Greater Than Zero: The Anti-AI Pushback On Gaming Preservation Efforts Makes No Sense
from the the-enemy-of-the-good dept
There is an old axiom you will have heard of before: don’t let the perfect be the enemy of the good. If we wanted to boil this down to a math equation, it might be described as something like: 0 < any positive integer. It’s not a difficult concept to grasp, typically, until you add in a dash of near-religious ideology into the equation. And that’s where the anti-AI crowd comes in.
Dustin Hubbard heads up Gaming Alexandria, a site dedicated to the preservation of obscure corners of video game history. Focused less on the actual games themselves, Gaming Alexandria instead focuses its efforts on media surrounding those games, such as manuals, box art, and old gaming journalism outputs. To that end, Hubbard’s group has amassed an impressive number of Japanese magazine scans throughout the years. To make this content useful to researchers elsewhere, he built a low-footprint app to make those scans searchable and, more importantly, to translate them. A Patreon page and subscriptions partially funded all of this.
And that’s what had Hubbard issuing apologies over this past weekend.
A day after that project went public, though, Hubbard was issuing an apology to many members of the Gaming Alexandria community who loudly objected to the use of Patreon funds for an error-prone AI-powered translation effort. The hubbub highlights just how controversial AI tools remain for many online communities, even as many see them as ways to maximize limited funds and man-hours.
“I sincerely apologize,” Hubbard wrote in his apology post. “My entire preservation philosophy has been to get people access to things we’ve never had access to before. I felt this project was a good step towards that, but I should have taken more into consideration the issues with AI.”
And this is where we enter the realm of the silly. I’m not some AI evangelist. I fully recognize that there are error and other problems with AI… and I imagine there always will be, to some extent. AI is not always, or perhaps even mostly, the right tool to use. Nor will it always have benefits that outweigh problems it creates for we human beings.
But a positive number is greater than zero. This was a tool that suddenly made all of this culture content accessible to a wider range of people. Before it was not available to anyone that didn’t have a high-level of knowledge on the Japanese language. Translation errors also happen with human translators, too. We need only look at the ancient religious texts, and the very real wars started over their translations, to understand that.
Hubbard himself attempted to make this point over the weekend.
Writing on Patreon this weekend, Hubbard said he has long been tinkering with an improved automated OCR and translation process that could help turn more of those magazine scans into useful tools for Western researchers. And when he put Google Gemini AI model to the task recently, he said he was “blown away” by the results. While he still recommended using a professional human translator before citing these magazines in any scholarly research, he said the output from the Gemini AI tool “gets you a large percentage of the way there quickly.”
Inspired by those results, Hubbard set to work on a self-described “vibe coded” interface to view the original PDF scans alongside their AI-generated text translations for easy comparison and editing. The result was the Gaming Alexandria Researcher tool, posted to GitHub on Friday and shared with the site’s Patreon backers as a “beta” on Saturday. The tool, which runs locally on Windows, Mac, or Linux, can search, download, and edit Gaming Alexandria’s files from the cloud or sort through local files stored on your own machine.
“This app has been something I never would have dreamed could exist,” Hubbard enthused. “Now I can finally read and enjoy these Japanese magazines I’ve been scanning for years. A large part of that is due to your believing in my work and funding me so thank you so much for that.”
The negative responses he got for all of this are wild. There were calls to boycott the project. Calls to rescind Patreon subscriptions. Max Nichols, a game designer, cancelled his own Patreon membership and decried the project as “worthless and destructive”, likening any output generated using AI-based translations as “looking at history through a clownhouse mirror.”
I would argue that I’d rather get that look than get no look at all. I’d also argue that we need to see very specific examples of AI-created translation errors to understand just how grounded these criticisms are in reality, versus all of this being a case of overstating the case.
Some fans of the site, at least, managed to understand the context here.
For some supporters, though, using machine translations—including ones aided by AI models—is a practical necessity given the size of the task at hand. “There’s no world in which they could ever get hundreds of thousands of pages translated by hand,” game preservationist Chris Chapman wrote on social media. “Error-prone searchability is more useful to more people than none at all.”
“Famitsu alone is over 1,900 issues, each with [a hundred-plus] pages,” journalist and author Felipe Pepe noted. “That’s one magazine from one country. [Human translation] would be ideal, but it’s impossible.”
On the Gaming Alexandria Discord, user asie wrote that people who use tools like Google Lens or DeepL are already using AI-powered OCR and translation tools. At this point, these kinds of tools are “just a fact of reality,” they added.
Again, any positive number is greater than zero. Don’t let the perfect be the enemy of the good. Something is better than nothing.
I don’t know how to explain the negative responses here as anything other than a ideological commitment to disliking anything that even remotely touches upon artificial intelligence. Absolute moral stances certainly have their place, but they sure ought to be used sparingly.
And this particular stance is silly.
Filed Under: ai, ai translations, preservation, translations, video games
Comments on “Greater Than Zero: The Anti-AI Pushback On Gaming Preservation Efforts Makes No Sense”
There are other tools
You don’t need to use AI for this.
OCR and translation tech have existed for years before AI did. Use those.
So yes, AI is stupid and shouldn’t have been used.
Using existing tools would produce no worse results, without supporting the billionaires and their AI.
Re:
Using existing tools would be cheaper and produce better results, but Techdirt, as a platform, can’t decide what opinion to have about AI and thinks that wildly bouncing between extremes is ‘neutral’.
Re:
Except those tools are inherently powered by “AI” (note the air quotes) and always have been.
Thank you for giving the class an example of the ideological rhetoric cited in the article.
Re: Re:
I don’t think it’s fair to blame Phoenix that we have conflated generative AI with AI in general (also I will not say “GenAI”, because LLM boosters also try to push the idea that they’re going to reach AGI).
But the conflation is unfortunate that more appropriate forms of AI get caught up in the backlash, while people who are not so against AI get funnelled towards generative AI.
Re: Re:
If you want to backport the definition of “AI”, or otherwise equivocatate types of “AI” for rhetorical purposes, that’s a you problem.
Re:
A.I. still doesn’t exist, and the “translation tech” you mention is what was being marketed as “A.I.” a decade or two ago.
Re: Re:
Yeah “AI” claims have always been hyperbolic puffery. LLMs are merely the latest grift.
Re: Re: Re:
“Fake it till you make it”, as they say; or “you only find out who is swimming naked when the tide goes out.”
It was only after the sea of grifters had moved on from neural networks that people quietly developed useful machine translation tools based on those networks. And I guess once the hype of “intelligence” from generative transformers has died down (i.e., when yet another “A.I. winter” begins), we’ll find a few things those transformers are actually kind of useful for.
Re: Re:
No, it was not called “AI” a decade or two ago.
Re:
I know what you mean but that’s a funny way of saying it. OCR and machine translation are based on ML; they work more or less the same as modern LLMs, just on a much smaller scale. I don’t think it really makes a lot of sense to categorize one as AI and the other as not-AI; what we’re really talking about is the tradeoff between quality and efficiency, and whether we want to support products like Gemini in the first place.
Re: Re:
And now you knowingly and specifically equivocate between rather distinct forms of “AI”, which have entirely different sets of costs. Good work.
Re: Re: Re:
What’s a useful definition of “A.I.” that differentiates between these things, and future things, and isn’t just a regurgitation of marketing points?
As has been mentioned on Techdirt from time to time, hucksters have been applying that term to a succession of disparate things since the 1950s. Without any obvious logic other than “we think calling it A.I. will get us funding”. And then it all falls apart—but research continues, and sometimes a useful thing later appears, which nobody wants to call “A.I.” because that’d just remind everyone of the failed promises of the past.
How about this.
Honestly lay out ALL the pros and cons of AI usage today. The slop, the undressing, the energy costs, the lost jobs, the shit apps, the black box denial of insurance, rising costs, and on and on.
Again, all of it. And honestly ask yourself if the minimum amount of pros that have been seen so far, are truly worth all the negatives.
Take a step back, and tell me what the difference is between the orange nazi and his nacho minions telling everyone the war is good, cheap, and we will all love it, and every single executive in AI telling consumers that we will all love it, it’s great, and we will benefit greatly.
AI isn’t going to go away, but what we have right now is a house of cards doused in gasoline. People are going to have their info stolen. Companies are going to be hacked. Technical issues are going to blow up massively.
Yeah, we should reserve them for things like billionaires wasting massive quantities of energy and water to inflate a bubble that’s going to trigger simultaneous environmental and economic catastrophe.
Boy, if something like that were happening I bet people would be really pissed about it.
Re:
Less sarcastically: I do actually see translation as an area where LLMs are a good fit; no substitute for a skilled human translator, but in many cases, as you say, better than nothing.
But I’ve got very little patience for this “But why would anyone be categorically opposed to using LLMs? Why are people so silly?” affectation you keep doing. If you don’t understand, it’s because you’re not listening.
Re: Re:
Because the goal of AI boosters is to smear all their opponent as unreasonable, hence their continual lumping of the Roko’s Basilisk doomers in with people with moral and environmental concerns, or tossing around terms like Luddite until people point out exactly why they did what they did and the ugly parallels with the present day.
This wouldn’t have caused as much backlash if Patreon donors were told ahead of time that this is a direction they were looking at and why, or decided to try and fund it as a separate project but instead they ploughed ahead and spent donor money on it because consent is never asked for when it comes to AI usage.
Re: Re: Re:
I’ll tell you why nobody is listening to ‘environmental concerns’ in turn – it sounds disingenuous to downright moral panicked. When shared common sentiments are both that even a data center is ‘too much’ and that ‘we need to bring back manufacturing’, it is no surprise they aren’t taken seriously when bringing back manufacturing would involve far worse than a data center. It is like saying you want plastic surgery but freezing off a wart would be too invasive.
Re: Re: Re:2
Nice, now tell me, are the people saying ‘we need to bring back manufacturing’ and the same people complaining about data centers one and the same because if they aren’t, you are putting words in people’s mouths. I somehow doubt that the minority communities who are fighting to stop data centers being built on top of them are the same ones giving grand political speeches about bringing back, or the ones who have learned too late that the data center they were told would use little water and power, would be clean and quiet will actually cause bills to skyrocket, run illegal gas turbines and sound like living next to Mordor.
You can write off a lot of views if you’re willing to cast a net wide enough to lump people into a collective ‘we’ to create hypocrisy, maybe don’t do that.
Re: Re:
How else can he show Masnick that he’s a very good boy?
Re: Re:
Geigner can’t listen; according to the Insider Chat, he’s too busy being wowed by the jangling keyring that is the DLSS 5 slop filter that Nvidia is making.
If the project is going to pull in funds from Patreon, then it makes sense to just hire translators to do the work. No, you aren’t going to get everything done overnight, but you are going to start working on the project piecemeal. With the funds, have someone translate a single issue, for instance. I’m honestly sure there are those out there who would be interested in doing some of that work for free because they want something to add to their resume. As a result, you’ll gradually build goodwill and have more people contribute either financially or through translation efforts.
Sounds like this guy wanted to use a shortcut and just get results now. This knowing full well how large of a task he set out to do. I’m not dismissing the volume of work involved, but just lazily declaring it “impossible” and resorting to AI is going to foster the very negative response he ended up getting. If people contributed financially, expecting the translations to be professionally done, then found out afterwards that it’s just him babysitting AI that is, by his own admission, error prone, then this is not necessarily ideological backlash and a declaration that anything AI is “bad”, but rather, getting people to contribute to something that was not as advertised.
Re:
I can’t find exact quotes, it’s not like buying sacks of potatoes, but even if you could get a whole hundred-page magazine translated, set up for display, etc for 10 bucks each, the article was talking about individual publications of thousands of issues each- I can’t imagine this niche hobby/research material could secure enough interested parties to pay for much compared to the vast trove of material.
OCR/ simple character translation is a relatively solved problem these days, lacking the drive to make randomized hallucinations. IMO this does not need to be pitchforked in the middle of a country lane like generative slop engines.
Yes, and no
I’m all for rational use of AI, but your comparison of generative AI-based translation errors with human translation (or any other automatic translation tool that’s not based on LLMs) is way, way off.
First of all, when your translator is translating from a second language they’ve mastered (maybe by having parents who speak it, and having been embedded in the source culture for many years) to their native language, the chances of getting something catastrophically wrong are pretty much zero. Nuance, double entendres, subtle subtext, might be lost from time to time, depending on their mastery of the source language and culture, but it won’t go beyond that.
Classic auto-translation tools can get things wrong more in the way a foreign speaker can get things wrong when translating to a non-native language.
GenAI-based translations will make things up. It’s not awkward sentence structure, or using the wrong word or expression, it will generate sentences that are not, in any way, in the original text.
It not a matter of being wrong more or less often, the issue is one of failure modes. When you read a poorly translated text, you can often spot something is off. When an LLM “hallucinates” a whole paragraph, you will not notice.
If the point is making the text available to people who can’t read the original language, use classic automatic translation. It won’t be as good, but when it fails in more acceptable ways.
To clarify, I don’t know, or care, about the real-life issue in the article, just the mischaracterization of both, human, and generative-AI translation capabilities.
Re:
This comparison of human and AI errors (followed by “well, humans do the same thing that AI does, so it’s fine!” or something along those lines) has always, always rubbed me the wrong way. Mainly because those kinds of comparisons are invalid to begin with. If you want to bring the errors humans make into the picture, then we must compare what came before to what the AI equivalent brings to the table. If what the AI brings to the table is substantially similar or worse than what existed before the AI came along to “replace” it, then AI has not, in fact, made any kind of improvement whatsoever, and has in fact made things worse. And unlike AI, humans actually learn from the mistakes we make. A human is (usually) not going to continuously make the same mistake (e.g., hallucinating entire paragraphs or pieces of information) repeatedly unless they are doing it deliberately.
This is something AI boosters seem to never grasp entirely. Or they deliberately ignore it. I can’t tell which, honestly.
Re: Re:
“If what the AI brings to the table is substantially similar or worse than what existed before the AI came along to “replace” it, then AI has not, in fact, made any kind of improvement whatsoever, and has in fact made things worse.”
….wut? If AI is the equivalent error-wise as humans, but is of course FASTER than humans, then that’s better incrementally, not worse. If you think it’s worse than humans in making errors, then that’s a claim requiring evidence and proof.
“And unlike AI, humans actually learn from the mistakes we make.”
AI does this too, on most platforms, actually. And largely with the help of human beings training it. I think the opposite is true: what AI learns is more permanent than when humans learn. A human will repeat mistakes often, while a trained AI largely won’t.
Re: Re: Re:
Making mistakes faster is not an incremental improvement.
Re: Re: Re:2
“Making mistakes faster is not an incremental improvement.”
I still think this framing is nonsense, but I’ll bite: why not?
If LLMs and humans reach the same endpoint (an error, in this case), why is it not better to reach that endpoint faster? The faster an error is made, the faster it can be recognized, the faster it can be corrected, the faster it can be used to enhance and train the LLM to not make the error again.
How is that NOT an incremental improvement?
Re: Re: Re:3 Error rates matter
Regardless of the hype cycle about LLMs, one of the problems that gets swept under the run a lot is the average “hallucination” rate. The best the industry has achieved as around a 30% hallucination rate; and apparently OpenAi is regressing and published a paper recently that they are now seeing a 40% error rate. While humans make mistakes, I don’t see it as a realistic comparison to say that trained / skilled humans are living in a 40% error rate zone.
Re: Re: Re:4
I’ll just point out that your claim of 30% hallucination rate is utter nonsense. First of all, there’s no single “hallucination” rate. Because it means all sorts of different things. You can see a more thoughtful discussion than your summary here:
https://suprmind.ai/hub/ai-hallucination-rates-and-benchmarks/
It really (again) comes down to what are you using it for, how complex is the task, and how much is a human in the loop (and also what are the downside risks of an error).
For many tasks, the tool is extraordinarily useful and accurate. For others it’s terrible. Translation is one of those categories where the error rate is extremely low.
Claiming there’s a 40% hallucination rate in response to an article about a narrow use case such as translation suggests (1) you have no clue what you’re talking about (2) you’re spewing nonsense because you have an irrational hatred of the tech you don’t understand or… both.
I know some of you refuse to believe there’s a way to have a nuanced conversation about this, but posting nonsense like “oh they get things wrong 40% of the time” is WHY people dismiss some AI haters. It’s just so far from the reality of how the tech is used it makes you look stupid.
Re: Re: Re:5
That’s only because the hallucination rate is, and always will be, 100%. What you find to be useful hallucinations which accidentally reflect reality (or not) is up to you. These word machines literally function by hallucinating, it is the only thing they can do.
(Yes, the 30% thing per whatever definnition seems like it comes from nowhere.)
But the real issue is that they do not and will not provide any kind of equitable value for their costs. The costs are far too high for any kind of output value. So sure, go on with the nuance. Have not seen any value in that, either. Make that argument about how it is in any way worth it.
Re: Re: Re:6
And this is why no one takes people like you seriously.
You have no idea what you’re talking about and it makes you look very silly. No, that’s not how they operate. I get that you’re trying to make a point in that they are generating things non-deterministically, which you define as a hallucination, but that’s changing definitions to fit your desired claim, not a statement based in reality.
You can just say “I don’t know how to use these systems properly, and therefore I assume no one does” and it’s a more accurate statement. The fact that you don’t bother to educate yourself is a not a condemnation of the tech. It’s a condemnation of your ability to learn and understand.
I mean, this is just provably untrue. I’m running a local model on the machine on a desk in my office. It’s not costing me a dime. It provides quite a bit of value.
I get it, you hate the tech, refuse to learn how it actually works and so you HAVE to buy into that world that the tech has no value. Good for you.
It is true that the tech is overhyped, and many people use it badly and in ways that have serious externalities. But your rush to condemn the whole thing is just silly and it’s why no one takes people like you seriously.
Re:
Nuance is the biggest thing that gets lost in translation. When translating an anime script for an English-speaking audience, a GenAI won’t bother to ask how big an insult “This soup tastes like water” is.
Re: Re:
Or to give an actual example some people complained about in a game from 2021, a character saying “bye” and leaving a conversation with no more than that is actually extremely rude, and was translated as “later, losers” in English because the rudeness doesn’t come across without the addition of the explicit insult.
Re: Re:
I don’t think a lot of human translators are asking those kinds of questions either. A lot of professional subtitles drop honorifics entirely, or translate them in a way that adds sexism where none was present originally. Or stuff like writing a person’s name when a character was clearly calling them just “brother”.
If I’m starting to notice discrepancies when I know only a small bit of Japanese words picked up from watching anime, I figure there must be a ton of dubious stuff going on that I’m not yet capable of noticing. Hell, even in English-language media, I don’t think I’ve ever seen subtitles that fully match the audio.
Japanese/English translation is commonplace. If this was even a Japanese/Yoruba translation it might be of some value (not to mention the many languages less common than that one), but as it is nothing of value has been added. Technology has societal value when it allows us to do more than was feasible to do before, not when it merely externalizes the costs of doing what is already easy to achieve.
Flipside
Never forget…….. All Your Base Are Belong to Us. and “Somebody Set us up the bomb”
Was translated by, and “coded” by humans.
The Butlerian Jihad had the right idea, and every piece going ‘tsk-tsk people are opposing the regurgitation engines, can you imagine‘ just proves the point.
Yeah, people might be getting sloppy on the distinctions between generative AI, ML stuff like protein folding, and arguably in-between stuff like translation. But I don’t give a shit about drawing those distinctions until people stop trying to use them to jam a shitty soul-destroying tech down everybody’s throats.
Stop trying to make fetch happen.
Re:
People unironically citing slavers like the Butlerian Jihad as ‘reasonable’ is a great failure in media literacy while showing the exact same fanaticism and hypocrisy being warned against.
Them considering mechanical calculators dangerous was supposed to be, and calling computers enslaving people bad but them enslaving them good a sign of how stupid and fanatical they were!
Re: Re:
I’m not sure you understand what “Butlerian Jihad” refers to. There’s no “them”; it’s from the novel “Dune” and refers not to a group of people, but to a conflict that led to “thinking machines” being considered taboo.
Per Dune Appendix II, “The Religion of Dune”, the central tenet was “Man may not be replaced.” That appendix says nothing at all about slavery; its existence in the book’s world is unrelated to the Butlerian Jihad.
Re: Re:
What in the world are you babbling about? Are you citing some bullshit from the KJA material or something?
Something about this topic really brings out the stupid. An entire comments section of people trying to convince themselves automatic translation for pennies is a bad thing.
Re:
Oh look yet another American who doesn’t understand externalities
It is interesting to have reached a time when potentially waiting for access to entertainment, not out of oppression or with intent to cause harm, is distressing enough that those who support it are declared to be objectively, categorically silly.
Purity culture
I thought Cory Doctorow’s second most recent (William Gibson vs Margaret Thatcher) blog post was an interesting take on (And one that I happen to agree with, to be fair.) the enshittification around the discussion of any use of LLMs or machine learning, and how this is a form of purity culture that brings little of value to the discussion.
“If technopolitics were immutable – if the original sin of a technology could never be washed away – then everything is beyond redemption. Somewhere in the history of the lever, the pulley and the wheel are some absolute monsters. Your bicycle’s bloodline includes some truly horrible ancestors. The computer is practically a crime against humanity.”
Should we use AI tools to commit genocide, algorithmically discriminate and destroy society? No, of course not. Should we use those tools to translate text, in an attempt to save culture from the dumpster of history? Of course we should. Do I share in Mark Zuckerberg’s crimes just because I used I have used LLMs to translate text, used it to point me towards knowledge and scrape the web? I sure hope not.
Of course many of the people that have poured money into AI so that line go up should be exiled from society and placed in international waters to fend for themselves, but the tools they leave behind shouldn’t just be thrown away with them.
If you really want to fight this crusade regarding this preservationist’s use of AI tools, get off the Internet, go learn Japanese and do it yourself. Or give him more money so that he can hire translators. Wagging your finger at people using these tools for a good cause doesn’t help anyone.
Re:
“Oh, your bike has sprockets and gears- Just like nazi tanks, obviously. You Monster”
/s
Re:
Orrrrrr… People planning to use donors money should ask before spending it on things they know there’s a good chance a large portion of the people providing the money would object to. It is not hard to ask permission first, it is, however, nearly impossible to regain trust after spaffing away money on things people find morally objectionable.
Re: Re:
I am the person you are replying to, by the way.
The preservation patreon is currently sitting at about $250USD a month, enough to pay for decent storage for all these magazines and not much else. Is this enough money to pay for the owners living expenses, provided he translated this for free? Is it enough to pay for a reliable translator? Of course it isn’t.
It’s sitting in storage, untranslated, unable to be accessed because people are saying “I would rather have nothing, than use a tool I morally oppose to gain something of value.” I wish I had the privilege in my day to day life to not have to drive a car, not buy electronics and apparel made by slaves, not use software and services made by monsters but the reality of the world is that I must, if I want to survive.
We aren’t talking about someone spamming forums with AI generated opinions to influence politics and culture, nor stealing art and damaging the ability for artists to live, we are talking about a tool that is bridging culture and saving art.
Re: Re: Re:
And again, he did not ask the people providing that money if they were okay with using something that is by it’s very nature, unethical. We may not be talking about an AI spammer, but we are talking about someone who took money from people who believed it would be used one way and used it another way without consultation, asking people if they think it is okay to use an error prone system to save art at the expense of the people and planet that produces it… Also those scans of the original works in Japanese that people thought they were paying to have archived weren’t going to stop existing if they weren’t AI translated immediately, so all they’ve really done is make it so there’s less money for them to spend on actual archiving.
Re: Re: Re:2
You’re telling me this is an immoral technology that shouldn’t ever be used, on a PC constructed from components that have been mined, refined and constructed by slaves. Something that is, by it’s very nature, unethical. If you live in a developed/less exploited country your entire society, like mine here in Australia, relies on the outsourcing of labour to countries with a lower minimum wage. My clothes, my boots, my car, my computer, my electric toothbrush and many of the products that I use to live are made by people paid dollars per day. It sucks dude, but am I supposed to throw these out and attempt to live off the land?
Even if it was 500 USD a month, it’s still not enough for this guy to live off of and translate for free.
Machine translation is better than it ever has been, thanks to machine learning and the cost of what he has done here is negligible. I’m not using Google’s Gemini from the tap, as I don’t care to pay Google for for attempts at it, but my testing using the Gemini 3 Flash (preview) model available to me on Kagi (Which passes on requests via API to the various providers) to translate a Japanese magazine cover cost me about 0.5 US cents for a page. The Rosetta Magazine translation program gets done once, then the translated text is placed along with the original characters and given to you in an easy-to-read user interface. This might be like 4 cents a page, at worst?
Are hundreds of thousands of pages to sit in limbo for decades while 250 USD per month trickles in, waiting for a human translator? I would say you’ve gone down a ridiculous purity spiral if the answer to that is “Yes”.
Re: Re: Re:3
If you can’t do it ethically, don’t do it.
Re: Re: Re:3
Nice goalpost moving there, love the use of the ‘you complain about capitalism, yet live in a capitalist system, hmmm’ style line of thought. Keep at it, you can go pro, kid.
None of that addresses the fact that they DID NOT ASK PATREON DONORS BEFORE USING THEIR MONEY FOR AN UNETHICAL THING THEY WOULD MOST LIKELY SAY NO TO. Asking for consent before doing something is not purity testing and pretending otherwise is quite frankly horrendous.
Jesus, if an AI enthusiast starts hitting on you in a bar, folks, cover your drink when you’re not taking a sip. They clearly cannot grasp that other people have the right to say no.
Seems perfectly reasonable
I don’t understand what’s silly. The people who funded a project learned about a controversial piece of technology being used to fulfill the task when clearly they’d assumed different processes would be leveraged. Especially in the gaming space it feels tone deaf to say “stupid Luddites, sometimes AI (which doesn’t exist) is actually good.” The corporations presently plastering all of our shared reality with their hallucinating chat bots sold as the precursor to super intelligence are responsible for making the gaming hardware space largely inaccessible for most consumers. These companies are killing the era of personal computing, in addition to all the other ills they’re propagating. But yeah, #!&% those grown adults sharing their opinions about what’s being done with their hard earned resources because checks notes you think they’re silly for having concerns.
They paid for translations (one would assume accurate) of novel pieces of human expression and found that someone who wasn’t a qualified expert (otherwise I’d assume they’d be translating a bunch of content on their own) was just going to run their database of scans through an LLM and deliver that as a product, even slapping terms like “Researcher tool” into the final title. Maybe they should have just been up front about what processes they planned to leverage for this work so their patrons could make an informed decision beforehand rather than finding they funded something that creates an inferior final product compared to what one would reasonably assume (not reliable translations, no expert review) and funnels some of their contributions to a machine which is destroying their hobby.
People: Wildfires are burning everything in sight! Put out the fire!
TechDirt: Well actually fire has some uses. Why do you hate fire?
Sometimes “burn it all down” is the best decision. LLMs are an abject disaster and deserve only failure. I’m completely fine throwing the baby out with the bath water here because the baby is a bit of an asshole.
Here comes Timmy with another nuclear bad take. Surprisingly, this time he foreswears his usual “pro-consumer integrity” screed to instead tell us that we’re wrong for… expecting integrity.
Re:
Unless Dustin Hubbard has stated in the past that he would never use AI for his preservation efforts, where does the integrity come into the picture, exactly?
'As the people currently funding this project what's your thoughts on using AI?'
If you’re currently doing something and getting funding via a patronage model to do so and it occurs to you to maybe involve AI in the process the next step after that should be asking those funding you if they’d be open to that.
With it’s currently not-great/toxic reputation jumping straight to using AI and only telling people after the fact is practically begging the whole thing to blow up in your face, just like happened here, and if they’d told their backers before pulling that trigger they might have been able to avoid the mess or at least significantly reduce it.
When it comes to AI, I can recommend this book:
“Computer power and human reason”
by Joseph Weizenbaum
But the english speaking world has forgotten him and his books are out of print.