on - medium term consequence is that the “economic blanket” for professional photographers will get even shorter …
hey, if I can get those greater than great Dyn.HighContrast, highly saturated and sharpened postcard pics out of my flagship phone, why should I pay a photographer for something similar, or “less flashy” (albeit way more accurate) … another market segment will (slowly?) melt away…
easy here when crying “protected by copyright” - what are practical and legal limitations to copyright?
Legislation??? Does the copyright of a Swiss guy in Namibia who snaps a pic of a burger apply in Bolivia, in Hong Kong?
… are their “fair use” clauses in copyright? what are their limitations?
what if you upload your burger pic to twitter/IG/FB/Straight Dope et al… do you waive your copyrights… ?
how do search engines fit in all this: (img. search for “photo of burger”) and have my AI look for pages after pages of burger-pics?
I believe all this “matters” (as per your post) - so it’s not as easy as “but its protected by copyright!!!”
I wasn’t “crying”, nor did I used multiple exclamation marks. I also don’t have a horse in the copyright dispute so, even if I was inclined to discuss this with someone who responds like that off the bat, I don’t see the point here. There’s also probably better threads for it in GD than cluttering up one about AI impact on the workplace.
Photographers can’t complain. They took the jobs away from all those illustration artists before photography. Imagine, spending all day drawing something, then some guy comes along with a device and pushes a button, and gets something better! There should be a law. Magazine Illustrators of the world, unite!
The more serious point is that AI is here, and it’s not going away. People learn from previous artists, and copy parts of the styles and expand the art. AI is no different, except maybe in the scale at which it can be used.
Plus ça change, plus c’est la même chose.
apologies offered if you feel my response was too strongly worded … (I gladly admit to interpunctation excesses!!! ← (hehe))
I also have no dog in this fight, but it seems this matter is very complex and the big dogs (who do have a dog in this fight) seem to just fight it out (NY-Times vs. the AIs) … in any case goes way beyond “my photos are protected by copyright”
fwiw - I feel this interchange fits quite nicely in the OP …
best regards
That’s not the core complaint of the creators involved.
Honestly, as I previously mentioned, I’m deep into using AI generation --probably in the top five active people on the board if not the top for actual methods & means – and the arguments from you guys are making ME cringe.
…exactly this.
Software automatically collected data from websites including stock image sites, social media, personal websites, artist showcases, many times in violation of the terms of services of the websites it scraped, without permission, or compensation.
Not “willy-nilly.”
Namibia, Boliva and Hong Kong are all signatories to the Berne Convention.
Fair use is a doctrine, it isn’t a “clause” in copyright. In the United States it allows limited use of copyrighted material under a “four-factor test” that includes the purpose of its use, the nature of the work, the amount and sustainability of the work that was used, and the effect of the usage of the works on the value of that work.
Fair use isn’t universal. For example, where I live, we don’t have fair use, we have fair dealing, which is a different doctrine that applies different tests, and is a lot less broad than fair use.
Using Facebook as an example:
You don’t waive copyright when you upload to Facebook. You retain ownership of all copyright and trademarks. You assign a licence to Facebook to be able to use it in different ways, for example to display your work.
Search engines post snippets of work that are largely considered to be fair use.
And your AI isn’t just “looking” at page after page of burger pics. That isn’t a useful analogy. It isn’t a person. It doesn’t have eyes. Its software.
I’ll repeat what I said:
I didn’t say anything about copyright. But what Jophiel says is correct. My work is protected by copyright.
I’m not sure what prompted you to be so flippant with your responses here. But copyright is the thing that allows creative people to earn a living. It’s been fundamental in giving space for writers and painters and photographers and musicians to do what they do and be able to get paid for it.
Of course we can.
I haven’t actually complained about AI taking away jobs in my last series of posts. I’ve argued that AI will have a marginal impact on food photography, and that food photography “won’t be toast.”
The thing is…I do have a dog in this fight.
I’m a small dog. There are hundreds of thousands of independent artists, photographers, writers, painters, illustrators who all have a dog in this fight. We are reliant on the “big dogs” fighting the fight on our behalf, even though we know that those “big dogs” really aren’t on our side.
Because it really doesn’t go much beyond, “my photos are protected by copyright." Thats the starting point here. Copyright is effectively the only tool that creative’s have to protect their work.
I’d actually be interested about this. DM/IM/PM me when you have some up. Really good food photography always impresses me.
…will do. At the rate we are going, it will be a very big while away LOL. But when I get to it, I’ll send a link. I’m still in the process of writing scripts, and it will be a while before we even start filming the basic stuff.
That’s exactly the problem with the argument. We’re in agreement that there won’t be a change at the low or high end. But change happens at the margins. There is always someone in the middle, debating whether spending X dollars on Y performance is worthwhile. And if you lower X by making things cheaper, or increase Y by making it better, then you will capture some of that market. Maybe each incremental step is so small to be unnoticeable. But tiny steps add up.
This has already happened with cell phone cameras themselves vs. previous digital cameras. And before that, digital cameras vs. film cameras. Early on, performance was bad, and basically only good enough for toy use. No professional would use that crap. But performance went up, and costs went down, and almost no one uses standalone cameras. Only at the tippy-top high end, or for very specific uses (telephoto) does anyone bother with standalone cameras.
At every point on that development spectrum, people were saying “no big deal, it’s still not good enough for application A, and application B people have already switched”. But incremental progress has a habit of eroding those gaps.
I was speaking rhetorically, and a little tongue in cheeck. But there are several real points in there. One is that technology has been disrupting industry since the start of the industrial revolution. But a more hopeful one is that people of talent won’t be shut out - they’ll be enabled to achieve greater heights. The people at risk aren’t the artists with real vision and talent, but the hordes of workaday illustrators, second unit productions, corporate graphics and video people, etc.
If AI makes some artistic tasks much easier, it enables artists of real vision to build new art on top of it, just as great artists did with photography and film.
Sure, but the argument is that the main factor in those “tiny steps” has been camera affordability and quality itself via cell phones. The advantages AI offers in that mid-space aren’t nearly as significant as to be “industry destroying” in the reasonable future. As previously stated:
If I got an Egg McMuffin that looked like that, I’d return it to the counter. The actual ones I get sit a lot closer to the top image than the bottom (except the sausage patty is the same size as the egg)
There’s a lot of overlap indeed. These algorithms are on a continuum. But there’s still a big difference between AI imaging and generative AI imaging.
The backend of stable diffusion is the same architecture as a NN denoiser or super-resolution. However the front-end of stable diffusion is unique; an encoder that maps the input image into internal features.
And that’s a big difference between current camera algorithms and generative AI (or the half step of moon inpainting).
There are tons of possible threads in which to update the crumbling of Stability AI/Stable Diffusion, but I thought I would pick this one. The hype cycle continues…hype a thing, attract tons of capital, fail to deliver a viable business model, fall to pieces.
Maybe the Stability company is just a little before its time. So the hype right now is just hype, and not a profitable business.
This reminds me of the pets dot com bust back in 1999. It was a good idea…and as Victor Hugo said, nothing is stronger than an idea whose time has come.
The problem was that the time for pets dot com had not yet come. It came a decade later, when a company called chewy dot com became a big success doing the same thing that pets dot com had failed at. But in the meantime, times had changed; the public’s shopping habits changed, the efficiency of home delivery had changed, etc. And so now there was a useful way to profit from the new technology.
AI is a totally new technology, and the world is still trying to figure out what to do with it, how to sell it and how to use it. Like in the 1980’s when computers were a new technology, and companies like Atari and Radio Shack tried to sell machines which nobody had figured out a good use for.
My guess is that pretty soon people will figure out a way to use AI for something practical. And when that happens, the Stablity company will have died off, like pets com. But somebody else will make a fortune; the hype will no longer be hype..it will be real.
AI scares me, because nobody seems to be able to control it. Hallucinations are a real problem. Mistakes made by AI could cause deaths (say, in self-driving cars), or destroy careers (say, in legal documents). So complete adoption of AI may take longer than we thought. (Just as fully autonomous cars are taking longer than we thought).
But for professions like the OP’s advertising .it seems to me that AI could be very useful, very soon. Let it run freely, create artwork and text. Then you will only need one human to supervise it, to catch the mistakes. Most of the other employees at the ad agency may lose their jobs.
Mistakes made by humans are already causing deaths in cars and destroying careers via the legal system. And I’m not just talking about it in a shallow whataboutism way of nothing can ever be perfect. US pedestrian deaths are up over 60% from the all time low with no similar rise in pedestrian deaths in other countries:
And anyone who has interacted with the court system knows there are thousands of common sense reforms that could greatly reduce the degree of injustice via simple errors in the courts that nobody is doing because most people don’t care and a very small group of people care very much that it doesn’t change.
It always fascinates me that something that’s been around for a while just fades into background noise and we develop learned helplessness about ever fixing it while some new thing crops up that has the same damages but 1000x less and all of a sudden, we are enthused about making sure that 1/1000th risk goes down to zero.
According to Malec, Champions has made “about $500K” in card sales so far. Its raison d’etre is that its cards are NFTs which can be traded or purchased with cryptocurrency, but the developer also sells “gems”—which can be traded in for card packs—in exchange for regular US dollars
Of course an NFT grifter would go all in on an AI Art grifter.
Captain Kirk sides with AI.