Adore your writing style! It was a breeze to read through.
I think AI is going to be hard to avoid. Like social media the network effect is huge. I don't want to be on facebook, but I get sucked back in anyway because everyone else is. I need to arrange that play date with my daughter's Mum. I need to send a quick email, the LLM makes the first draft so easy. And so it continues....
It is hard to avoid already, but I really want to try. Amazon is very hard to avoid in the US/Canada, but we've been managing—for some reason, I see it as a good analogy.
As a software developer, I have started using LLMs more, as a tool to help me out of problems, where normally I would have done an internet search. I have to admit I have found some benefits to this, though I think these are ultimately structured questions/answer sessions, so makes some sense. I've never outright copied the responses I get back, because I don't trust it entirely, and sometimes it just gives a made up answer which sounds right. But in that particular context, I can see it as a useful aid.
Content generation though, as you've rightly said, is another matter. People seem to see GenAI as a "cheap" way of making something for free, except for two problems. One, there is a cost, and it's an environmental impact far exceeding any other form of computing at the moment. Two, it's just obvious that the images or videos you're looking at are fake, the uncanny valley effect immediately jumps out at you so you instinctively know it's "wrong". Yet judging by the YouTube shorts my daughter is obsessed with, people don't seem to care so long as something exists to consume.
AI may be able to replicate, or reproduce, but it can never create.
I understand that use case, Peter, and have used genAI (when requested by clients) the same way, realizing that ChatGPT was acting like a glorified Google for me. I would not mind light use like that if LLMs were trained more ethically, and perhaps at some point we'll get to that. I hope.
When it comes to creativity though... 💯 what you said. Regurgitation and no soul. Too high a price.
Strongly agree that, if GenAI is to be used in gaming at all, it should be clearly marked as such so that users can make an informed choice. I can sort of see the place for it, e.g. when used to upscale textures in old games, but these should be free updates (because it took no labour to create them), should be optional, and above all users should be clearly informed. To present AI textures as the work of humans feels like a kind of fraud, and should be treated as such. In the case of Expedition 33, I'm maybe willing to give the studio the benefit of the doubt, since it seems like the AI textures really were just placeholders that were never supposed to be in the final product, but in general I hope the public will maintain a firm anti-AI consensus in the arts.
You're also very right to talk about the role of AI as a crutch that leads to creative atrophy. Great art comes from limitations, and I can think of a number of amazing games made by solo developers who were not trained artists, but self-taught through necessity and produced amazing art as a result. For example, Miro Haverinen is not, as far as I know, a formally trained artist, but produced very distinctive, original, and beautiful drawings for Fear and Hunger. Lucas Pope wasn't trained as an artist either, but produced brilliant character portraits for Papers, Please.
I am also very surprised by how quickly so many people have accepted ChatGPT and are willing to hold even quite intimate and personal conversations with it. I have never communicated with any of these LLMs and probably never will: it's creepy to me on a visceral level, and I also suspect LLMs will privilege some (bad) sources of information over others, so they're not going to be a reliable research assistant let alone a therapist.
You know, I appreciate you pointing out that there could be a place for it—I didn't touch upon it in the article. I know it will never go away completely, and I wouldn't oppose the usage of it to automate, do QA, even playtest and push updates if need be, as long as clearly stated and not done at the expense of human jobs, the environment, and the privacy and intellectual property considerations. A tall order, but a girl can dream.
Art and creativity are a different matter. I want human-made games still. I realize that I now view games that came out before 2021 as a different beast; I tend to trust them more than new games. Inevitable, probably.
What a contrast between the comments on this piece and my LinkedIn profile.
Being in the tech industry for my main gig I’m always at the bleeding edge of adoption. I can’t escape the GenAI discussion, and navigating it in my client work where my main output is writing has been interesting to say the least. I’ve been fronting the daily challenge of: “Maybe AI can do a better job than you now” for at least six months.
As I’ve found with PR, there’s always a line, and I think the games industry is trying to find it?
Is it obvious? I’m not sure.
On this thread the answer is very clear, any AI use is a negative for the game. But in other circles where I game, I’ve raised this issue and gotten a more muted response? I do agree that 11 Bit’s PR here was tone deaf to say the least!
Clear pieces like this and the response you are getting for it help a lot in adjusting the industry’s barometer.
I’m now more curious than ever to get the Aus game industry’s take on AI — if I can.
Having come to the journalism shores from the tech industry as a marketer myself, I know exactly what you mean. I still occasionally pop onto LinkedIn, and the tone about GenAI over there couldn't be more different, so it's a breath of fresh air to hear the response to my piece on Substack. However! I recognize that we're in a bit of an echo chamber, both here and there—I'm not surprised to see like-minded readers here agree with my stance, just as I know it's a much different ballgame on LinkedIn.
Balance isn't something I touched on in the piece, but I agree, we'll eventually have to find it, and I hope we will. I'm under no illusion that GenAI will disappear completely; I hope and want to fight for a more ethical and just version of it that doesn't do harm.
Please do write an Aus lens take on AI, too, I'd love to read it!
I'm gonna yell about this until the day I die: art is good because it connects me to the person/people who made it. If it was made by genAI, why would I care? We are definitely in a period where a healthy dose of skepticism seems to be necessary when consuming reading, playing, watching almost anything.
I'll admit I even had my suspicions of Sandfall because of the small team, scope of the game, and when doing some digging right after the announcement, I found a passage in their mission statement along the lines of "... using the latest technologies to make our games," which to me was kinda saying it without saying it. Now it seems that's confirmed. That they patched it out at least is a positive sign and they certainly have the resources after the game's success to not feel the need to do it again. Or at least we can hope.
You're wiser than many of us, Josh! In hindsight, the line about the latest technologies does sound sus as hell, although pre-genAI I would have taken it to mean Unreal Engine 5.6 or something, not shitty abstract visuals "painted" by DALLE-E.
Agree about art. Give me all the imperfections of the human behind it. I don't want machine-generated art because it's not art worth consuming, or art at all.
Thank you for writing this! I can't tell you how lonely I sometimes feel when it comes to GenAI. People are really okay with rotting their brains and remain nonchalant over the simple fact GenAI is theft! I'm disappointed to hear about Clair Obscur as I haven't bought or played it yet! Gah! 😭
Yeah! It can be a daunting subject to broach, but you did a brilliant job! 💖 The lack of apology from the Clair Obscur devs must be really grating, especially for fans 😩 It's great to see so many gamers won't stand for it though!
I am probably going to get some flak for this but there is one way that I think an LLM could be used in games -- to power NPC conversations.
This isn't to cut out games writers -- but to enable them to create the rich world in their vision, which might otherwise be technologically impossible to create. I imagine the writers playing a director role, deciding the role that each NPC will play and how they might act. There may be certain lines that need to be said, but the rest is "up to the actor (LLM)" in the same way a director might leave the nuances of the acting up to the actor.
It doesn't get around the ethicalness of being trained on data that it shouldn't have accessed or the environmental impact that using LLMs have. But for me, genAI is a human-directed filler of content, which I think, if we can figure out the ethics and environmental impact, would be a really cool way to give writers superpowers.
Just to be clear, I'm in total agreement with you about using genAI assets or text straight up, especially at the expense of human input. I'm looking to be entertained by humans, not robots!
I'm all for an ethically trained and energy-efficient LLM that would allow NPCs to have unscripted conversations, that would be very cool, but I'm having a very hard time imagining how it would be possible to make one like that. I'd imagine that an LLM would have to be in-game world-specific and trained on human-written content specifically for the game to be immersive... and that would be so prohibitively expensive I don't think any studio would go for it. We can dream though!
Agreed on all points. The nonchalant way that the people around me talk about GenAI is kind of unsettling. I'm a little tired of being the 'odd one out' in these conversations.
The Clair Obscur one made me sad, though, especially since I've been raving about the game so much. I actually did check out those posters when I was playing just a while ago and they containted nice little easter eggs, so I had felt like I learned more about the world of Clair Obscur. But to learn that they actually used random AI assets at first just leaves a bad taste in my mouth. Like you, I'll be giving Sandfall the benefit of the doubt for the same reasons.
It's not exactly GenAI but this reminds me of the time when I was still working as a translator, I had a few jobs where I was instructed to use MTPE (Machine Translation Post Editing) and wow, that was the worst time I ever had on that job lol. I would basically have to re-translate everything from scratch because MT either couldn't keep up with the context of the game (since it doesn't know the characters or the settings) or the way it translated things sounded really unnatural. It also paid less than translating from scratch and it was easier to miss bad translations and I got a warning for missing a few. Ever since then, this topic has been a sore spot for me lol.
Being the odd one out is exactly how I feel too. AI haters unite 🤗 And oh my gosh, your experience with MTPE is just such a perfect example of the stupidity of this whole craze, it totally applies. I understand completely! Mopping up machines' mistakes is so much more time-consuming than just doing a good job yourself in the first place.
I was thinking about you when writing that passage about Clair Obscur, because you have just finished the game and because of how deeply it affected you. Here's hoping they truly don't make the same mistake again! The correct textures they patched in are so much more meaningful and beautiful than the AI weirdness that was there before, too.
And here I was, just about to finish up an article on the themes of Clair Obscur 🥲 the game is still phenomenal imo and most likely my GOTY, but it IS disappointing, but I also hope that we won't see such a thing again! And besides, the correct textures look so much more aesthetically pleasing than the AI ones too, even when you ignore the worldbuilding contents on them.
Anyway, I wanted to say thanks for bringing this discussion to Substack! It is a very nuanced topic and not discussed as much as it should be. You brought up some great points and I think they resonated with a lot of people! I want to see what the future brings and hopefully it'll be the most environmentally friendly and ethical option possible!
It IS phenomenal and it IS going to win GOTY or else I'm gonna go to LA and slap someone :D I'm very much looking forward to your all-spoilers analysis of it! The AI thing should hopefully remain only a footnote on what otherwise is a stellar experience.
Thank you so much, I appreciate it! I'm relieved to see other people feeling similarly to how I'm feeling. Only lost one subscriber so far from today's post lol
That so many businesses across various fields are rushing to embrace AI over people is just crazy to me. I understand typical CEO desire for less workers, but the whole "risking ChatGPT making up shit," could quickly turn into a serious problem for any company that has gone all in on AI.
There are good AI apps for sure, but they should be recognised as tools humans use to ease or improve their daily work, not as free human replacements.
Sadly, there are many people who gleefully view AI like this. Still, it's only going to take one or two high-profile AI-led company disasters for that way of thinking to drastically change.
Also, it feels like the pushback against generative AI in the creative industries by audiences is only getting stronger. Humans aren't dead yet!
I love this! Thank you for taking a hard stance. Especially because your point about people acting like there's nothing we can do, or that it's merely a technological future, is infuriating to me.
Let's not forget though how biases are ubiquitous in genAI. When you pull information from everywhere and give it all equal weight, we find LLMs steeped in racism, sexism, transphobia and homophobia. Hiring AIs also often reinforce old stereotypes and increase wage and diversity gaps.
Plus, of course, the constant, never ending hallucinations.
Absolutely to all of that. I didn't focus on that part in my list of why genAI sucks, but the biases it reinforces are... something else. I want to believe that the future is still not written in stone. Fight, fight, fight!
The older I get, the more I realise that most modern technology, not just AI is an enshittified version of something older.
Modern LCD TV's have worse contrast, poorer durability, worse burn-in and a poorer amount of ports/connections than CRT's but use less power and can be more easily churned out in a factory so they took over.
My specialisation in University was translation so it's one I get pissed off about quite easy. It is so obvious when something has been machine translated, it always reads like an Email from HR.
Completely agree it should be out of games entirely. If you can't be assed to make your own art and assets then I'll play somebody's game who could be bothered.
Yup, exactly. I want human-made, imperfections and all. Agreed re: ebshittified technology—"they don't make them like they used to" seems to hold too true, especially now.
Translation is such a noble craft! Which language(s) did you study it for?
It's gotten to the point where I tend to view any innovation with an instinctive, knee-jerk reaction of "Okay, in what way will my user experience be sacrificed to facilitate this?"
I was a Russian and English Literature student. I didn't complete my degree though, just really disliked my time at University and decided to cut it short.
I've tried French a few times and just couldn't get into it (despite having been there) I just... don't enjoy speaking it? I'm sure you get what I mean. Some languages are just fun to speak, Spanish is very fun to speak so far
I'm missing how easy past tense was in Russian though! It's hard to wrap my head around compared to Russian where it's so easy
I liked this post before I started reading, and I'm glad I did. Based as hell. Thanks for this, Katya. Good to know we aren't alone. These softwares are only as culturally important as we let them.
I don't even like calling LLMs "AI." Like you said, they aren't intelligent. And they can't be. They reproduce the visual appearance of text - it has no meaning and was built by no thought. Signs without signifiers, phenomena without essence. Angela Collier (physicist YouTuber) has a great video about it.
Thanks Evan, means a lot! I wasn't sure how this post would be received but I'm glad there are people who think like me out there (and here, of course).
Agree; AI is a convenient shorthand, but it's silly. The artificial part is true, the intelligence part is not. I'll check out the video!
I really think AI is going to nosedive. Because, at some point - AI is going to start training on itself. It's inescapable, once it creates enough content it won't be able to identify what it created versus what it didn't. And AI trained on AI produces the most gibberish possible.
Adore your writing style! It was a breeze to read through.
I think AI is going to be hard to avoid. Like social media the network effect is huge. I don't want to be on facebook, but I get sucked back in anyway because everyone else is. I need to arrange that play date with my daughter's Mum. I need to send a quick email, the LLM makes the first draft so easy. And so it continues....
Thank you so much 😊
It is hard to avoid already, but I really want to try. Amazon is very hard to avoid in the US/Canada, but we've been managing—for some reason, I see it as a good analogy.
As a software developer, I have started using LLMs more, as a tool to help me out of problems, where normally I would have done an internet search. I have to admit I have found some benefits to this, though I think these are ultimately structured questions/answer sessions, so makes some sense. I've never outright copied the responses I get back, because I don't trust it entirely, and sometimes it just gives a made up answer which sounds right. But in that particular context, I can see it as a useful aid.
Content generation though, as you've rightly said, is another matter. People seem to see GenAI as a "cheap" way of making something for free, except for two problems. One, there is a cost, and it's an environmental impact far exceeding any other form of computing at the moment. Two, it's just obvious that the images or videos you're looking at are fake, the uncanny valley effect immediately jumps out at you so you instinctively know it's "wrong". Yet judging by the YouTube shorts my daughter is obsessed with, people don't seem to care so long as something exists to consume.
AI may be able to replicate, or reproduce, but it can never create.
I understand that use case, Peter, and have used genAI (when requested by clients) the same way, realizing that ChatGPT was acting like a glorified Google for me. I would not mind light use like that if LLMs were trained more ethically, and perhaps at some point we'll get to that. I hope.
When it comes to creativity though... 💯 what you said. Regurgitation and no soul. Too high a price.
Strongly agree that, if GenAI is to be used in gaming at all, it should be clearly marked as such so that users can make an informed choice. I can sort of see the place for it, e.g. when used to upscale textures in old games, but these should be free updates (because it took no labour to create them), should be optional, and above all users should be clearly informed. To present AI textures as the work of humans feels like a kind of fraud, and should be treated as such. In the case of Expedition 33, I'm maybe willing to give the studio the benefit of the doubt, since it seems like the AI textures really were just placeholders that were never supposed to be in the final product, but in general I hope the public will maintain a firm anti-AI consensus in the arts.
You're also very right to talk about the role of AI as a crutch that leads to creative atrophy. Great art comes from limitations, and I can think of a number of amazing games made by solo developers who were not trained artists, but self-taught through necessity and produced amazing art as a result. For example, Miro Haverinen is not, as far as I know, a formally trained artist, but produced very distinctive, original, and beautiful drawings for Fear and Hunger. Lucas Pope wasn't trained as an artist either, but produced brilliant character portraits for Papers, Please.
I am also very surprised by how quickly so many people have accepted ChatGPT and are willing to hold even quite intimate and personal conversations with it. I have never communicated with any of these LLMs and probably never will: it's creepy to me on a visceral level, and I also suspect LLMs will privilege some (bad) sources of information over others, so they're not going to be a reliable research assistant let alone a therapist.
You know, I appreciate you pointing out that there could be a place for it—I didn't touch upon it in the article. I know it will never go away completely, and I wouldn't oppose the usage of it to automate, do QA, even playtest and push updates if need be, as long as clearly stated and not done at the expense of human jobs, the environment, and the privacy and intellectual property considerations. A tall order, but a girl can dream.
Art and creativity are a different matter. I want human-made games still. I realize that I now view games that came out before 2021 as a different beast; I tend to trust them more than new games. Inevitable, probably.
What a contrast between the comments on this piece and my LinkedIn profile.
Being in the tech industry for my main gig I’m always at the bleeding edge of adoption. I can’t escape the GenAI discussion, and navigating it in my client work where my main output is writing has been interesting to say the least. I’ve been fronting the daily challenge of: “Maybe AI can do a better job than you now” for at least six months.
As I’ve found with PR, there’s always a line, and I think the games industry is trying to find it?
Is it obvious? I’m not sure.
On this thread the answer is very clear, any AI use is a negative for the game. But in other circles where I game, I’ve raised this issue and gotten a more muted response? I do agree that 11 Bit’s PR here was tone deaf to say the least!
Clear pieces like this and the response you are getting for it help a lot in adjusting the industry’s barometer.
I’m now more curious than ever to get the Aus game industry’s take on AI — if I can.
Having come to the journalism shores from the tech industry as a marketer myself, I know exactly what you mean. I still occasionally pop onto LinkedIn, and the tone about GenAI over there couldn't be more different, so it's a breath of fresh air to hear the response to my piece on Substack. However! I recognize that we're in a bit of an echo chamber, both here and there—I'm not surprised to see like-minded readers here agree with my stance, just as I know it's a much different ballgame on LinkedIn.
Balance isn't something I touched on in the piece, but I agree, we'll eventually have to find it, and I hope we will. I'm under no illusion that GenAI will disappear completely; I hope and want to fight for a more ethical and just version of it that doesn't do harm.
Please do write an Aus lens take on AI, too, I'd love to read it!
Will do! But thank you for penning this and setting barometer on the subject too.
It helps my work immensely in understanding what to ask, and how to ask it!
I'm gonna yell about this until the day I die: art is good because it connects me to the person/people who made it. If it was made by genAI, why would I care? We are definitely in a period where a healthy dose of skepticism seems to be necessary when consuming reading, playing, watching almost anything.
I'll admit I even had my suspicions of Sandfall because of the small team, scope of the game, and when doing some digging right after the announcement, I found a passage in their mission statement along the lines of "... using the latest technologies to make our games," which to me was kinda saying it without saying it. Now it seems that's confirmed. That they patched it out at least is a positive sign and they certainly have the resources after the game's success to not feel the need to do it again. Or at least we can hope.
You're wiser than many of us, Josh! In hindsight, the line about the latest technologies does sound sus as hell, although pre-genAI I would have taken it to mean Unreal Engine 5.6 or something, not shitty abstract visuals "painted" by DALLE-E.
Agree about art. Give me all the imperfections of the human behind it. I don't want machine-generated art because it's not art worth consuming, or art at all.
Thank you for writing this! I can't tell you how lonely I sometimes feel when it comes to GenAI. People are really okay with rotting their brains and remain nonchalant over the simple fact GenAI is theft! I'm disappointed to hear about Clair Obscur as I haven't bought or played it yet! Gah! 😭
My pleasure 🩷 So happy it resonated, although I had no doubt it would with at least a few special peeps on here!
Maybe we can forgive Clair Obscur this one time. Maybe. Wish they apologized or something, but still.
Yeah! It can be a daunting subject to broach, but you did a brilliant job! 💖 The lack of apology from the Clair Obscur devs must be really grating, especially for fans 😩 It's great to see so many gamers won't stand for it though!
I am probably going to get some flak for this but there is one way that I think an LLM could be used in games -- to power NPC conversations.
This isn't to cut out games writers -- but to enable them to create the rich world in their vision, which might otherwise be technologically impossible to create. I imagine the writers playing a director role, deciding the role that each NPC will play and how they might act. There may be certain lines that need to be said, but the rest is "up to the actor (LLM)" in the same way a director might leave the nuances of the acting up to the actor.
It doesn't get around the ethicalness of being trained on data that it shouldn't have accessed or the environmental impact that using LLMs have. But for me, genAI is a human-directed filler of content, which I think, if we can figure out the ethics and environmental impact, would be a really cool way to give writers superpowers.
Just to be clear, I'm in total agreement with you about using genAI assets or text straight up, especially at the expense of human input. I'm looking to be entertained by humans, not robots!
I'm all for an ethically trained and energy-efficient LLM that would allow NPCs to have unscripted conversations, that would be very cool, but I'm having a very hard time imagining how it would be possible to make one like that. I'd imagine that an LLM would have to be in-game world-specific and trained on human-written content specifically for the game to be immersive... and that would be so prohibitively expensive I don't think any studio would go for it. We can dream though!
Agreed on all points. The nonchalant way that the people around me talk about GenAI is kind of unsettling. I'm a little tired of being the 'odd one out' in these conversations.
The Clair Obscur one made me sad, though, especially since I've been raving about the game so much. I actually did check out those posters when I was playing just a while ago and they containted nice little easter eggs, so I had felt like I learned more about the world of Clair Obscur. But to learn that they actually used random AI assets at first just leaves a bad taste in my mouth. Like you, I'll be giving Sandfall the benefit of the doubt for the same reasons.
It's not exactly GenAI but this reminds me of the time when I was still working as a translator, I had a few jobs where I was instructed to use MTPE (Machine Translation Post Editing) and wow, that was the worst time I ever had on that job lol. I would basically have to re-translate everything from scratch because MT either couldn't keep up with the context of the game (since it doesn't know the characters or the settings) or the way it translated things sounded really unnatural. It also paid less than translating from scratch and it was easier to miss bad translations and I got a warning for missing a few. Ever since then, this topic has been a sore spot for me lol.
Being the odd one out is exactly how I feel too. AI haters unite 🤗 And oh my gosh, your experience with MTPE is just such a perfect example of the stupidity of this whole craze, it totally applies. I understand completely! Mopping up machines' mistakes is so much more time-consuming than just doing a good job yourself in the first place.
I was thinking about you when writing that passage about Clair Obscur, because you have just finished the game and because of how deeply it affected you. Here's hoping they truly don't make the same mistake again! The correct textures they patched in are so much more meaningful and beautiful than the AI weirdness that was there before, too.
And here I was, just about to finish up an article on the themes of Clair Obscur 🥲 the game is still phenomenal imo and most likely my GOTY, but it IS disappointing, but I also hope that we won't see such a thing again! And besides, the correct textures look so much more aesthetically pleasing than the AI ones too, even when you ignore the worldbuilding contents on them.
Anyway, I wanted to say thanks for bringing this discussion to Substack! It is a very nuanced topic and not discussed as much as it should be. You brought up some great points and I think they resonated with a lot of people! I want to see what the future brings and hopefully it'll be the most environmentally friendly and ethical option possible!
It IS phenomenal and it IS going to win GOTY or else I'm gonna go to LA and slap someone :D I'm very much looking forward to your all-spoilers analysis of it! The AI thing should hopefully remain only a footnote on what otherwise is a stellar experience.
Thank you so much, I appreciate it! I'm relieved to see other people feeling similarly to how I'm feeling. Only lost one subscriber so far from today's post lol
With how things are these days, I'd say losing only one is pretty good :P jokes aside, it's good to see so many supportive comments!
That so many businesses across various fields are rushing to embrace AI over people is just crazy to me. I understand typical CEO desire for less workers, but the whole "risking ChatGPT making up shit," could quickly turn into a serious problem for any company that has gone all in on AI.
There are good AI apps for sure, but they should be recognised as tools humans use to ease or improve their daily work, not as free human replacements.
Sadly, there are many people who gleefully view AI like this. Still, it's only going to take one or two high-profile AI-led company disasters for that way of thinking to drastically change.
Also, it feels like the pushback against generative AI in the creative industries by audiences is only getting stronger. Humans aren't dead yet!
Heck yes we're alive! Long live humans ✊
I didn't have a good spot to link this article, but it's such a great example of why AI replacing workers is a bad idea and is, to be quite charitable, premature: https://www.pcgamer.com/software/ai/anthropic-tasked-an-ai-with-running-a-vending-machine-in-its-offices-and-it-not-only-sold-some-products-at-a-big-loss-but-it-invented-people-meetings-and-experienced-a-bizarre-identity-crisis/ (holy cow what a long slug)
I love this! Thank you for taking a hard stance. Especially because your point about people acting like there's nothing we can do, or that it's merely a technological future, is infuriating to me.
Let's not forget though how biases are ubiquitous in genAI. When you pull information from everywhere and give it all equal weight, we find LLMs steeped in racism, sexism, transphobia and homophobia. Hiring AIs also often reinforce old stereotypes and increase wage and diversity gaps.
Plus, of course, the constant, never ending hallucinations.
Absolutely to all of that. I didn't focus on that part in my list of why genAI sucks, but the biases it reinforces are... something else. I want to believe that the future is still not written in stone. Fight, fight, fight!
Corey Doctorow, a Toronto writer, is a great author on this stuff especially - and almost all of his books are free on his website.
Brilliant piece, fuck yes! I absolutely agree on your points
Thank you Alex!
The older I get, the more I realise that most modern technology, not just AI is an enshittified version of something older.
Modern LCD TV's have worse contrast, poorer durability, worse burn-in and a poorer amount of ports/connections than CRT's but use less power and can be more easily churned out in a factory so they took over.
My specialisation in University was translation so it's one I get pissed off about quite easy. It is so obvious when something has been machine translated, it always reads like an Email from HR.
Completely agree it should be out of games entirely. If you can't be assed to make your own art and assets then I'll play somebody's game who could be bothered.
Yup, exactly. I want human-made, imperfections and all. Agreed re: ebshittified technology—"they don't make them like they used to" seems to hold too true, especially now.
Translation is such a noble craft! Which language(s) did you study it for?
It's gotten to the point where I tend to view any innovation with an instinctive, knee-jerk reaction of "Okay, in what way will my user experience be sacrificed to facilitate this?"
I was a Russian and English Literature student. I didn't complete my degree though, just really disliked my time at University and decided to cut it short.
Yeah, I hear that. Enshittification made us all cynical
That's right, Russian and English! The two best languages, basically 😄
Basically covers the entire Northern Hemisphere since English is so common in the EU now, so useful, am working on picking up Spanish next!
Do it, Spanish is my third most used language. It makes travel to so many destinations easier. My French and Italian are weak for a reason 🤣
I've tried French a few times and just couldn't get into it (despite having been there) I just... don't enjoy speaking it? I'm sure you get what I mean. Some languages are just fun to speak, Spanish is very fun to speak so far
I'm missing how easy past tense was in Russian though! It's hard to wrap my head around compared to Russian where it's so easy
I liked this post before I started reading, and I'm glad I did. Based as hell. Thanks for this, Katya. Good to know we aren't alone. These softwares are only as culturally important as we let them.
I don't even like calling LLMs "AI." Like you said, they aren't intelligent. And they can't be. They reproduce the visual appearance of text - it has no meaning and was built by no thought. Signs without signifiers, phenomena without essence. Angela Collier (physicist YouTuber) has a great video about it.
Thanks Evan, means a lot! I wasn't sure how this post would be received but I'm glad there are people who think like me out there (and here, of course).
Agree; AI is a convenient shorthand, but it's silly. The artificial part is true, the intelligence part is not. I'll check out the video!
I really think AI is going to nosedive. Because, at some point - AI is going to start training on itself. It's inescapable, once it creates enough content it won't be able to identify what it created versus what it didn't. And AI trained on AI produces the most gibberish possible.
Stands to reason that this might happen, and I can't bloody to wait to see it all crumble