Considering WotC has doubled down on ai art being a non issue with the fact "ai Art" is creeping into both MTG and D&D how do you feel as a Player/DM feel about this very real problem creeping into the game?
Are you good, bad or indifferent about this as a player/DM?
That's not a very accurate portrayal of the situation. They were lax about detecting it, but have apologized and promised to update their ban to include promotional art.
Wizards has absolutely not said it is a “non issue”—to the contrary, they have made it very clear they believe it is an issue and it has no place into the game.
Now, will it creep into the game? Sure, that is inevitable—you are always going to have bad actors who are lazy and try to pull a fast one. With AI also becoming more ubiquitous in regular tools like Photoshop, the chance of mistakes is also going to be fairly high as both artists and the art review team are unfamiliar with new technology. Laziness and mistakes are not new problems. Folks have been violating copyrights for the same reasons for the past few centuries—AI is just the most modern weapon in the lazy or careless man’s arsenal.
Only a fool will demand perfection from Wizards—and I am sure with each mistake, we will see such folks rise from the woodworks as they always do. For me, all I care about is that Wizards acknowledges their mistakes and take steps to ensure they do not happen again. Considering Wizards has done that with the hiccups so far—and considering Wizards has historically been rather honest, open, and draconian when errors are made when it comes to their “don’t use copyrighted materials in our art” policy, their art team has more than earned the benefit of the doubt that they will treat their AI policy with the same reverence.
Not that I want to be labelled an "AI bro" or whatever... but I'm afraid there's no putting this particular genie back in the bottle. It's a new technology that's out in the wild now. People demanding that generative AI technology be "banned" from use are just as delusional as the people who were against trains and cars because they'd disrupt the horse and buggy trades, or the people who said cameras would destroy the visual arts... or... and I'm old enough to remember this one personally: that art "made on teh computer" was "not real art". Trying to put in short-sighted protective measures always backfires in the long run: ask the UK's car production industry about that.
Did people stop riding horses when the car was popularized? No. Did people stop painting when the camera was invented? No. Did we stop making candles when the light bulb was invented? No. BUT: there were "less" of those jobs to go around. If you want something specific, designed, and of high quality; a skilled human is almost always going to be better than the best generative AI; exactly the same way a hand-m,made Swiss watch is of higher quality than one that comes off an assembly line, or a custom-build computer performs better than a pre-build. However; in both of those cases, the "factory made" one is cheaper, and as they say "close enough for government work". If you want something "disposable" like a video thumbnail or background for a display, or "tedious" like a pile of rubble: then this AI tech is actually a colossal time and money saver for artists included.
I find it rather funny; being someone who has studied history at a university level; the greatest wailing and gnashing of teeth always seems to come about when something once exclusive is democratized; and that's precisely what we are seeing. I have a friend with motor control problems; and he uses AI to produce illustrations that would take him weeks, months, or never to do by hand; it's incredible. On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
To paraphrase the great George Carlin: "We must adapt to the new paradigm: The Earth, plus generative AI."
On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
It's Hanlon's Razor honestly. "Sorry they got caught" implies intentional malfeasance when in reality they just have bad review processes, especially for things that aren't part of the core product pipeline.
Did people stop riding horses when the car was popularized? No. Did people stop painting when the camera was invented? No. Did we stop making candles when the light bulb was invented? No. BUT: there were "less" of those jobs to go around. If you want something specific, designed, and of high quality; a skilled human is almost always going to be better than the best generative AI; exactly the same way a hand-m,made Swiss watch is of higher quality than one that comes off an assembly line, or a custom-build computer performs better than a pre-build. However; in both of those cases, the "factory made" one is cheaper, and as they say "close enough for government work". If you want something "disposable" like a video thumbnail or background for a display, or "tedious" like a pile of rubble: then this AI tech is actually a colossal time and money saver for artists included.
There'll always be some luddism, but a charitable interpretation of gen AI detractors is this: yes genrative AI will be a thing, but we still have a long way to go before its role in our society is determined. While it's in its infancy, governments and other institutions have a chance to create an environment that "aligns" AI with out goals for a just society. The unfortunate truth, as I see it, is that the glide path we're on for AI policy is to essentially let private actors do whatever they want and hope that the result aligns well, which is just an exercise in faith.
Not that I want to be labelled an "AI bro" or whatever... but I'm afraid there's no putting this particular genie back in the bottle. It's a new technology that's out in the wild now. People demanding that generative AI technology be "banned" from use are just as delusional as the people who were against trains and cars because they'd disrupt the horse and buggy trades, or the people who said cameras would destroy the visual arts... or... and I'm old enough to remember this one personally: that art "made on teh computer" was "not real art". Trying to put in short-sighted protective measures always backfires in the long run: ask the UK's car production industry about that.
Did people stop riding horses when the car was popularized? No. Did people stop painting when the camera was invented? No. Did we stop making candles when the light bulb was invented? No. BUT: there were "less" of those jobs to go around. If you want something specific, designed, and of high quality; a skilled human is almost always going to be better than the best generative AI; exactly the same way a hand-m,made Swiss watch is of higher quality than one that comes off an assembly line, or a custom-build computer performs better than a pre-build. BUT: if you want something "disposable" like a video thumbnail or background for a display, or "tedious" like a pile of rubble: then this AI tech is actually a colossal time and money saver for artists included.
I find it rather funny; being someone who has studied history at a university level; the greatest wailing and gnashing of teeth always seems to come about when something once exclusive is democratized; and that's precisely what we are seeing. I have a friend with motor control problems; and he uses AI to produce illustrations taht would take him weeks, months, or never to do by hand; it's incredible. On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
To paraphrase the great George Carlin: "We must adapt to the new paradigm: The Earth, plus generative AI."
Your post would hold more weight if you did not ignore most of the issue with AI. AI is not just a new technology that makes folks’ lives easier—if that were the case, I would be there with you in thinking it was a bit overhyped.
But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not. The biggest issue? Edison could tell you how a lightbulb worked; Ford could tell you how his assembly line was able to produce a car. But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of.
As a secondary concern, AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from. In the olden days, if you built a disruptive technology that stole from someone else, you absolutely are getting your product banned or you are being forced to pay a license—with AI, the law has not caught up to the tech yet, and that means there is no way to solve intellectual property theft problems that historically have had mechanisms to make the wronged party whole.
I do not think it is really fair to just focus just on the Luddites who fear technological change, while ignoring the very real, non-“delusional”, legal and technological issues relatively unique to this emergent technology.
On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
It's Hanlon's Razor honestly. "Sorry they got caught" implies intentional malfeasance when in reality they just have bad review processes, especially for things that aren't part of the core product pipeline.
I mean; how is that surprising? Considering that we know from recent experience they have poor review processes for things that ARE part of the core product pipeline.
To me it's the naked dishonesty that screams "yeah we're guilty": in the case of the MTG promo background teh attempt to say "no no, you're all crazy! That's not AI generated!" only to walk it back a bit later. As they say: "don't pee on my head and then try to tell me it's raining." ... I'd add "Especially if you want me to "trust" you or what you say later."
On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
It's Hanlon's Razor honestly. "Sorry they got caught" implies intentional malfeasance when in reality they just have bad review processes, especially for things that aren't part of the core product pipeline.
I mean; how is that surprising? Considering that we know from recent experience they have poor review processes for things that ARE part of the core product pipeline.
To me it's the naked dishonesty that screams "yeah we're guilty": in the case of the MTG promo background teh attempt to say "no no, you're all crazy! That's not AI generated!" only to walk it back a bit later. As they say: "don't pee on my head and then try to tell me it's raining." ... I'd add "Especially if you want me to "trust" you or what you say later."
Except Wizards did not make the art in question—a third party did, and it looks like, based on the statements of Wizards, their third party did not realise that some of Photoshops new tools were AI based. Wizards repeated incorrect information given to them by a third party who themselves made a mistake—that is hardly “naked dishonesty”, but a pretty easy, honest mistake for Wizards to make.
Or, to use your analogy, someone told Wizards it was raining, Wizards passed that wrong info along, then later corrected themselves. Not really any malice to be found—though the anti-Wizards folks will see some regardless.
Your post would hold more weight if you did not ignore most of the issue with AI. AI is not just a new technology that makes folks’ lives easier—if that were the case, I would be there with you in thinking it was a bit overhyped.
But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not. The biggest issue? Edison could tell you how a lightbulb worked; Ford could tell you how his assembly line was able to produce a car. But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of.
As a secondary concern, AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from. In the olden days, if you built a disruptive technology that stole from someone else, you absolutely are getting your product banned or you are being forced to pay a license—with AI, the law has not caught up to the tech yet, and that means there is no way to solve intellectual property theft problems that historically have had mechanisms to make the wronged party whole.
I do not think it is really fair to just focus just on the Luddites who fear technological change, while ignoring the very real, non-“delusional”, legal and technological issues relatively unique to this emergent technology.
Okay, let me tackle this one at a time:
1) "But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not." - Light, noise and air pollution, Consumption of fossil fuels and rare earth metals. Every new technology has "trade-offs" that we make because the benefits outweigh the costs; or at least, we think they do at the time. Every year, about 1.3 million people are killed on the roads around the world; nobody wants to stop driving though. Sometimes, we DO find that the costs outweigh the benefits, or something better comes along (cigarettes, lead/radium paint, compact florescent bulbs, aerosol cans etc.); but we only know that in retrospect.
2) "But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of. " - THIS is actually a real issue, I'll agree there... Thing is though: that's an issue with all "learning algorithms"; not just "generative AI" technology; from the ones that bring you your Youtube recommended videos, to the ones your bank and credit card companies use to detect fraud; going well afield of the "AI art issue". Mainly because we don't even have that tight a grasp on how humans learn yet, much less how our emulations of that process work. "Emergent complexity" as it turns out is very hard for our poor monkey brains to get a hold of.
3) "AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from." - This is PARTIALLY true; but the idea that AI "takes from" people is ridiculous. A generative AI model contains exactly zero of the things that it studied to be created; we'd have to have data compression technology light years ahead of what we do for that; it woudl actually be far more impressive if we'd created something that COULD "stitch things together" and get the results we do. This is going to be a bit of a hot button to push but 'ere goes: this is actually much closer to how humans "create" than we'd all like to admit. When we make things; we more often than not combine things that we've already seen into something "new": Star wars for instance is "Samurai film, Flash Gordon, and Western in a blender", Star Trek is "Pioneers, but in space"; that's literally where the name comes from, Superman is "What if a circus strong-man (that's where the costume and cape come from by the way)... but REALLY REALLY strong? ... and also from space." D&D itself is "what if Chain mail but Lord of the Rings?". Has the law caught up to this yet? Oh I'll totally agree that it has not... the law has barely come around to the concept of the internet itself; never mind distributed generative AI models living ON it.
Except Wizards did not make the art in question—a third party did, and it looks like, based on the statements of Wizards, their third party did not realise that some of Photoshops new tools were AI based. Wizards repeated incorrect information given to them by a third party who themselves made a mistake—that is hardly “naked dishonesty”, but a pretty easy, honest mistake for Wizards to make.
Or, to use your analogy, someone told Wizards it was raining, Wizards passed that wrong info along, then later corrected themselves. Not really any malice to be found—though the anti-Wizards folks will see some regardless.
All good and fair, until we look at the statements from WotC about "ai art"; they assured both consumers and artists it was not acceptable for ai art in anything published yet it keeps finding it's way into published material AND they still defend it when called out for it. Artists are starting to not work with WotC over this.
It only looks good for WotC in the vacuum of individual circumstance, as a whole with their published statements WotC has plenty of egg on their face when it comes to published ai art.
That's not a very accurate portrayal of the situation. They were lax about detecting it, but have apologized and promised to update their ban to include promotional art.
Is it though? this is hardly the first time it has happened and they have stuck to their guns in how they handle it vs not letting it happen at the very next marketing oppertunity.
Your post would hold more weight if you did not ignore most of the issue with AI. AI is not just a new technology that makes folks’ lives easier—if that were the case, I would be there with you in thinking it was a bit overhyped.
But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not. The biggest issue? Edison could tell you how a lightbulb worked; Ford could tell you how his assembly line was able to produce a car. But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of.
As a secondary concern, AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from. In the olden days, if you built a disruptive technology that stole from someone else, you absolutely are getting your product banned or you are being forced to pay a license—with AI, the law has not caught up to the tech yet, and that means there is no way to solve intellectual property theft problems that historically have had mechanisms to make the wronged party whole.
I do not think it is really fair to just focus just on the Luddites who fear technological change, while ignoring the very real, non-“delusional”, legal and technological issues relatively unique to this emergent technology.
Okay, let me tackle this one at a time:
1) "But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not." - Light, noise and air pollution, Consumption of fossil fuels and rare earth metals. Every new technology has "trade-offs" that we make because the benefits outweigh the costs; or at least, we think they do at the time. Every year, about 1.3 million people are killed on the roads around the world; nobody wants to stop driving though. Sometimes, we DO find that the costs outweigh the benefits, or something better comes along (cigarettes, lead/radium paint, compact florescent bulbs, aerosol cans etc.); but we only know that in retrospect.
2) "But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of. " - THIS is actually a real issue, I'll agree there... Thing is though: that's an issue with all "learning algorithms"; not just "generative AI" technology. Mainly because we don't even have that tight a grasp on how humans learn yet, much less how our emulations of that process work. "Emergent complexity" as it turns out is very hard for our poor monkey brains to get a hold of.
3) "AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from." - This is PARTIALLY true; but the idea that AI "takes from" people is ridiculous. A generative AI model contains exactly zero of the things that it studied to be created; we'd have to have data compression technology light years ahead of what we do for that; it woudl actually be far more impressive if we'd created something that COULD "stitch things together" and get the results we do. This is going to be a bit of a hot button to push but 'ere goes: this is actually much closer to how humans "create" than we'd all like to admit. When we make things; we more often than not combine things that we've already seen into something "new": Star wars for instance is "Samurai film, Flash Gordon, and Western in a blender", Star Trek is "Pioneers, but in space"; that's literally where the name comes from, Superman is "What if a circus strong-man (that's where the costume and cape come from by the way)... but REALLY REALLY strong? ... and also from space." D&D itself is "what if Chain mail but Lord of the Rings?". Has the law caught up to this yet? Oh I'll totally agree that it has not... the law has barely come around to the concept of the internet itself; never mind distributed generative AI models living ON it.
This is my issue with ai, just worded far better than I could ever put it.
Rollback Post to RevisionRollBack
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
Except Wizards did not make the art in question—a third party did, and it looks like, based on the statements of Wizards, their third party did not realise that some of Photoshops new tools were AI based. Wizards repeated incorrect information given to them by a third party who themselves made a mistake—that is hardly “naked dishonesty”, but a pretty easy, honest mistake for Wizards to make.
Or, to use your analogy, someone told Wizards it was raining, Wizards passed that wrong info along, then later corrected themselves. Not really any malice to be found—though the anti-Wizards folks will see some regardless.
All good and fair, until we look at the statements from WotC about "ai art"; they assured both consumers and artists it was not acceptable for ai art in anything published yet it keeps finding it's way into published material AND they still defend it when called out for it. Artists are starting to not work with WotC over this.
It only looks good for WotC in the vacuum of individual circumstance, as a whole plenty of egg on the face for WotC.
Just my take.
I do not think anyone is saying it is a good look for Wizards, just that the folks trying to paint it as something catastrophic or malicious are either lacking in knowledge about the circumstances, misreading the situation, or actively trying to create a controversy where there really is nothing more than a series of fairly regular mistakes in the use of new technology.
Regarding the artist not working with Wizards anymore? Not really sure why you think that is a useful data point--one artist deciding to leave the company really does not mean all that much. Many artists are terrified of AI and acting a bit irrationally about things (see CaptainCorvid's posts about the Luddite reaction) and, having dealt with artists in a professional environment and having plenty of friends and family in the arts, I feel pretty confident in pointing out that some artists are prima donnas and love to thrive in controversy. A single artist or even a small group of artists means pretty much nothing, other than proving that some members the arts community can sometimes be prone to attention-seeking melodrama (as can be true in any profession).
I think this is a situation where a comedy of errors is making everyone look bad--the third-party content creator who apparently did not know what their own tools did, Wizards' quality control for not being trained to recognize AI, the third-party content creator for conveying wrong information to Wizards that resulted in Wizards making a wrong statement, an artist trying to capitalize on the drama to make a name for themselves (and probably hurting their own prospects at the same time since they willingly chose to burn a bridge with the biggest name in the fantasy art business), etc.
3) "AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from." - This is PARTIALLY true; but the idea that AI "takes from" people is ridiculous. A generative AI model contains exactly zero of the things that it studied to be created; we'd have to have data compression technology light years ahead of what we do for that; it woudl actually be far more impressive if we'd created something that COULD "stitch things together" and get the results we do. This is going to be a bit of a hot button to push but 'ere goes: this is actually much closer to how humans "create" than we'd all like to admit. When we make things; we more often than not combine things that we've already seen into something "new": Star wars for instance is "Samurai film, Flash Gordon, and Western in a blender", Star Trek is "Pioneers, but in space"; that's literally where the name comes from, Superman is "What if a circus strong-man (that's where the costume and cape come from by the way)... but REALLY REALLY strong? ... and also from space." D&D itself is "what if Chain mail but Lord of the Rings?". Has the law caught up to this yet? Oh I'll totally agree that it has not... the law has barely come around to the concept of the internet itself; never mind distributed generative AI models living ON it.
The part you are missing here--while similar in nature to how humans create, the black box problem creates very particularized problems. With a human, you have some recourse if the "homage" crosses the line to copyright theft--after all, copyright caselaw is pretty clear that theft goes beyond simply stitching items together, and even a new creation can be considered theft under certain circumstances. Far from being "ridiculous", this is a pretty substantial problem that many of the best legal minds are wrestling with on a daily basis.
I want to be clear on this--I do not think any of this is an insurmountable problem, and I think it is going to get solved sooner rather than later. Right now, there is a class action lawsuit by creators against AI generators which should provide some answers. More importantly, it might provide answers for the entire "class" of content creators whose content AI dredges, which has certain advantages. We actually saw this before with a different tech company--Spotify. Spotify lost a ton of money in a big class action lawsuit back when they started... and it was the best thing that ever happened to their company, as it created rules that governed the entire class of content creators, even those they had no way of reaching out to and cutting individual deals with.
I expect something similar will happen with AI art--and, once there is a legal answer and there are some rules everyone has to play by, that is going to take away a lot of my reservations about the usage of AI.
Also, another flaw I should have mentioned in my prior post--Wizards cares fairly deeply about its intellectual property (as every content creator should), and items generated by AI cannot be copyrighted. Wizards has a vested interest in only using copyrightable art--it gives them protections against folks trying to capitalize on Wizards' game or make counterfeit/fake products. This probably will take a while to get solved--while the Courts might move fast in some things, they have consistently held that AI cannot be copyrightable, and it is going to take a legislative act to change that, and legislative changes in tech are, as you pointed out, absurdly slow to happen. Accordingly, I expect Wizards will keep their anti-AI policy in place for quite some time--it is the smartest thing to do from a self-preservation stance.
Wizards has absolutely not said it is a “non issue”—to the contrary, they have made it very clear they believe it is an issue and it has no place into the game.
Now, will it creep into the game? Sure, that is inevitable—you are always going to have bad actors who are lazy and try to pull a fast one. With AI also becoming more ubiquitous in regular tools like Photoshop, the chance of mistakes is also going to be fairly high as both artists and the art review team are unfamiliar with new technology. Laziness and mistakes are not new problems. Folks have been violating copyrights for the same reasons for the past few centuries—AI is just the most modern weapon in the lazy or careless man’s arsenal.
Only a fool will demand perfection from Wizards—and I am sure with each mistake, we will see such folks rise from the woodworks as they always do. For me, all I care about is that Wizards acknowledges their mistakes and take steps to ensure they do not happen again. Considering Wizards has done that with the hiccups so far—and considering Wizards has historically been rather honest, open, and draconian when errors are made when it comes to their “don’t use copyrighted materials in our art” policy, their art team has more than earned the benefit of the doubt that they will treat their AI policy with the same reverence.
I disagree with "Only a fool will demand perfection from Wizards"
It is disingenuous to pose everyone who takes WotC at face value for published statements is a "fool".
Except Wizards did not make the art in question—a third party did, and it looks like, based on the statements of Wizards, their third party did not realise that some of Photoshops new tools were AI based. Wizards repeated incorrect information given to them by a third party who themselves made a mistake—that is hardly “naked dishonesty”, but a pretty easy, honest mistake for Wizards to make.
Or, to use your analogy, someone told Wizards it was raining, Wizards passed that wrong info along, then later corrected themselves. Not really any malice to be found—though the anti-Wizards folks will see some regardless.
All good and fair, until we look at the statements from WotC about "ai art"; they assured both consumers and artists it was not acceptable for ai art in anything published yet it keeps finding it's way into published material AND they still defend it when called out for it. Artists are starting to not work with WotC over this.
It only looks good for WotC in the vacuum of individual circumstance, as a whole plenty of egg on the face for WotC.
Just my take.
I do not think anyone is saying it is a good look for Wizards, just that the folks trying to paint it as something catastrophic or malicious are either lacking in knowledge about the circumstances, misreading the situation, or actively trying to create a controversy where there really is nothing more than a series of fairly regular mistakes in the use of new technology.
Regarding the artist not working with Wizards anymore? Not really sure why you think that is a useful data point--one artist deciding to leave the company really does not mean all that much. Many artists are terrified of AI and acting a bit irrationally about things (see CaptainCorvid's posts about the Luddite reaction) and, having dealt with artists in a professional environment and having plenty of friends and family in the arts, I feel pretty confident in pointing out that some artists are prima donnas and love to thrive in controversy. A single artist or even a small group of artists means pretty much nothing, other than proving that some members the arts community can sometimes be prone to attention-seeking melodrama (as can be true in any profession).
I think this is a situation where a comedy of errors is making everyone look bad--the third-party content creator who apparently did not know what their own tools did, Wizards' quality control for not being trained to recognize AI, the third-party content creator for conveying wrong information to Wizards that resulted in Wizards making a wrong statement, an artist trying to capitalize on the drama to make a name for themselves (and probably hurting their own prospects at the same time since they willingly chose to burn a bridge with the biggest name in the fantasy art business), etc.
Not really sure it is "a comedy of errors" more like a we're gonna do what we do and try to pass it off as a mistake when we get called out.
Rollback Post to RevisionRollBack
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
Wizards has absolutely not said it is a “non issue”—to the contrary, they have made it very clear they believe it is an issue and it has no place into the game.
Now, will it creep into the game? Sure, that is inevitable—you are always going to have bad actors who are lazy and try to pull a fast one. With AI also becoming more ubiquitous in regular tools like Photoshop, the chance of mistakes is also going to be fairly high as both artists and the art review team are unfamiliar with new technology. Laziness and mistakes are not new problems. Folks have been violating copyrights for the same reasons for the past few centuries—AI is just the most modern weapon in the lazy or careless man’s arsenal.
Only a fool will demand perfection from Wizards—and I am sure with each mistake, we will see such folks rise from the woodworks as they always do. For me, all I care about is that Wizards acknowledges their mistakes and take steps to ensure they do not happen again. Considering Wizards has done that with the hiccups so far—and considering Wizards has historically been rather honest, open, and draconian when errors are made when it comes to their “don’t use copyrighted materials in our art” policy, their art team has more than earned the benefit of the doubt that they will treat their AI policy with the same reverence.
I disagree with "Only a fool will demand perfection from Wizards"
It is disingenuous to pose everyone who takes WotC at face value for published statements is a fool.
You may disagree, but, as the old saying goes, "to err is human." Everyone makes mistakes; everyone particularly makes mistakes when they are dealing with something new and constantly evolving. You can--and should--take Wizards published statements at face value... but assuming they will be flawless in executing those statements, particularly with an emergent and rapidly evolving technology? Or getting angry at Wizards during this period of growing pains? Those are positions which, at this time, can only be justified if one completely ignores the human propensity for simple, innocent error--and that, I would suggest, is foolish.
Now, if we see the same problems repeat multiple times under the same circumstances? Then, of course, Wizards should be held to task for that, and I will happily join in calling them out when they have shown patterns of consistent failures. But, as things currently stand, we have not actually seen Wizards make the same mistake twice (one was a third-party independent artist for their books; the other was third-party company for promotional materials, each of which was likely reviewed by completely different teams)--which indicates they are making mistakes, but they are at least learning from them as they are not repeating the same mistake twice.
That is what I want to see--not perfection, since perfection is only an obtainable goal in fiction, but mistakes that Wizards learns from and works to prevent from reoccurring.
Wizards has absolutely not said it is a “non issue”—to the contrary, they have made it very clear they believe it is an issue and it has no place into the game.
Now, will it creep into the game? Sure, that is inevitable—you are always going to have bad actors who are lazy and try to pull a fast one. With AI also becoming more ubiquitous in regular tools like Photoshop, the chance of mistakes is also going to be fairly high as both artists and the art review team are unfamiliar with new technology. Laziness and mistakes are not new problems. Folks have been violating copyrights for the same reasons for the past few centuries—AI is just the most modern weapon in the lazy or careless man’s arsenal.
Only a fool will demand perfection from Wizards—and I am sure with each mistake, we will see such folks rise from the woodworks as they always do. For me, all I care about is that Wizards acknowledges their mistakes and take steps to ensure they do not happen again. Considering Wizards has done that with the hiccups so far—and considering Wizards has historically been rather honest, open, and draconian when errors are made when it comes to their “don’t use copyrighted materials in our art” policy, their art team has more than earned the benefit of the doubt that they will treat their AI policy with the same reverence.
I disagree with "Only a fool will demand perfection from Wizards"
It is disingenuous to pose everyone who takes WotC at face value for published statements is a fool.
You may disagree, but, as the old saying goes, "to err is human." Everyone makes mistakes; everyone particularly makes mistakes when they are dealing with something new and constantly evolving. You can--and should--take Wizards published statements at face value... but assuming they will be flawless in executing those statements, particularly with an emergent and rapidly evolving technology? Or getting angry at Wizards during this period of growing pains? Those are positions which, at this time, can only be justified if one completely ignores the human propensity for simple, innocent error--and that, I would suggest, is foolish.
Now, if we see the same problems repeat multiple times under the same circumstances? Then, of course, Wizards should be held to task for that, and I will happily join in calling them out when they have shown patterns of consistent failures. But, as things currently stand, we have not actually seen Wizards make the same mistake twice (one was a third-party independent artist for their books; the other was third-party company for promotional materials, each of which was likely reviewed by completely different teams)--which indicates they are making mistakes, but they are at least learning from them as they are not repeating the same mistake twice.
That is what I want to see--not perfection, since perfection is only an obtainable goal in fiction, but mistakes that Wizards learns from and works to prevent from reoccurring.
Oh, I am not angry, however "Now, if we see the same problems repeat multiple times under the same circumstances?"
I mean, exactly how many instances of AI art slipping by are we getting, relative to the total volume of art they're taking in? As has been said, demanding their review system be perfect is an unrealistic standard. If they're consistently averaging a 1 in 10 rate of pieces turning out to have AI art, that's a problem. If we're talking 1 in 1000, then it's less that WotC is clearly failing to provide an adequate review and more that no one bats 1000. Even if it does recur, the simple number of recurrences is insufficient data for assessing the efficacy of their review system.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Considering WotC has doubled down on ai art being a non issue with the fact "ai Art" is creeping into both MTG and D&D how do you feel as a Player/DM feel about this very real problem creeping into the game?
Are you good, bad or indifferent about this as a player/DM?
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
That's not a very accurate portrayal of the situation. They were lax about detecting it, but have apologized and promised to update their ban to include promotional art.
Wizards has absolutely not said it is a “non issue”—to the contrary, they have made it very clear they believe it is an issue and it has no place into the game.
Now, will it creep into the game? Sure, that is inevitable—you are always going to have bad actors who are lazy and try to pull a fast one. With AI also becoming more ubiquitous in regular tools like Photoshop, the chance of mistakes is also going to be fairly high as both artists and the art review team are unfamiliar with new technology. Laziness and mistakes are not new problems. Folks have been violating copyrights for the same reasons for the past few centuries—AI is just the most modern weapon in the lazy or careless man’s arsenal.
Only a fool will demand perfection from Wizards—and I am sure with each mistake, we will see such folks rise from the woodworks as they always do. For me, all I care about is that Wizards acknowledges their mistakes and take steps to ensure they do not happen again. Considering Wizards has done that with the hiccups so far—and considering Wizards has historically been rather honest, open, and draconian when errors are made when it comes to their “don’t use copyrighted materials in our art” policy, their art team has more than earned the benefit of the doubt that they will treat their AI policy with the same reverence.
Not that I want to be labelled an "AI bro" or whatever... but I'm afraid there's no putting this particular genie back in the bottle. It's a new technology that's out in the wild now. People demanding that generative AI technology be "banned" from use are just as delusional as the people who were against trains and cars because they'd disrupt the horse and buggy trades, or the people who said cameras would destroy the visual arts... or... and I'm old enough to remember this one personally: that art "made on teh computer" was "not real art". Trying to put in short-sighted protective measures always backfires in the long run: ask the UK's car production industry about that.
Did people stop riding horses when the car was popularized? No. Did people stop painting when the camera was invented? No. Did we stop making candles when the light bulb was invented? No. BUT: there were "less" of those jobs to go around. If you want something specific, designed, and of high quality; a skilled human is almost always going to be better than the best generative AI; exactly the same way a hand-m,made Swiss watch is of higher quality than one that comes off an assembly line, or a custom-build computer performs better than a pre-build. However; in both of those cases, the "factory made" one is cheaper, and as they say "close enough for government work". If you want something "disposable" like a video thumbnail or background for a display, or "tedious" like a pile of rubble: then this AI tech is actually a colossal time and money saver for artists included.
I find it rather funny; being someone who has studied history at a university level; the greatest wailing and gnashing of teeth always seems to come about when something once exclusive is democratized; and that's precisely what we are seeing. I have a friend with motor control problems; and he uses AI to produce illustrations that would take him weeks, months, or never to do by hand; it's incredible. On the one hand: I will say that WOTC more than deserves the drumming it's getting for "taking a moral stand against AI!" publicly, yet employing it repeatedly internally; WOTC are sorry they got caught, not that they saved a few dollars. On the other, I'll actually partially echo Caerwyn_Glyndwr; to quote the old adage: "Don't let the perfect become the enemy of the good".
To paraphrase the great George Carlin: "We must adapt to the new paradigm: The Earth, plus generative AI."
It's Hanlon's Razor honestly. "Sorry they got caught" implies intentional malfeasance when in reality they just have bad review processes, especially for things that aren't part of the core product pipeline.
There'll always be some luddism, but a charitable interpretation of gen AI detractors is this: yes genrative AI will be a thing, but we still have a long way to go before its role in our society is determined. While it's in its infancy, governments and other institutions have a chance to create an environment that "aligns" AI with out goals for a just society. The unfortunate truth, as I see it, is that the glide path we're on for AI policy is to essentially let private actors do whatever they want and hope that the result aligns well, which is just an exercise in faith.
Your post would hold more weight if you did not ignore most of the issue with AI. AI is not just a new technology that makes folks’ lives easier—if that were the case, I would be there with you in thinking it was a bit overhyped.
But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not. The biggest issue? Edison could tell you how a lightbulb worked; Ford could tell you how his assembly line was able to produce a car. But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of.
As a secondary concern, AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from. In the olden days, if you built a disruptive technology that stole from someone else, you absolutely are getting your product banned or you are being forced to pay a license—with AI, the law has not caught up to the tech yet, and that means there is no way to solve intellectual property theft problems that historically have had mechanisms to make the wronged party whole.
I do not think it is really fair to just focus just on the Luddites who fear technological change, while ignoring the very real, non-“delusional”, legal and technological issues relatively unique to this emergent technology.
I mean; how is that surprising? Considering that we know from recent experience they have poor review processes for things that ARE part of the core product pipeline.
To me it's the naked dishonesty that screams "yeah we're guilty": in the case of the MTG promo background teh attempt to say "no no, you're all crazy! That's not AI generated!" only to walk it back a bit later. As they say: "don't pee on my head and then try to tell me it's raining." ... I'd add "Especially if you want me to "trust" you or what you say later."
Except Wizards did not make the art in question—a third party did, and it looks like, based on the statements of Wizards, their third party did not realise that some of Photoshops new tools were AI based. Wizards repeated incorrect information given to them by a third party who themselves made a mistake—that is hardly “naked dishonesty”, but a pretty easy, honest mistake for Wizards to make.
Or, to use your analogy, someone told Wizards it was raining, Wizards passed that wrong info along, then later corrected themselves. Not really any malice to be found—though the anti-Wizards folks will see some regardless.
Okay, let me tackle this one at a time:
1) "But it is not just a new technology and it has inherent flaws planes, lightbulbs, etc. do not." - Light, noise and air pollution, Consumption of fossil fuels and rare earth metals. Every new technology has "trade-offs" that we make because the benefits outweigh the costs; or at least, we think they do at the time. Every year, about 1.3 million people are killed on the roads around the world; nobody wants to stop driving though. Sometimes, we DO find that the costs outweigh the benefits, or something better comes along (cigarettes, lead/radium paint, compact florescent bulbs, aerosol cans etc.); but we only know that in retrospect.
2) "But not even those who created AI understand how their systems produce results—this is the “black box” problem even AI experts are afraid of. " - THIS is actually a real issue, I'll agree there... Thing is though: that's an issue with all "learning algorithms"; not just "generative AI" technology; from the ones that bring you your Youtube recommended videos, to the ones your bank and credit card companies use to detect fraud; going well afield of the "AI art issue". Mainly because we don't even have that tight a grasp on how humans learn yet, much less how our emulations of that process work. "Emergent complexity" as it turns out is very hard for our poor monkey brains to get a hold of.
3) "AI bases the entirety of its output on dredging information from sources not owned or licensed by the AI creator, and often without a way to give credit to those the AI takes from." - This is PARTIALLY true; but the idea that AI "takes from" people is ridiculous. A generative AI model contains exactly zero of the things that it studied to be created; we'd have to have data compression technology light years ahead of what we do for that; it woudl actually be far more impressive if we'd created something that COULD "stitch things together" and get the results we do. This is going to be a bit of a hot button to push but 'ere goes: this is actually much closer to how humans "create" than we'd all like to admit. When we make things; we more often than not combine things that we've already seen into something "new": Star wars for instance is "Samurai film, Flash Gordon, and Western in a blender", Star Trek is "Pioneers, but in space"; that's literally where the name comes from, Superman is "What if a circus strong-man (that's where the costume and cape come from by the way)... but REALLY REALLY strong? ... and also from space." D&D itself is "what if Chain mail but Lord of the Rings?".
Has the law caught up to this yet? Oh I'll totally agree that it has not... the law has barely come around to the concept of the internet itself; never mind distributed generative AI models living ON it.
All good and fair, until we look at the statements from WotC about "ai art"; they assured both consumers and artists it was not acceptable for ai art in anything published yet it keeps finding it's way into published material AND they still defend it when called out for it. Artists are starting to not work with WotC over this.
It only looks good for WotC in the vacuum of individual circumstance, as a whole with their published statements WotC has plenty of egg on their face when it comes to published ai art.
Just my take.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
That's not a very accurate portrayal of the situation. They were lax about detecting it, but have apologized and promised to update their ban to include promotional art.
Is it though? this is hardly the first time it has happened and they have stuck to their guns in how they handle it vs not letting it happen at the very next marketing oppertunity.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
I am for AI art for personal use at your table. I am against passing AI Art off as your own or making money from it as it samples other peoples work.
This is my issue with ai, just worded far better than I could ever put it.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
I do not think anyone is saying it is a good look for Wizards, just that the folks trying to paint it as something catastrophic or malicious are either lacking in knowledge about the circumstances, misreading the situation, or actively trying to create a controversy where there really is nothing more than a series of fairly regular mistakes in the use of new technology.
Regarding the artist not working with Wizards anymore? Not really sure why you think that is a useful data point--one artist deciding to leave the company really does not mean all that much. Many artists are terrified of AI and acting a bit irrationally about things (see CaptainCorvid's posts about the Luddite reaction) and, having dealt with artists in a professional environment and having plenty of friends and family in the arts, I feel pretty confident in pointing out that some artists are prima donnas and love to thrive in controversy. A single artist or even a small group of artists means pretty much nothing, other than proving that some members the arts community can sometimes be prone to attention-seeking melodrama (as can be true in any profession).
I think this is a situation where a comedy of errors is making everyone look bad--the third-party content creator who apparently did not know what their own tools did, Wizards' quality control for not being trained to recognize AI, the third-party content creator for conveying wrong information to Wizards that resulted in Wizards making a wrong statement, an artist trying to capitalize on the drama to make a name for themselves (and probably hurting their own prospects at the same time since they willingly chose to burn a bridge with the biggest name in the fantasy art business), etc.
The part you are missing here--while similar in nature to how humans create, the black box problem creates very particularized problems. With a human, you have some recourse if the "homage" crosses the line to copyright theft--after all, copyright caselaw is pretty clear that theft goes beyond simply stitching items together, and even a new creation can be considered theft under certain circumstances. Far from being "ridiculous", this is a pretty substantial problem that many of the best legal minds are wrestling with on a daily basis.
I want to be clear on this--I do not think any of this is an insurmountable problem, and I think it is going to get solved sooner rather than later. Right now, there is a class action lawsuit by creators against AI generators which should provide some answers. More importantly, it might provide answers for the entire "class" of content creators whose content AI dredges, which has certain advantages. We actually saw this before with a different tech company--Spotify. Spotify lost a ton of money in a big class action lawsuit back when they started... and it was the best thing that ever happened to their company, as it created rules that governed the entire class of content creators, even those they had no way of reaching out to and cutting individual deals with.
I expect something similar will happen with AI art--and, once there is a legal answer and there are some rules everyone has to play by, that is going to take away a lot of my reservations about the usage of AI.
Also, another flaw I should have mentioned in my prior post--Wizards cares fairly deeply about its intellectual property (as every content creator should), and items generated by AI cannot be copyrighted. Wizards has a vested interest in only using copyrightable art--it gives them protections against folks trying to capitalize on Wizards' game or make counterfeit/fake products. This probably will take a while to get solved--while the Courts might move fast in some things, they have consistently held that AI cannot be copyrightable, and it is going to take a legislative act to change that, and legislative changes in tech are, as you pointed out, absurdly slow to happen. Accordingly, I expect Wizards will keep their anti-AI policy in place for quite some time--it is the smartest thing to do from a self-preservation stance.
I disagree with "Only a fool will demand perfection from Wizards"
It is disingenuous to pose everyone who takes WotC at face value for published statements is a "fool".
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
Not really sure it is "a comedy of errors" more like a we're gonna do what we do and try to pass it off as a mistake when we get called out.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
You may disagree, but, as the old saying goes, "to err is human." Everyone makes mistakes; everyone particularly makes mistakes when they are dealing with something new and constantly evolving. You can--and should--take Wizards published statements at face value... but assuming they will be flawless in executing those statements, particularly with an emergent and rapidly evolving technology? Or getting angry at Wizards during this period of growing pains? Those are positions which, at this time, can only be justified if one completely ignores the human propensity for simple, innocent error--and that, I would suggest, is foolish.
Now, if we see the same problems repeat multiple times under the same circumstances? Then, of course, Wizards should be held to task for that, and I will happily join in calling them out when they have shown patterns of consistent failures. But, as things currently stand, we have not actually seen Wizards make the same mistake twice (one was a third-party independent artist for their books; the other was third-party company for promotional materials, each of which was likely reviewed by completely different teams)--which indicates they are making mistakes, but they are at least learning from them as they are not repeating the same mistake twice.
That is what I want to see--not perfection, since perfection is only an obtainable goal in fiction, but mistakes that Wizards learns from and works to prevent from reoccurring.
Oh, I am not angry, however "Now, if we see the same problems repeat multiple times under the same circumstances?"
This is exactly what is happening in real time.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
I mean, exactly how many instances of AI art slipping by are we getting, relative to the total volume of art they're taking in? As has been said, demanding their review system be perfect is an unrealistic standard. If they're consistently averaging a 1 in 10 rate of pieces turning out to have AI art, that's a problem. If we're talking 1 in 1000, then it's less that WotC is clearly failing to provide an adequate review and more that no one bats 1000. Even if it does recur, the simple number of recurrences is insufficient data for assessing the efficacy of their review system.