel like you’ll be able to not only get better results, but possibly create 3D models using AI, based off inputting 2d images. This is where stuff gets real scary, and the gates come down for our careers.
Suddenly, anyone is able to create ‘roughly’ anything they want in 3d, with no effort. This will ignore design constraints, technical limitations, safety, etc, but …. maybe it will be good enough to hand off to engineering to start modifying it for production?
It’s terrifying to see such a hugely rewarding task that requires a ton of hardwork/practice to acquire talent become meaningless…. BUT.
Design is not just a pretty picture. Maybe it’s because I am industrial not transportation, but imho automotive design has been kind of boring and awful to me for a while now. Most of today’s cars have creases and fake vents and scallops and complex surfacing to hide the fact the cars have shit proportions. Instead of design being pushed to create gorgeous cars, it’s instead being used to create controversy (cough BMW cough), hide proportions, or just look “different” so consumers feel like their current car is stale, BUT IT IS NOT SOLVING PROBLEMS.
IMHO, design is looking at the world, finding problems, and creating solutions. For transportation design, I see AI as being a real threat in the future considering how quickly it is improving, but for industrial design, where you have so many other aspects, I’m not so sure. Even if you take out sketching AND cad modeling from my job, a lot of it is talking to engineering and finding compromises to whatever they’ve come up with, a lot of ideating in 3d, user testing, trying to find ways to fix problems and make experiences better.
I don’t think AI is going to take that way. I hope.
The creative phase of car design forms only one small part of the process. If I got to spend one day a week purely sketching and rendering, that was a good week.
The rest of the week I was in meetings with engineers, solving problems for production, or with suppliers pushing them for better parts. Or taping up clay models, getting models ready for review, supervising and directing modellers, or prepping images for reviews.
Well, I guess you’ll be okay too. For a while. What freaks me out is HOW FAST it got ‘pretty good’. Imagine the progress a decade from now.
Your hot rod is badass! Weird how the robot just started making Batmobiles. I want one of those underwater Renaults, but I’ll never buy one. I’m anti-robot-overlord. I’ll only ever consider a personal submarine designed by a human.
Thank you. It was extremely well received, but didn’t get me a job. The model survives and is in my friends garage.
I’d have hired you for that design. It’s a Lightning Rod. A fabulous one that RuPaul would have been proud to drive.
I really like that also. A few years ago, or 10 or something, I think it was Audi that did a concept for a little 2 seater that looked similar to yours. I’ve never been able to find it again. May you or someone here has a link to it? It could have also be Volvo or Bugatti maybe. European for sure, but pretty sure I remember the Audi logo.
There was a hard top, like your concept and an open top version. Or maybe I just had a good dream one night and I still remember flashes of it. I’ve been trying to find those images on and off for a couple of years and no luck so far. Yours is really close to the shape.
Maybe it was this. But it is nothing like I remember it to be. I’m thinking more like the grill and headlights from that first front 3/4 view on your body.
http://www.ausmotive.com/2011/08/30/audi-continues-urban-concept-tease.html
Yes the Audi Urban concept. They made a drivable prototype, that I think Car magazine drove.
I’m very curious to see if AI can get out of the uncanny valley where it is currently entrenched. I have a suspicion it won’t, at least not any time soon, for the same reason autonomous cars won’t. Throwing a massive amount of data at an algorithm can yield some impressive results, but getting through that last few percent of accuracy where it can reliably provide excellent results is much, much harder.
I always look back at the IBM Watson appearance on Jeopardy. Yes, it was impressive that it worked as well as it did…right up until it gave an incredibly obvious wrong answer. At the time I think the assumption was they would clean up whatever bug caused it and everything would be amazing in our AI-controlled utopia. In reality, IBM sold off Watson because it could never figure out how to make it work in a practical scenario. Maybe that will turn out to be a mistake, but I suspect the people in charge realized at some point that reliable AI is a lot harder than they expected. Creating meme images on the internet? Sure. Diagnosing cancer? Ehhhh, not so much.
I love that it is stuck in the uncanny valley, precisely because by being stuck there, it is perfect for creating eldritch abominations that are total eyesores. I love that sort of imagery, and AI is currently excellent at generating it as a result of not being good at what it is tasked with doing.
If it evolves beyond that, I hope the older algorithms are retained as an option. Otherwise, we will lose something truly wonderful.
I agree with everything that you said, however I have a heart and a soul.
Bean counters have neither.
Capitalism doesn’t even understand the words.
A designer’s output is always a reflection of their lived experiences, memories, preferences and bias. Very human qualities no silicon chip can emulate.
Well said. Very well said.
I mean, at least till we start implanting memories into the replicants…
Qualities themselves, yes, but all AI does is emulate human work. Creators are suing AI companies for unauthorized use of their images for just this reason–when an AI produces compelling results, it is really just a pastiche of the various source materials. There is no original content here–just an algorithmic theft of human work and inspiration.
Its a kind of strange argument, and I think the wrong way of going about arguing against AI art.
In music, especially modern music, remixes are common, sampled songs and sounds and snippets are common. Or, if I take a photo that is inspired by someone else’s, down to the same color palette, subject matter, and composition then it is my work. Unless I am literally in line behind the other person and using their tripod, it is obvious that I’m the one behind the lens. It can be completely derivative of their work, but it is still mine.
But when an AI does it, which is really just another human somewhere unleashing an algorithm on a database, it somehow isn’t original content?
Seems like plagiarism or other similar laws would be the better claim than saying “it isn’t original content.” (and maybe that’s how they are suing them, I just keep hearing the O.C. argument, so maybe the messaging needs to change.)
Not sure what the legal argument is, to be honest, just pointing out that these AI’s aren’t creating from nothing. As far as music goes, incorporating another artist’s work is a mix of fair use and licensing. There is no licensing with these AI’s, and the line gets blurry somewhere with digital manipulation. If you photocopy the Mona Lisa; it’s not your artwork.
My real point is that AI’s respond to the source material, copying an mixing according to the algorithm. No source material means no results. While the humans behind the original works brought all of the human qualities to their works, the AI can only copy what they did, not the factors behind their work, making the results a pale, mutated imitation.
…unless you can convince a judge that putting the Mona Lisa on paper makes it a transformative work, and you might win that.
My understanding is that samples and remixes have to credit or licence the original artist in some way? Certainly I’ve heard of cases where they haven’t been and the original artist has sued and won. I’m not a music expert though, nor a lawyer.
The issue with AI art at the moment is that it might use someone’s likeness or an original artwork without that persons permission. Certainly the individual likeness is an argument I’ve seen in the EU because of GDPR laws.
I remembered taking an art appreciation course at a local junior college just for the hell of it one summer when I was already enrolled in a major university elsewhere, and I remember the first day vividly. The professor showed the huge lecture hall some paintings from an elephant. She said,
“This famous elephant creates paintings by choosing colors and brushes and applying them to canvas and they look beautiful and they sell for thousands of dollars…IS IT ART???”
I thought it was a typical right-brained, open-ended question that had no real right answer except for your interpretation and justification, and she let us discuss it for a while, as I expected. But after about 5 minutes of debate, she surprised us all with an exact answer: “NO, IT IS NOT ART. Art has context. It comes from somewhere. It has reference points and is the result of a creative process. An elephant, no matter how ‘talented’ with a brush, can’t create art.”
AI is a very, very talented elephant with a paintbrush.
Except – except – the elephant painting has all that, it’s just *not provided by the elephant*. The elephant is middleware; the artist is the one that gave the elephant the paintbrush and trained it to paint in the first place.
This is one of the things that I find *infuriating* about the current AI debate. It’s the same debate as CGI art from the last century. Did you know Tron was disqualified from a Visual Effects oscar for using digital art? Think about how ludicrous that sounds today. It took hundreds of hours of human labor and inventiveness to create the 15 minutes or so of CGI in Tron, but at the time people treated it like someone had just typed ‘amazing movie’ into a terminal the size of a washing machine and smacked enter. All the amazing AI artwork making the news isn’t made by someone typing in ‘an amazing art’ into the engine. Aside from a few lucky results, it’s all from *people* using the *tool* for tens to hundreds of hours, learning, refining, getting the right result, then using photoshop, *an already respected instrument for creating art* to clean it up.
It’s just another tool. Learn to integrate it into your creative process, or don’t. That’s your decision as an artist. Personally, I’m excited. I’m already experimenting with ways to use it to improve my own work. It’s difficult; learning new things always is, but it’s a hell of a lot easier to clean up after than an elephant.
You make great points, and I agree with the ridiculousness of denying Tron the chance to compete on the grounds that it’s not art.
I think the sticking point with people is that AI produces finished products (or more accurately, ‘products with a finished appearance’), whereas the other tools (Photoshop, etc) do not. Those older tools more closely emulate a brush. But you know, that doesn’t disqualify AI generated art from being art, since it’s directed by humans. Artists have (for CENTURIES!) used understudies and others from their large studios to help them create great works of art that are still attributed to the directing artist even if they were not physically created by the lead-artist’s hands. That’s another thing the art appreciation course taught us–it’s still art as long as it is directed by the MIND of the artist, even if the artist does nothing but provide direction. That’s the difference between a craftsman and an artist (as I understand it–as a complete art novice, I might add). So I think your analogy with a human directing the elephant is apt.
Visual AI is the new NFT. Renault launched a NFT collection together with R5 concept last year and now this..
Thanks, I enjoyed this article a lot!
“Juicero is my prosecutorial star witness in how to spunk $120 million of other people’s money in the pursuit of solving a non-problem.”
Yes, true. But as the great uncle Bumblefuck found out, Juicero was also insanely overbuilt and over-engineered, the damn thing would survive an atomic bomb and still make a cup of juice just fine. In an era of cheaply designed and cheaper manufactured planned-obsolescence consumer goods, we should recognize those few things that buck the disposable trend.
It’d only keep making that juice if you could get the packets, though, right? It may have bucked the disposable trend, but only because they figured it would be more profitable to keep selling the packets than the juicers.
Unless I misunderstood the whole thing.
Somehow that makes it even worse, as they didn’t consider BOM when designing it. It didn’t and shouldn’t have been built that way, but they probably wanted to rush to market and disrupt, rather than focus on driving the unit cost down to an acceptable level. If they’d done that it might’ve stood a chance (I mean it’s still a fucking stupid idea, but if it was cheap enough suckers would have bought it).
I mean, at least in terms of being overbuilt and over-engineered, it WAS designed to be a commercial unit and to be in use constantly.
I get that, but I think in consumer’s minds there’s a dichotomy between ‘buy once at a higher price and never by again’ or ‘buy cheaply and keep having to pay out’. Or at least, there should be.
In that case (and I’m making this argument as a thought exercise, not in anyway should it be considered as an endorsement) Juicero should have eaten the up front cost and made it up on the back end (much like the business model of games consoles or printers).
“It can only reformulate something that already exists in a manner it’s programmed to, nothing else. It can’t look around and write a design brief or see a problem to be solved.”
This. AI art/design/etc. NEEDS artists and designers to work. And it’s going to be derivative because its limits make it derivative. The movement to replace designers/artists with AIs is incredibly short-sighted and fails to understand how new styles come about.
Besides, once the engineers start modifying things for practical reasons, how’s the AI going to understand those inputs and adjust the design? It’s not. You still need designers, and it’s going to become expensive to have them fix derivative AI designs instead of creating innovative ones.
Hypothesis: there will be a premium on people who can write, rather than draw. Both will still need to know the reference material.
I did once meet a husband and wife team of illustrators, at a Y2K dinner party, who saw no point in learning to use a Wacom tablet. I worry about them quite a lot.
I think there is certainly going to be a market for people who can do both. There’s something to be said for keeping up, but I think that keeping up might be having the ability to feed reference drawings in and correct AI outputs. I don’t think that using both will ultimately save companies money.
I know someone who does writing work and a company asked him to edit output (which is to say rewrite to sound less AI generated) from Chat GPT. He rightfully said that he could do that for the same rate he had been charging for writing, since rewriting AI writing is not really saving as time vs writing copy from scratch and is a different skill. AI design is going down that same path.
Yes. Fix this AI generated image in Photoshop will probably take the same amount of time as creating something original from scratch.
“None of this will stop the world’s tech bros from using AI to replace as many humans as they can, until this technology inevitably fizzles out just like crypto and self-driving cars.”
– I agree with you completely on the first half of this statement. And I can only hope that the second half comes true.
“We all carry a personalized, powerful and configurable personal assistant and communication device with us all the time.”
I don’t have a cell phone. I assume my recalcitrance will delay the arrival of the Singularity by a good 15-20 ns, which means I’m doing my part. You’re welcome.
Fair play to you, and I kind of admire you being able to get by these days without one.
I use a $10 flip phone from 2 decades ago. It initiates or receives calls, which is all I need it for.
As is so often the case, the key is diminished expectations, not necessarily my own.
No, I get that. I don’t bother texting Mother Dearest or sending her photos because I know her mobile is turned off at the bottom of her handbag.
With no credit.
I’m still going to get an alert for it either way, with both a buzz and a banner, but thanks much!
That shitty, shitty, shitty analogy: i lost the article 1/3 in. Dobby driving a Landspeeder is whizzing through my brain. Bikini’d Griphook dancing for Jabba, even worse.
“Catch the snitch with the Force, ‘arry!”
“Yer a wizard, Luke”
My mind. She is aflame.
Flitwick fired first?
That’s actually not a halfbad Batmobile
Interesting exercise!