They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
Alternatively, they'd create a set of tools that make it easier for DMs to do the fun creative work that the AI can't replicate. That's the worker-augmenting change I'd like to see.
Why not both? Isn't that the most profitable?
Note that I'm not against automation per se. I just want to point out that there is a slippery slope here. The more DM functions are effectively replicated by A.I., the more superfluous human DMs become. They don't need to entirely replace human DMs either, just make it so that they can "Mechanical Turk" the system so that a human makes micro-decisions here and there for multiple live games at a time.
Back on the topic of the thread, Beyond has a post on the homepage reaffirming their commitment to not using AI in writing books or in their art. A similar statement was made over on Magic’s site.
I expect this was directly inspired by the conversation surrounding layoffs—there has been a lot of rampant speculation that laying off members of the art department might mean more AI content is coming. I am not sure that speculation was really warranted—Wizards stated they would not use AI art just a couple months ago and one of the individuals laid off was an art director who quire possibly approved someone’s controversial AI art submission to D&D —but it certainly existed.
I think that is a positive step for the game, and shows a commitment to protecting the independent artist contractors who make up a core component of what makes this game and Magic great. It also is nice to see the commitment to not having AI produce any written content for the game—it was hinted that Wizards was anti-AI across the board, but I believe this is the first time the actively said “no, we won’t be using AI in any parts of production.”
While I appreciate that execs at WotC recognize the industry PR problems of allowing AI art to be published with the Wizards logo associated with it, I don't see how this would dissuade them from developing (or contracting to be developed) an AI to replace some of the major roles played by DMs. They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
The only problem being that until the fundamental nature of an “AI” program is massively improved, they’re unlikely to have a viable product. DMing anything more advanced than “proceed down passage, kill monsters, repeat” calls for initiative and flexibility of thought, and current alleged “AIs” possess neither.
While I appreciate that execs at WotC recognize the industry PR problems of allowing AI art to be published with the Wizards logo associated with it, I don't see how this would dissuade them from developing (or contracting to be developed) an AI to replace some of the major roles played by DMs. They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
The only problem being that until the fundamental nature of an “AI” program is massively improved, they’re unlikely to have a viable product. DMing anything more advanced than “proceed down passage, kill monsters, repeat” calls for initiative and flexibility of thought, and current alleged “AIs” possess neither.
They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
Alternatively, they'd create a set of tools that make it easier for DMs to do the fun creative work that the AI can't replicate. That's the worker-augmenting change I'd like to see.
Why not both? Isn't that the most profitable?
Note that I'm not against automation per se. I just want to point out that there is a slippery slope here. The more DM functions are effectively replicated by A.I., the more superfluous human DMs become. They don't need to entirely replace human DMs either, just make it so that they can "Mechanical Turk" the system so that a human makes micro-decisions here and there for multiple live games at a time.
Making both is not necessarily the most profitable because that results in a higher initial investment of resources, and if one is demonstrably or even just likely/perceived to be significantly more profitable while targeting largely the same niche in the same market segment, it’s better to focus your resources on the most profitable venture rather than allocate a portion of your resources to something that will offer a lower return to the detriment of the more profitable offering.
The only problem being that until the fundamental nature of an “AI” program is massively improved, they’re unlikely to have a viable product. DMing anything more advanced than “proceed down passage, kill monsters, repeat” calls for initiative and flexibility of thought, and current alleged “AIs” possess neither.
The word to emphasize here is currently.
Sure, but in the long run we're all dead, yeah? The current iteration of Chat-GPT cost over $100 million to train and has a trillion parameters floating around, and it isn't even close to viable as a DM. The cost to train something that could be viable as a DM in terms of cash and computational power would be potentially orders of magnitude more.
While I appreciate that execs at WotC recognize the industry PR problems of allowing AI art to be published with the Wizards logo associated with it, I don't see how this would dissuade them from developing (or contracting to be developed) an AI to replace some of the major roles played by DMs. They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
The only problem being that until the fundamental nature of an “AI” program is massively improved, they’re unlikely to have a viable product. DMing anything more advanced than “proceed down passage, kill monsters, repeat” calls for initiative and flexibility of thought, and current alleged “AIs” possess neither.
The word to emphasize here is currently.
Yes, in much the same way that “currently” manned intrasolar travel to anything outside of Earth’s orbit is not viable; it’s not a question of simply improving the existing tech, it’s a matter of making a breakthrough that completely changes the paradigm. If WotC is considering it at all, it’s on the “maybe someday” board, not the “what’s our 5 year plan” board.
Fully agreed. What do you think about AI-DM's? If the growth of the game is no longer bottle-necked by the number of available DM's, that could potentially do wonders for the number of users. I wonder if their recent statement completely excludes all forms of AI, not limited just art and writing for final products, and they're nipping it in the bud early.
AI DMs are still a pipe dream at the moment. If you give an AI a simple open-ended prompt it can stitch together a generic response without glaring issues, but they objectively do not have the capacity to actually formulate or run an ongoing narrative.
One day one can hope, it'll only get better from now.
It's not something that can be done with the current generation of "AI" tech. Large language models are fundamentally incapable of doing it. (Also, they're hideously expensive to train, expensive to run, and the huge datasets that they need make them easy to poke into spitting out some pretty hideous stuff, which is the sort of PR nightmare that WotC has the sense to avoid.)
Much more limited stuff like running monsters in combat could be done with machine learning, but you'd need a full rule model first, which is difficult, and it will only work if groups are playing fully by the rules in expected conditions. And sometimes you don't want the monsters to be doing the optimal thing.
Fully agreed. What do you think about AI-DM's? If the growth of the game is no longer bottle-necked by the number of available DM's, that could potentially do wonders for the number of users. I wonder if their recent statement completely excludes all forms of AI, not limited just art and writing for final products, and they're nipping it in the bud early.
AI-DMs are currently not a good option, they can't construct a logical long-term narrative currently and most are biased to agreeing with whomever interacts with them which makes it impossible for them to act as a fair and consistent arbiter of the rules while still preserving the open-ended-ness that is what make TTRPGs different from regular RPGs. And implementing them would be a slap in the face of the human DMs who are the ones who buy most of the content WotC produces. The main thing to really consider is why do people choose to play D&D rather than a digital RPG? To me it comes down to two things: 1) the human connection - D&D is a community, you play it with your friends and may make new friends by playing it. 2) the creative freedom - there are no invisible walls in D&D, no pre-programmed move set that are the only things your character can do.
AI DMs destroy #1 to a large extent, and would either have to limit #2 or would be easily broken / driven insane. But I fully acknowledge other people want other things out of the game, but IMO most of those people are the combat-focused power-gamers - who just want to make that character that can do 10,000 damage on turn 1, because big numbers make them happy. For those people, I'd be happy for them to have an AI to beat up with their OP builds.
Something like a "D&D Arena" where you can pit your character (or a party of character you design) against a random set of monsters on a random map, where the enemies are controlled by AI would be doable and I suspect relatively viable financially. The quality of the AI will be limited and I'm certain players would find ways to exploit it, but the kind of player it would appeal to are the kind that would find figuring out how to exploit it fun and empowering.
They are already aware that one of the biggest constraint to further expansion is the # of people who want to DM (leaving aside the economic constraints fr a minute). It's logical to extrapolate that the easiest solution to that is the mechanize as many aspects of DM work as possible. This could very well be sold using a "We're doing this for our customers" angle, since there will always be a much larger percentage of people who want to "just play the game already" than people who want to DM.
Alternatively, they'd create a set of tools that make it easier for DMs to do the fun creative work that the AI can't replicate. That's the worker-augmenting change I'd like to see.
Why not both? Isn't that the most profitable?
Note that I'm not against automation per se. I just want to point out that there is a slippery slope here. The more DM functions are effectively replicated by A.I., the more superfluous human DMs become. They don't need to entirely replace human DMs either, just make it so that they can "Mechanical Turk" the system so that a human makes micro-decisions here and there for multiple live games at a time.
With respect, it is a bit silly to talk about slippery slopes… while presently engaging in one. You took their silence on AI DMs as cause to be concerned about their recent statement about AI (which focused on things that are real and not speculative).
AI is likely never going to be able to sit at the top of the table and express human emotions with its own mannerisms. It can’t look at or listen to their players and make decisions based on their emotional state, can change the game to avoid problems that seem to be brewing, and can react and adjust to their individual players’ specific interests and desires. It can’t make goofy faces at the players. It will never be able to laugh with the players when something funny happens. It will never be able to react with a joke that doesn’t feel artificial—since everyone will know any joke did not come from the heart.It will never make those human errors which make the game fun—the accidentally slipping into the wrong accent, the “Vex” “No, I’m Vax” mistakes that make people enjoy D&D. It will never be fully human—at least, not in anything close to the foreseeable future.
I expects one day, we will have an AI DM that can provide a better Mansions of Madness experience—something fun, but ultimately a bit soulless. Something that can help scratch the D&D itch, but never fully cures it.
But the idea that an AI DM might render human DMs superfluous? Even the most bullish AI advocates do not think we are anywhere close to replicating the true human element—and many experts think we never will.
But the idea that an AI DM might render human DMs superfluous? Even the most bullish AI advocates do not think we are anywhere close to replicating the true human element—and many experts think we never will.
You're only wrong here because "the most bullish AI advocates" are seriously off the rails.
But the idea that an AI DM might render human DMs superfluous? Even the most bullish AI advocates do not think we are anywhere close to replicating the true human element—and many experts think we never will.
You're only wrong here because "the most bullish AI advocates" are seriously off the rails.
The problem isn't whether or not "AI" is capable or when it will be, but rather when will it be good enough for people not to care. There are plenty of post on this site where people have used it as a DM or in character creation, it is interesting to see the number that think it does a fine job or better.
"AI" isn't gonna be the problem human laziness and "not perfect but good enough" mentality will.
I put "AI" in quotes because what people call "AI" is not even close to AI, and that is alarming in itself.
It's like chat gpt is the result of a drunken night between Google and Miss Cleo.
Rollback Post to RevisionRollBack
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
It's not something that can be done with the current generation of "AI" tech. Large language models are fundamentally incapable of doing it. (Also, they're hideously expensive to train, expensive to run, and the huge datasets that they need make them easy to poke into spitting out some pretty hideous stuff, which is the sort of PR nightmare that WotC has the sense to avoid.)
Much more limited stuff like running monsters in combat could be done with machine learning, but you'd need a full rule model first, which is difficult, and it will only work if groups are playing fully by the rules in expected conditions. And sometimes you don't want the monsters to be doing the optimal thing.
Ah of course, naturally the risk of having it run unregulated is scary. Enemy-AI could be interesting, especially if it can be tuned to the liking of the DM and homebrew-able. As others have pointed out, I don't see having it ease or aid the job of the DM as a bad thing. If it increases the number of players willing to DM their own campaigns and thus more people able to play, more power to it.
But the idea that an AI DM might render human DMs superfluous? Even the most bullish AI advocates do not think we are anywhere close to replicating the true human element—and many experts think we never will.
You're only wrong here because "the most bullish AI advocates" are seriously off the rails.
The problem isn't whether or not "AI" is capable or when it will be, but rather when will it be good enough for people not to care. There are plenty of post on this site where people have used it as a DM or in character creation, it is interesting to see the number that think it does a fine job or better.
"AI" isn't gonna be the problem human laziness and "not perfect but good enough" mentality will.
I put "AI" in quotes because what people call "AI" is not even close to AI, and that is alarming in itself.
It's like chat gpt is the result of a drunken night between Google and Miss Cleo.
Actually, the issue is if WotC thinks they’ll get a return to justify the cost. As been pointed out, training and running an AI is expensive, so they’d need a large buy-in to make it viable. There’s a lot of automatic pushback against AI in the creative community right now thanks to the issue of data scraping, and the fact that existing programs already seem to be accounting for the novelty factor makes it hard for them to draw interest to a more commercialized model. Not exactly auspicious conditions.
The problem isn't whether or not "AI" is capable or when it will be, but rather when will it be good enough for people not to care. There are plenty of post on this site where people have used it as a DM or in character creation, it is interesting to see the number that think it does a fine job or better.
Honestly, that's a wider conversation over how much the bar has been lowered on what's considered acceptable in pretty much every avenue of life
Rollback Post to RevisionRollBack
Active characters:
Carric Aquissar, elven wannabe artist in his deconstructionist period (Archfey warlock) Lan Kidogo, mapach archaeologist and treasure hunter (Knowledge cleric) Mardan Ferres, elven private investigator obsessed with that one unsolved murder (Assassin rogue) Xhekhetiel, halfling survivor of a Betrayer Gods cult (Runechild sorcerer/fighter)
Actually, the issue is if WotC thinks they’ll get a return to justify the cost. As been pointed out, training and running an AI is expensive, so they’d need a large buy-in to make it viable. There’s a lot of automatic pushback against AI in the creative community right now thanks to the issue of data scraping, and the fact that existing programs already seem to be accounting for the novelty factor makes it hard for them to draw interest to a more commercialized model. Not exactly auspicious conditions.
That push back is a good thing as far as I am concerned. This "AI" reminds me of a joke.
Cliff notes for the joke are: people challenged their god to a contest to build a human as they knew the ingredients and could do a better job. After hearing their list of ingredients and the improvements they would make the god accepted the challenge with one caveat, the humans had to make their own ingredients.
This is the issue with "AI" it makes nothing just stands on those that have made something and takes all of the credit. Which sadly is ok with many.
Not sure how the AI policy is related to Hasbro's layoffs. If you guys wanna discuss hat, I'd advice starting another thread.
”They fired a lot of people in creative, they must be replacing them with AI” is a pretty common refrain in the layoff conversation, including being made a number of times on this thread. I think the conversation here has gone a bit far beyond that—but showing one of the areas of concern regarding the layoffs is no longer much of a concern is of value.
Hard to say. DMs are the ones who buy the lion's share of D&D books and peripherals. Cut out the DM and you lose those sales.
They are - under the current paradigm. And yet, I doubt Baldur's Gate's meteoric success was built more on the backs of prospective DMs than prospective players. There is considerable demand for D&D experiences that don't need a dedicated human DM to function.
But I'm just as much in favor of augmenting human DMs with AI too.
Sure, but in the long run we're all dead, yeah? The current iteration of Chat-GPT cost over $100 million to train and has a trillion parameters floating around, and it isn't even close to viable as a DM. The cost to train something that could be viable as a DM in terms of cash and computational power would be potentially orders of magnitude more.
Yeah, why should we bother planting trees or doing reseach? It's not like we'll live to see the results. And it's not like a hobby can last for decades or anything, let's all only care about the short term!
I expects one day, we will have an AI DM that can provide a better Mansions of Madness experience—something fun, but ultimately a bit soulless. Something that can help scratch the D&D itch, but never fully cures it.
D&D doesn't have to just be one thing. There's room for traditional sitting around the table in a basement or at a convention, there's room for VTTs, there's room for solo modules that don't need a DM at all - and yes, there's room to experiment with what AI can do for this hobby, up to and including functioning as the DM for a group that doesn't have one.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Why not both? Isn't that the most profitable?
Note that I'm not against automation per se. I just want to point out that there is a slippery slope here. The more DM functions are effectively replicated by A.I., the more superfluous human DMs become. They don't need to entirely replace human DMs either, just make it so that they can "Mechanical Turk" the system so that a human makes micro-decisions here and there for multiple live games at a time.
The only problem being that until the fundamental nature of an “AI” program is massively improved, they’re unlikely to have a viable product. DMing anything more advanced than “proceed down passage, kill monsters, repeat” calls for initiative and flexibility of thought, and current alleged “AIs” possess neither.
Hard to say. DMs are the ones who buy the lion's share of D&D books and peripherals. Cut out the DM and you lose those sales.
The word to emphasize here is currently.
Making both is not necessarily the most profitable because that results in a higher initial investment of resources, and if one is demonstrably or even just likely/perceived to be significantly more profitable while targeting largely the same niche in the same market segment, it’s better to focus your resources on the most profitable venture rather than allocate a portion of your resources to something that will offer a lower return to the detriment of the more profitable offering.
Sure, but in the long run we're all dead, yeah? The current iteration of Chat-GPT cost over $100 million to train and has a trillion parameters floating around, and it isn't even close to viable as a DM. The cost to train something that could be viable as a DM in terms of cash and computational power would be potentially orders of magnitude more.
Yes, in much the same way that “currently” manned intrasolar travel to anything outside of Earth’s orbit is not viable; it’s not a question of simply improving the existing tech, it’s a matter of making a breakthrough that completely changes the paradigm. If WotC is considering it at all, it’s on the “maybe someday” board, not the “what’s our 5 year plan” board.
It's not something that can be done with the current generation of "AI" tech. Large language models are fundamentally incapable of doing it. (Also, they're hideously expensive to train, expensive to run, and the huge datasets that they need make them easy to poke into spitting out some pretty hideous stuff, which is the sort of PR nightmare that WotC has the sense to avoid.)
Much more limited stuff like running monsters in combat could be done with machine learning, but you'd need a full rule model first, which is difficult, and it will only work if groups are playing fully by the rules in expected conditions. And sometimes you don't want the monsters to be doing the optimal thing.
AI-DMs are currently not a good option, they can't construct a logical long-term narrative currently and most are biased to agreeing with whomever interacts with them which makes it impossible for them to act as a fair and consistent arbiter of the rules while still preserving the open-ended-ness that is what make TTRPGs different from regular RPGs. And implementing them would be a slap in the face of the human DMs who are the ones who buy most of the content WotC produces. The main thing to really consider is why do people choose to play D&D rather than a digital RPG?
To me it comes down to two things:
1) the human connection - D&D is a community, you play it with your friends and may make new friends by playing it.
2) the creative freedom - there are no invisible walls in D&D, no pre-programmed move set that are the only things your character can do.
AI DMs destroy #1 to a large extent, and would either have to limit #2 or would be easily broken / driven insane. But I fully acknowledge other people want other things out of the game, but IMO most of those people are the combat-focused power-gamers - who just want to make that character that can do 10,000 damage on turn 1, because big numbers make them happy. For those people, I'd be happy for them to have an AI to beat up with their OP builds.
Something like a "D&D Arena" where you can pit your character (or a party of character you design) against a random set of monsters on a random map, where the enemies are controlled by AI would be doable and I suspect relatively viable financially. The quality of the AI will be limited and I'm certain players would find ways to exploit it, but the kind of player it would appeal to are the kind that would find figuring out how to exploit it fun and empowering.
With respect, it is a bit silly to talk about slippery slopes… while presently engaging in one. You took their silence on AI DMs as cause to be concerned about their recent statement about AI (which focused on things that are real and not speculative).
AI is likely never going to be able to sit at the top of the table and express human emotions with its own mannerisms. It can’t look at or listen to their players and make decisions based on their emotional state, can change the game to avoid problems that seem to be brewing, and can react and adjust to their individual players’ specific interests and desires. It can’t make goofy faces at the players. It will never be able to laugh with the players when something funny happens. It will never be able to react with a joke that doesn’t feel artificial—since everyone will know any joke did not come from the heart.It will never make those human errors which make the game fun—the accidentally slipping into the wrong accent, the “Vex” “No, I’m Vax” mistakes that make people enjoy D&D. It will never be fully human—at least, not in anything close to the foreseeable future.
I expects one day, we will have an AI DM that can provide a better Mansions of Madness experience—something fun, but ultimately a bit soulless. Something that can help scratch the D&D itch, but never fully cures it.
But the idea that an AI DM might render human DMs superfluous? Even the most bullish AI advocates do not think we are anywhere close to replicating the true human element—and many experts think we never will.
You're only wrong here because "the most bullish AI advocates" are seriously off the rails.
The problem isn't whether or not "AI" is capable or when it will be, but rather when will it be good enough for people not to care. There are plenty of post on this site where people have used it as a DM or in character creation, it is interesting to see the number that think it does a fine job or better.
"AI" isn't gonna be the problem human laziness and "not perfect but good enough" mentality will.
I put "AI" in quotes because what people call "AI" is not even close to AI, and that is alarming in itself.
It's like chat gpt is the result of a drunken night between Google and Miss Cleo.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
Ah of course, naturally the risk of having it run unregulated is scary. Enemy-AI could be interesting, especially if it can be tuned to the liking of the DM and homebrew-able. As others have pointed out, I don't see having it ease or aid the job of the DM as a bad thing. If it increases the number of players willing to DM their own campaigns and thus more people able to play, more power to it.
Free Content: [Basic Rules],
[Phandelver],[Frozen Sick],[Acquisitions Inc.],[Vecna Dossier],[Radiant Citadel], [Spelljammer],[Dragonlance], [Prisoner 13],[Minecraft],[Star Forge], [Baldur’s Gate], [Lightning Keep], [Stormwreck Isle], [Pinebrook], [Caverns of Tsojcanth], [The Lost Horn], [Elemental Evil].Free Dice: [Frostmaiden],
[Flourishing], [Sanguine],[Themberchaud], [Baldur's Gate 3], [Lego].Actually, the issue is if WotC thinks they’ll get a return to justify the cost. As been pointed out, training and running an AI is expensive, so they’d need a large buy-in to make it viable. There’s a lot of automatic pushback against AI in the creative community right now thanks to the issue of data scraping, and the fact that existing programs already seem to be accounting for the novelty factor makes it hard for them to draw interest to a more commercialized model. Not exactly auspicious conditions.
Honestly, that's a wider conversation over how much the bar has been lowered on what's considered acceptable in pretty much every avenue of life
Active characters:
Carric Aquissar, elven wannabe artist in his deconstructionist period (Archfey warlock)
Lan Kidogo, mapach archaeologist and treasure hunter (Knowledge cleric)
Mardan Ferres, elven private investigator obsessed with that one unsolved murder (Assassin rogue)
Xhekhetiel, halfling survivor of a Betrayer Gods cult (Runechild sorcerer/fighter)
That push back is a good thing as far as I am concerned. This "AI" reminds me of a joke.
Cliff notes for the joke are: people challenged their god to a contest to build a human as they knew the ingredients and could do a better job. After hearing their list of ingredients and the improvements they would make the god accepted the challenge with one caveat, the humans had to make their own ingredients.
This is the issue with "AI" it makes nothing just stands on those that have made something and takes all of the credit. Which sadly is ok with many.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
Not sure how the AI policy is related to Hasbro's layoffs. If you guys wanna discuss hat, I'd advice starting another thread.
BoringBard's long and tedious posts somehow manage to enrapture audiences. How? Because he used Charm Person, the #1 bard spell!
He/him pronouns. Call me Bard. PROUD NERD!
Ever wanted to talk about your parties' worst mistakes? Do so HERE. What's your favorite class, why? Share & explain
HERE.”They fired a lot of people in creative, they must be replacing them with AI” is a pretty common refrain in the layoff conversation, including being made a number of times on this thread. I think the conversation here has gone a bit far beyond that—but showing one of the areas of concern regarding the layoffs is no longer much of a concern is of value.
Since WotC has used "AI" in products, it is an easy argument that it could have played a role in the layoffs, but a new thread may be a good idea.
CENSORSHIP IS THE TOOL OF COWARDS and WANNA BE TYRANTS.
They are - under the current paradigm. And yet, I doubt Baldur's Gate's meteoric success was built more on the backs of prospective DMs than prospective players. There is considerable demand for D&D experiences that don't need a dedicated human DM to function.
But I'm just as much in favor of augmenting human DMs with AI too.
Yeah, why should we bother planting trees or doing reseach? It's not like we'll live to see the results. And it's not like a hobby can last for decades or anything, let's all only care about the short term!
D&D doesn't have to just be one thing. There's room for traditional sitting around the table in a basement or at a convention, there's room for VTTs, there's room for solo modules that don't need a DM at all - and yes, there's room to experiment with what AI can do for this hobby, up to and including functioning as the DM for a group that doesn't have one.