AI Megathread
-
@Roz said in AI Megathread:
If you program a calculator, you instruct it on immutable facts of how numbers work.
I mean… kinda? A calculator app doesn’t really know math facts the way a third grader does. It doesn’t intuitively know that 1x1=2. It just responds to keypresses, turns them into bits, and shuffles the bits around in a prescribed manner to get an answer.
I don’t point that out to be pedantic, but just to further contrast it with the way a LLM handles “what is 1+1”. Like you said, it’s based on statistical associations. It may conclude that 1+1=2 because that’s most common, but it could just as easily land on 1+1=3 because that’s a common joke on the internet. LLMs contain deliberate randomization to keep the outputs from being too repetitive. This is the exact opposite of the behavior you want in a calculator or fact-finder. And if you ask it some uncommon question, like 62.7x52.841, you’ll just get nonsense.
Now sure, some GenAI apps have put some scaffolding around their LLMs to specifically handle math problems. But a LLM itself is still ill-suited for such a task. And when we understand why, we can start to understand why it’s also ill-suited for giving other accurate information.
-
@MisterBoring now I have to add to my +finger I don’t use LLMs to write my poses sometimes I just suck as a writer
thx -
@Faraday said in AI Megathread:
@Roz said in AI Megathread:
If you program a calculator, you instruct it on immutable facts of how numbers work.
I mean… kinda? A calculator app doesn’t really know math facts the way a third grader does. It doesn’t intuitively know that 1x1=2. It just responds to keypresses, turns them into bits, and shuffles the bits around in a prescribed manner to get an answer.
Yeah, sorry, my point is more that – computers know as much or as little as they’re programmed to know. The calculator is given strict rules to calculate input, whereas LLMs are literally just guessing at a probable answer.
-
@Faraday said in AI Megathread:
But a LLM itself is still ill-suited for such a task. And when we understand why, we can start to understand why it’s also ill-suited for giving other accurate information.
Sorry for double post, but wanted to add:
When we understand why it’s bad at giving accurate information, then we have to ask: So what IS it good for? And apart from a few niche word-parsing and pattern-matching tasks, the only answer I’ve seen is: replace the humans who generate content (artists, authors, customer support agents, narrators, etc.) with a machine that generates worse content. And that core idea is fundamentally the problem I have with GenAI.
-
@Faraday Getting a computer as close to passing the Turing Test than any other computer has managed before. That’s about it.
-
AI cannot improve your writing. I’m not saying the LLMs are bad at writing, or worse than you are, but I am saying that if you use AI it’s no longer your writing. Being that I am only RPing with you to interact with you and read your writing, kindly keep that shit away from me.
If you want a sentence to hit different then think harder.
-
I’ve said it before (possibly in this thread, idk, it’s been A Year) and I’ll say it again, all I want from someone using ChatGPT to write poses/descs/thematic stuff is disclosure that they’re doing that. Now, I want this because I don’t want to engage with LLM content in my travels through this hobby and it saves me time and frustration, but at this point this stuff isn’t going away and clearly a lot of people don’t care. So as long as it’s on the tin there’s not much I can complain about.
What I have observed is that nobody doing this, in even in very blatant ways, cops to it.
Which is something I find worth interrogating. People are a lot more upfront about using Midjourney and the like to create PBs and wiki art, maybe because they think it’s more obvious. I guess there’s a level of embarrassment involved that isn’t present for some people when it comes to visual art, idk.
-
While I’m not trying to change anyone’s opinion, I thought I’d share my personal experiences with generative AI accusations both in MU environments and the real world.
First, I’m an author by trade. I make money and pay bills by writing and selling novels. It’s not my sole source of income yet, but I’m hopeful it will be one day. Sadly, the “anti-AI” witch hunt often impacts authors like myself who don’t even use such tools.
I’ve received negative reviews claiming my work is “AI slop” for the following reasons:
- Using the phrase “with practiced ease” once
- Publishing 2 books in a 4-month period
- Having chapter lengths that are too consistent
- Creating covers that are “obviously AI”
I don’t know what to say about the “with practiced ease” comment. It is what it is, I guess. As for the other points, I write every day and can produce 3,000-8,000 words daily. I’m also very particular about my novel structure - I try to keep chapters below 2,500 words and my novels around 80,000 words. That’s just how I work.
I’d like to think these reviews haven’t negatively impacted my growth as an author, but I’ll never really know. I’ll never know how many people checked out my book on Amazon, read those reviews, and said “Oh, AI slop? Forget that.”
I know artists in the same boat. Take my covers, for instance - my wife hand paints all of them. Every single one. Yet I still got a negative review because the cover was “obviously AI.” I know many creators who’ve had their work criticized as “possibly AI.” Even when you show documented proof of every step in the creation process, people still attack you.
Finally, I do know people who use AI to touch up their RP poses. For many reasons, most of them come down to insecurity and simply wanting to tell better stories. They’re not mustache-twirling villains looking to ruin the MUSH world—they’re real people trying their best to engage meaningfully with others. The poses still come from a human with emotions, motivations, and a genuine desire to connect, regardless of what tools they used to craft them.
Why don’t they tell everyone they’re doing it? Looking at some of the responses in this very thread gives me a pretty clear answer—they justifiably fear being attacked, called names, having their motivations questioned, and being treated as morally bankrupt or creatively empty. There’s a world of difference between encouraging AI-free spaces (which is completely valid) and demonizing the people who use these tools as if they’re somehow less human or less worthy of respect in creative spaces.
I guess my only point is, the AI witch hunt can be just as damaging to people as AI itself. I think it’s perfectly valid to question AI and all things related to it, but we cross a line when we attack people for using it.
And this constant looking for AI in everything—even when it isn’t there—creates a toxic environment where genuine human creativity gets dismissed, artistic choices get questioned, and people’s enjoyment of creative spaces gets diminished. We’re reaching a point where people are so busy hunting for AI that they’re no longer engaging with the actual content or the humans behind it.
-
@Raistlin said in AI Megathread:
First, I’m an author by trade. I make money and pay bills by writing and selling novels. It’s not my sole source of income yet, but I’m hopeful it will be one day. Sadly, the “anti-AI” witch hunt often impacts authors like myself who don’t even use such tools.
That’s a bummer. Blame AI and the people using it to churn out slop.
The poses still come from a human with emotions, motivations, and a genuine desire to connect, regardless of what tools they used to craft them.
They don’t, though? If someone’s using AI to write their poses they’re not coming from a person, they’re coming from a bullshit machine, and anyone trying to pass it off as anything else is showing such absolute disrespect for their scene partner that I can’t even put it into words. Why would I bother to read something they couldn’t be bothered to write?
-
@Raistlin said in AI Megathread:
they’re real people trying their best to engage meaningfully with others. The poses still come from a human with emotions, motivations, and a genuine desire to connect, regardless of what tools they used to craft them.
No, that’s not genuine. That is literally disingenuous, particularly when it is undisclosed to other participants.
Engaging meaningfully requires thought, creation, intentionality, and vulnerability, not throwing half an idea into the word blender to come out with whatever sounds slick in the algorithm today.
As @Flitcraft said, if they can’t spend the time and effort to write something, why should I spend time and effort interacting with it?
-
This response perfectly illustrates my points. The focus is squarely on continuing the witch hunt rather than considering how it impacts real creators or understanding the actual humans you’re supposedly trying to engage with on a game. There’s no empathy for false accusations or collateral damage to innocent writers.
I’m not saying anyone should RP with anyone else for any reason. There are a host of reasons to not RP with someone, including using AI for any part of the RP process. I know people who won’t RP on Ares games that focus on async RP. That’s fine. Everyone is entitled to their own boundaries.
My point is how people address those boundaries. There’s a world of difference between “I prefer not to RP with people who use AI tools” and attacking others as “disrespectful” or their writing as “bullshit.” Some are simply using AI as an excuse to engage aggressively with other people. That’s what I feel is wrong.
Thankfully this forum allows us to unfollow/watch threads. I’ll be doing that here because there clearly is no actual intent to have a meaningful discussion—just shaming, name calling, and attacking. I have zero interest in participating in that kind of environment.
-
@Raistlin To be fair, I think it started with the intent to have meaningful discussion. But it just devolved into hyperbolic rhetoric with all the name calling and shaming after a while. Sometimes people who “feel passionately” about a subject justify less-than-great actions with their emotions.
Maybe that’s most threads that go on long enough. But I do think it started with good intent.
-
@Raistlin @Warma-Sheen
I also think it’s a trust issue, and a trust issue that is almost entirely in the “you just have to trust me” category. Just like trusting the person on the other side of the screen isn’t a stalker who keeps changing their identity so they can chargen a new character to interact with you or trusting that people you are having a private scene with or worse TSing aren’t sharing those logs with their friends to laugh at your writing, there isn’t a way to outright protect yourself and verify that it isn’t happening. It’s too easy for the other person to obfuscate. So instead of allowing for grey area, people have to make it a firm black or white issue, because if there is grey area, when you catch someone, it becomes a whole ordeal figuring out if what they did was wrong. When it’s black and white, there doesn’t have to be any time spent on nuance. -
@Raistlin said in AI Megathread:
rather than considering how it impacts real creators
@Raistlin said in AI Megathread:
or understanding the actual humans you’re supposedly trying to engage with on a game. There’s no empathy for false accusations
Of course there is. Most people (in this community) know that AI detector tools are not a silver bullet. Most people would feel badly upon finding out that someone accused of using AI had not, in fact, used AI, in the same way that all false accusations garner empathy if they are disproved.
What most people are expressing is the desire for transparency, to know if AI was used to create the content they are engaging with in the MU* space, and if there was more of that, I expect there would be less false accusations to go along with it.
he/him
-
@Raistlin said in AI Megathread:
There’s a world of difference between “I prefer not to RP with people who use AI tools” and attacking others as “disrespectful” or their writing as “bullshit.”
I don’t personally feel that calling a behavior “disrespectful” is a personal attack. I feel that online piracy is disrespectful to creators. I push back against those who feel that it’s a “victimless crime”. That doesn’t mean I think everyone who pirated GoT rather than pay for HBO is evil incarnate.
In other news, ChatGPT fails hilariously at MapMaking 101.
-
I just desperately want a solid science / math AI to do something newsworthy, like invent a 3d printer resin with no toxic fumes, or cure a disease. I feel like by the time that comes around, people are going to distrust AI in general so much that they just ignore it. I can see it now:
Science AI model releases an open source schematic for a water powered drive system for cars that gets 450 miles to a gallon of water and can drive at freeway speeds. crickets
Proud Member of the Pro-Mummy Alliance
-
@Trashcan said in AI Megathread:
Most people would feel badly upon finding out that someone accused of using AI had not, in fact, used AI, in the same way that all false accusations garner empathy if they are disproved.
Just to touch on this note a little, without directly correcting you in particular: While it’s true that most people would feel bad upon finding out it was a false accusation, I imagine that actually finding that out would be extraordinarily difficult. The lie, or false accusation in this case, being halfway around the world before the truth has even gotten its pants on. With the additional problem being that any ‘evidence’ offered to the contrary would likely be deemed insufficient or an additional lie. So making those accusations should be done with utmost care and the expectation of being wrong, if they’re made at all.
-
@MisterBoring said in AI Megathread:
Science AI model releases an open source schematic for a water powered drive system for cars that gets 450 miles to a gallon of water and can drive at freeway speeds. crickets
That’s not driving advertising dollars and making anyone’s money. ChatGPT does…somehow…
Honestly, I see a lot of promise in narrowly-tailored AI designed to wade through data and do pattern-recognition, but that’s not atm a money-maker for anyone.
I want something else to get me through this
Semi-charmed kinda life, baby, baby
I want something else, I'm not listening when you say good-byeShe/Her or They/Them
-
@Third-Eye Agreed. I recall reading somewhere that an AI program could detect the very, very, very early stages of some kinds of cancers far more quickly and consistently than a human. That’s the sort of thing AI should be used for, not art.
-
@Third-Eye said in AI Megathread:
@MisterBoring said in AI Megathread:
Science AI model releases an open source schematic for a water powered drive system for cars that gets 450 miles to a gallon of water and can drive at freeway speeds. crickets
That’s not driving advertising dollars and making anyone’s money. ChatGPT does…somehow…
Honestly, I see a lot of promise in narrowly-tailored AI designed to wade through data and do pattern-recognition, but that’s not atm a money-maker for anyone.
And to say this (bolded part) is beyond ironic is a colossal understatement. How much money would be saved having a properly trained AI do the work of what would otherwise be thousands of data analysts? Part of the point of an AI is to be able to do the things a human cannot do, because a computer’s speed is significantly higher.
AI has potential, but oh boy are “we” more determined to churn out slop than quality with it, right now.
