AI fan wikis riddled with errors are muddying the waters of games culture
eurogamer.net
eurogamer.net
If the solution merely makes it high friction to post AI content that is probably good enough. In other words if someone has to copy/paste the output of an LLM into a text field and even then the tech detects that and throws an error that would be a big step in the right direction.
Reddit, Wikipedia, fan wikis, reviews, Hacker News comments - I'm sure there are a dozen more categories I'm missing.
We can put these together to conclude that every game and community that lacks that kind of dedication will die out and be replaced by psychologically personalized short-form dopamine-agonist videos, as exemplified in an infamous comic:
I don't think the boundary should live at human-written or AI-written, it should just be factual.
This weekend I was playing Stalker 2 and got frustrated that the trader prices in the game seem nonsensical. I spent a lot of time trying to research the mechanic but most guides and some wikis are outright wrong. They either buy into the games marketing or just assume mechanics work as expected.
So what I did was I let Claude read through my game files. 5 minutes later I ended up with a document that very concretely explains all the factors that influence trader prices with references to the game scripts to back it up.
My point is that having AI write game wikis is fine, as long as the information is verifiably correct.
Incentives will make sure that this new web will become the old one given enough time.
The only reason there's a connection between the richness of the content, and the likelihood of it being an ad, is because advertisers attract attention by making things shiny. But if we short circuit that upstream, by not running untrusted js to begin with , then there's no reason that we can't have shiny content downstream of where the trust was verified.
Information doesn't need to be shiny.
If the platform can show shiny content, then ads WILL FIND A WAY.
I've never been served an ad over bittorrent, for instance, because it only gives me what I ask for and I only ask for things that are trusted by my peers. The ad problem is structural. If we change the structure, we change it too.
In other words, signatures don't address the fact that a human can ask an AI model for content and then sign it to pass it as their own.
The most famous attempt at a web of trust was PGP, but it just attempted to bind a public key to a name. That's too generic. Only privacy nerds are actually going to have a key signing party if that's all they get out of it.
I think that it'll go better if the we're attesting about common interests. Like, go to the event because you're enthusiastic about games, end up with some public keys as a side effect.
It'll be hard if you're the only person you know with a particular interest and it's impractical to travel to wherever people with that interest gather. Then you'll have to take a gamble on some internet handle you've never met. But really that's just falling back to the dubiously trustworthy web we currently have.
https://en.wikipedia.org/wiki/There%27s_a_sucker_born_every_...
For any of this to work we have to already be gossiping metadata about each other's content. "I think this key was compromised on Nov 1, 2026" is exactly that kind of metadata. That sort of thing rarely happened in PGP because key revocation was necessarily a side channel that you were often ignoring. But if the only way to avoid being inundated by slop is to assemble the page from the gossip of trusted peers, then the untrusted nature of the recently compromised key is gonna show up right next to its dubiously trustworthy signatures, and your browser can then interpret it without you having to do any sleuthing on your own.
How hard are you really gonna work to compromise a key that you can only use for a few days? I mean, the going rate for placing an ad on a page is like, fractions of a cent per view. Maybe it'll be more valuable in a post-ad world, but I have a hard time imagining that getting your content in front of some eyeballs for a day or two will ever be valuable enough to be worth hitting somebody with a wrench over. What you have to gain by doing so just isn't valuable enough to be worth the risk.
The architects of the original web gave up on the trustworthiness problem pretty early. All we got is : "is this server lying about its name?". We're going to have to pick that problem back up and bake solutions to it into our browsers. Except instead of attestations about server names, I'll be about the content. Was it made by somebody who is known to be sloppy with AI? Or maybe they're the sort that would put out content that doesn't disclose that its secretly an ad.
People have been gossiping this sort of stuff about each other for a very long time. If were gonna have a useful web again, we'll need to find a way to scale that capability.
I think its working so far.
Wikipedia has always been different from other sites for leaning more on human labour instead of automated systems (so-called "soft security" https://meatballwiki.org/wiki/SoftSecurity ).
Alternatively and for more general communities, the solution is invite-only with tracked invites (so it’s easy to disinvite a bot or otherwise malicious actor) and federation (if you’re invited to one group, you’re automatically trusted by many more).
It needs to be verified. That's all. That could be by a human, but it could be by AI. The point is that any content that's written by AI might be wrong, so if you're not checking it you can't trust it. AI is pretty good at checking it's own work.
I imagine with a fan wiki (or company docs in my experience) no one bothers to put the effort in. That's probably because the rewards are for writing the content, and there's absolutely no reward for checking it's correct.
If that were even remotely true AI fan wikis wouldn't be riddled with errors.
Case in point: the proliferation of furcons.
Which has been good & bad with some of the shadier smaller conventions, mismanagement and overgrowth of some of the larger conventions into worse hotel/con experiences.
But to go beyond furcons, nerd conventions in general are now dealing with gen-ai art showing up in lazy con hosts who are just trying to scrape money from the hobby. Struggling to police makers who just download common models online and try to sell budget 3d printed crap at dealer dens / artist allies, taking spaces from actual artists.
In a healthy community people will notice crap and simply won't buy it, which discourages merchants from selling crap. If that's not the case, then it's quite likely that the community has bigger issues because they cannot distinguish crap from good items.
I mean, if someone shows up with hand-made crap then the community should also push it out, whether it's hand-made or mass-produced is pretty much besides the point.
Regarding the cons themselves, of course there are good ones and bad ones. That's how it is with everything.
In some ways that’s the web, everyone posts their stuff. Or at least it was.
The unspoken assumptions that underpinned the web included the idea that it was made by humans and that content generation couldn’t be mechanized.
There are already YouTube ads for services that Humanize AI output.
The best analogy is that we are at the start of the Dark Forest era of the net.
I assume people will adapt, based on the things that don’t change. If you cannot make out authenticity of participants, then you index on correctness of argument. If the output is useful to you, then you got what you needed and can move on. A more transactional and less community oriented web.
I think fundamentally you are right. People or AI will generate artifacts of value and a system will evlove to find that value. Partially because there will likely be a financial incentive to do so.
We are no longer on a web that assumed humans first.
LLMs can shitpost like the best of us. I’ve had to point out content was AI generated so many times that its futility is apparent.
Consumers have more to consume.
A new category of producers can provide.
If the content produced is close enough to the real thing, is cheaper, then that’s enough to change consumption patterns.
You would need to identify some attribute of content that cannot be replicated at scale, for the old system to persist.
Currently the average output of an LLM matches the average output of an online conversation. It’s only when it’s pushed or put in front of an expert in a domain that its weaknesses are apparent.
Unless everyone becomes an expert, I’ll output is a good enough substitute.
Except that Astra will happily lie to your face just like it lies in the wikis. It might even "fix" things that are perfectly correct. The problem here isn't that wikis have errors, the problem is that AI gives wrong information.
I'm afraid your dystopian future where the zone is flooded with shit and nobody can trust anything seems eminently plausible, partly because the internet was built without identity.
I think to solve this you'd have to have strong proofs of identity locked to biometric identifiers. We'll probably need that anyway for other reasons.
It's not a trivial problem. Websites like this could require webauthn tied to a hardware identifier of their choice (like a fingerprint or faceid) - I'm not sure if webauthn provides that level of detail to providers (which method was used), and how secure that attestation is. Of course that's not going to be popular but it might be the lesser of two evils.
That's what surveillance-capitalism wants, but what regular people need is something which is easier to supply and less likely to be turned against them: A lesser proof of "I invested enough time/effort/money into this identity that I'll be hurt if its reputation craters".
I'm not sure I share your concern about surveillance capitalism - that ship sailed long ago - you are stored in many nation-state and corporate databases under all the aliases, phone numbers, IPs you've ever used. I sincerely doubt your opsec has always been even near good enough to evade this surveillance, it's a little late to start worrying about it now, particularly when there are other concerns.
I'd be more concerned about billionaires spending on bots impersonating humans to sway elections in my local state than a nation state snooping on forum comments and linking them with something else.
Proof of identity is badly needed in order to use the online spaces we've come to know and trust, and time is running out.
If my options are "give all my data to malicious corporations and their bots" or "wonder if the pathological liar I'm talking with is a bot" I'll stick with guessing about the liar. Proof of identity is a scam, not a solution.
We shouldn't abandon the entire idea of an internet where people can find information and speak to each other without being identified and monitored 100% of the time just because the companies who want us to be identified and monitored filled the web with bullshitting bots
Maybe you've already given up and no longer care about that freedom, but don't count on everyone else giving up theirs.
The internet does not provide anything close to anonymity and never really has. Even a simple IP address is enough to out the majority of users, when combined with fingerprinting, metadata tracing and the use of metadata from other users, there really is no possibility of hiding your identity from a determined attacker.
I would choose the opposite option personally, and would welcome it if this site for example allowed biometric identification and used that to weight comments. While I wouldn't want to force that on anyone I do think it's going to be necessary given the rise of LLMs.
If there's no possible way to hide your identity then as long as we're determined to identify humans and bots we already can. In truth, most people are able to hide their identity pretty well from most people most of the time. The government can pretty much always track down who you are if they feel like it. If you're using a cell phone apple and google can find out who you are. If you use windows Microsoft will have a pretty good idea of who you are, but the vast majority of websites can't get anything more identifying than an IP address and a user-agent unless you tell it to them without trying to correlate your traffic with other traffic. That's where browser fingerprinting comes in really, it tells them the same browser a person used to log in with their name and address at one website is using another website. That lets them make some educated guesses, but they can't be sure who is behind the screen and there are easy steps a person can take to make it even harder for them to guess.
The fact that anonymity on the internet isn't perfect doesn't make it useless. Giving that up because you're afraid of reading something a bot wrote seems crazy.
Utilizing terms like surveillance capitalism, doesn’t serve as a useful deterrent, because the underlying system has changed.
Previously Bots were piloted to be marginally effective at appearing to be normal users. With LLMs, that’s basically out of the window.
At this point your false positive detection rate is going to start hitting human users.
Users are also going to accept AI content. Simply because their feeds are going to provide it to them.
It’s cheap, and it achieves the same dopamine hits that human interaction will achieve.
Unless you are knowledgeable in a field, you wouldn’t be able to make out if something is off.
The most likely positive outcome is we see bots that take on bridging roles in conversations.
The question I would ask is why would and who would, pay to verify a user is human ?
Maybe in the future that will be true, but right now AI is constantly making very obvious mistakes and errors. That might even eventually become one way we tell if something was written by AI. If human pathological liars get flagged as AI by mistake, that wouldn't even be so bad.
There is no solution (though there will be plenty of grifters trying to sell one).
The open internet we currently know will be lost to mountains of slop just like usenet and email were even prior to the AI era.
The ones I came across just contained nebulous prose with variously incorrect info about the game on a few barebones pages, with a custom domain name. They weren't using any wiki software.
I have uBlock Origin installed so not sure if the purpose was ads but I was baffled as to what the goal would be otherwise since they're useless for proper information, let alone a wiki.
Maybe there are other people selling the tools, who try to convince the former people that what they're doing is smart and profit will roll in any day now.
Anyway it would be a mistake to think something is profitable just because someone bets on it.
AI is rotting pretty much everything that was carefully curated over the last couple of decades. If it can be expressed in bits it will be degraded. This is one of the most annoying side effects of AI, the general decrease in quality. Quality of discourse, quality in literature, quality of information, quality of software. It's everywhere.
The problem I'm already having with LLMs (in miniature) is code/comments/docs where I don't know whether it's someone's intent or whether its un-reviewed LLM babble. That creates a continual stumbling block for figuring out whether the thing should be the way it is, or whether I'm looking at an unusually-coherent error.
Take the exact same messy pile of bytes but promise: "Everything in this made sense to a human at some point in the process", and that changes how one approaches the problem.
Whether AI written code is going to be better in the longer term remains to be seen. With what I've looked at so far I'm not super optimistic about this.
To me it has become clear that LLMs will, on average, improve the code quality by a lot.
Which model did you use? I hold Opus 5.5 in quite high regard, finding it very reliable in terms of cider quality.
For example, try searching about what sort of nails you should be using to attach cedar lap siding.
On Google, you might find a few good results, but you’ll really need to scrutinize them to make sure they’re reliable. Most are slop on the ‘blogs’ of siding and roofing contractors.
On DDG, it’s even worse. Usually nailerguy.com ranks highly - it seems to be completely AI generated, complete with contradictions and impossible pictures. And, often, nailicy.com, a (seriously) nail salon in New Hampshire that I’m sure is confused about why all their web traffic isn’t translating into clients.
YouTube is about the only ‘reliable’ source for recent advice.
Don’t get me started on Google’s propensity to show a full page of unwanted shopping results now without any way to show websites unless you turn verbatim on. “Brand picks” what garbage.
The first result I looked at suggested stainless steel splitless siding nails. The Gemini overview said something similar.
So what's the actual answer?
When AI started I got a copy of 6 months before of wikipedia and open street map.
At some point you cannot trust anything because everything has been touched by everything as a regurgitation of everything.
You know it’s not gonna be updated, because it was created on a whim and the author has already moved onto 79 other projects since.
If it was all clearly completely incorrect, it wouldn't be such a huge problem, but it is plausible and hard for humans to detect in a short time unless it is long form text.
Even search providers are giving us this new kind of generated result at the top of their results, which most people will be using blindly to verify answers. So it's turtles all the way down.
If you're not a domain expert you have no idea which things are right and which are wrong so it's very hard to determine the truth or trust results.
To top all that, LLMs are being trained on this new untrustworthy data without differentiating it from older data, and trained by other LLM instances in a sort of recursive loop to improve performance on very narrow goals. What could possibly go wrong!
It’s the shit tier free or super low cost AI hallucination engines used by the masses that are the problem.
The problem here is Google. Bad websites have always been around.
For any search have always been a few or several valuable results, and more than enough bad websites to fill the first 10 or 100 or more pages of results - more than any user ever sees. Multiplying the number of bad sites by 1,000 due to people using LLMs doesn't change the search user's experience - there's still more than enough.
It's still the same problem for Google: Find the few valuable results and put them on the top. Does Google even need different tools to identify them? Link farms and spam pages are old technologies.
Perhaps the more important problem for google though is that people really like a query style interface AND response to a list of links. They don't even care if it's sometimes wrong (apparently). Google have added that, but it doesn't really fit with their current design.
The entire premise of google is now under threat.
this is such an emblematic thing for the slop peddlers
Bonus points for also filtering out paywalls.
I'm tired of search results being useless. The signal to noise ratio on internet search is much worse now a days.