If you're using AI to make content you're saving time and money but you're making shitty content. If you can't see the shittyness of the content, it's all the more reason you shouldn't be doing it.
Oct 13, 2026, 3:43 AM@grumpygamer Dunning-Kruger amplifier?
@grumpygamer the word I have been using is "Mediocracy"
@grumpygamer
There are cases when scientists publish papers that they have written with ai and don't even understand themselves. But the papers are good. That happens a lot as humanity has already produced a lot of knowledge and just hasn't connected the dots a lot of times.
That's is a good example why you are wrong with your absolutist statements. Good content can and does come from ai. You just hate ai.
@Noisecolor No good scientist is releasing a paper they don't understand. Doing so makes them not a good scientist. You thinking that is true kind of proves my point.
@grumpygamer
Your argument comes from your feeling.
My argument is sort of a truism: If the content is good, the content is good.
You can debate the ethical side of the source, but the actual content.
Oct 13, 2026, 5:09 AM@Noisecolor @grumpygamer As a researcher, hell no. You should understand every word and idea you put in a paper. If you don't, you're not doing good science. This is basic shit.
And yes, I'd say the same thing even if they didn't use AI. That doesn't somehow magically turn bullshit into great science.
@chiraag @grumpygamer
Id say good content is good content. If the research paper discovers something new and especially if that leads to new discoveries, thats good content.
@Noisecolor @chiraag @grumpygamer It doesn’t ”discover” anything if it’s not comprehensible.
@ahltorp @chiraag @grumpygamer
You can discover something and not understand it. Thats true about most discoveries ever made.
@Noisecolor @chiraag @grumpygamer I can find a pebble on the beach. Does that mean that I have now discovered physics? No.
@ahltorp @chiraag @grumpygamer
If you describe how its falling or moving, yes you have! And you dont need to even understand special relativity.
@Noisecolor @chiraag @grumpygamer Yes, but if you don't understand what you are describing, you haven't discovered anything.
@ahltorp
Semantics. The content has value.
@chiraag @grumpygamer
@Noisecolor I’m glad you’re starting to describe what words mean.
@ahltorp That was the original point. Can content have value when its written by AI.
@Noisecolor @ahltorp there was a lot of good content in that report that linked autism to vaccines. It wasn’t the conclusion, or any of the summaries. It wasn’t the methodology or critical thinking, but many of the datapoints were correct. The author didn’t u see stand the information so the paper wasn’t good, regardless of the data used inside of it. If it can’t be trusted you can’t say it’s good content.
@passwordsarehard4 @ahltorp
It can be when its verified. No matter how it was generated.
@Noisecolor @ahltorp if you have to recreate a report to verify its contents each time, the AI didn’t generate it you did. The unverified report was useless, it had to be translated by you to turn it into something that could be trusted and used. The AI output is raw ore with 20% impurities, you are judging the results after you smelt it down.
Oct 13, 2026, 9:46 AM@Noisecolor @chiraag @grumpygamer a paper doesn’t discover anything. A paper describes what has been discovered and the process by which it was discovered. What you seem to be saying is that if enough shit is thrown at the wall some of it might stick.
@spzb @chiraag @grumpygamer
No. That is not what im saying at all. Its strange thats you would say that.
Oct 13, 2026, 2:44 PM@Noisecolor @chiraag @grumpygamer literal quote : "If the research paper discovers something new"
Oct 13, 2026, 3:03 PM@spzb @Noisecolor @grumpygamer Papers don't discover anything though...researchers do. And someone who doesn't understand _what_ they discovered and _how_ (as well as ideally having some plausible explanation of _why_ it might make sense) is just a shitty "researcher".
Oct 13, 2026, 12:21 PM@Noisecolor @grumpygamer Science isn't "content" and scientists aren't influencers holy shit. If you don't understand what you are publishing, you are a bad scientist, and that has nothing to do with AI.
Even if *others* don't understand the ramifications, that's fine. History is replete with those. But if *you* don't understand and can't validate what you're publishing, you're a bad scientist and a fraud.
No scientist I know would agree with you (thank God).
@chiraag @grumpygamer
People have been publishing science they dont fully understand since ever. Its kind of what we do. Some stuff we understand, some stuff we just observe, but still use in the future.
Its not a black and white thing and it was just use an extreme example.
Contended point was whether AI can create valuable content. I think it can and discoveries made with ai prove that.
Oct 13, 2026, 2:44 PM@Noisecolor @grumpygamer Scientists not fully understanding _why_ something happens (but having hypotheses) is different from someone not understanding what they are putting in their paper. The two are not comparable at all.
In one case, we understand _that_ something happens (b/c of experiments) and have educated guesses as to why. In the other, the "scientist" can't even explain _what they did_, let alone why it happened. Those are not the same thing.
@Noisecolor @grumpygamer you seriously think that anything from AI can be 'good' in any meaningful sense of the word? Boy, have you been hoodwinked.
@Noisecolor @grumpygamer how can you claim content is good if no one understands it?
@Noisecolor @grumpygamer Unfortunately, that argument doesn't fly on Mastodon. Content is good, until you hear it's written by AI, then you have always found it flawed anyway. No place for objectivism here. 😢
@erwinrossen @Noisecolor @grumpygamer
Ayn Rand can sit and spin, Objectivism doesn't deserve any respect on Mastodon.
@Noisecolor @grumpygamer your truism is a "false premise" fallacy.
Science is about understanding (and explaining).
The only way we know scientific content is actually good is if at least a few scientists validate it by understanding it and eventually successfully replicate or exploit the results.
AI content has not been understood by the meat proxy that shared it, so it lacks the very first step. The humans which prove (or disprove) that, are making the actual science. And why should anyone have their research agenda dictated by random enterprises *for free*? It's already as bad as it is with proper funding.
@grumpygamer @Noisecolor Exactly.
The entire point of a scientific paper is to communicate scientific results to (at least) other people working in the field. If not even the authors of the paper understand it then no one else can be expected to either, and it's a complete failure at the one job that a scientific paper has.
@spacelizard @grumpygamer Not if it leads to other discoveries.
@Noisecolor @grumpygamer If the paper is incomprehensible even to its own authors then it is impossible for it to lead to anything. It's just number-go-up for bibliometrics, and does nothing to contribute to the body of knowledge. Again, the sole purpose of a scientific paper is communication. If it cannot be understood even by experts in the field then it is by definition not good, in fact it is completely pointless.
@spacelizard @grumpygamer
@uastronomer
I think we have made such discoveries since the beginning of time. We had two things we didnt fully understand and we connected them to something we could reliably observe. Arguably we dont fully understand anything yet we make discoveries.
@Noisecolor @grumpygamer You're conflating two very different things. Experimental results that cannot be fully understood with current theories is one thing, using a text generator to generate a document that no one can understand is something entirely different.
@spacelizard @grumpygamer
Im not conflating. You are strawmaning.
@Noisecolor Your words: "There are cases when scientists publish papers that they have written with ai and don't even understand themselves."
Papers that the authors don't understand. Not results. Papers. I'm only arguing with that you actually wrote.
Pointless talking with your further anyway, you're clearly very attached to your imaginary scenario and you're not going to listen to actual scientists telling you that's not how good science happens.
Oct 13, 2026, 7:47 AMExactly. GenAI produces something that looks plausibly like a good paper if you don't understand what you're reading. No true scientist would publish a paper without understanding it, without having tried to disprove its central thesis, or without being prepared to answer questions about it.
Oct 13, 2026, 3:28 PM@grumpygamer @Noisecolor remember Feynman's "cargo cult science" speech?
This is non-science that the authors themselves cannot make meaning of.
Imagine if these people built houses and didn't know why they'd keep standing.
"but people can live in the ones that don't fall over"
@Noisecolor @grumpygamer codswallop. Random joining of the whole of what the tool has been fed on (including misinformation, proven falsehoods, disproved hypotheses, flawed studies and opinions) is most unlikely to come up with anything usable, and the percentage of 'possible' 'good' stuff would be miniscule. Tell me you don't understand what AI does without telling me...
@UkeleleEric @grumpygamer
So unlikely that it had no chance of solving any mathematical problems? 😂
@Noisecolor @grumpygamer it didn't. Read up the truth, not what the AI PR is telling you...
@UkeleleEric @grumpygamer
I am and it did. 100%. There is some controversy the way it did it was combining previous work, which is true, but there is no doubt that it succeeded where humans did not and there are other examples across science.
@Noisecolor @grumpygamer It did no more than a simple program could have done with 1% of the effect on the environment (that would have been possible under a simple COBOL program in the 60s with enough resources, albeit it would have taken longer then). And, as per usual, it is the 1 in 10,000 cases (probably underestimating) that you use as an example... If that is your reasoning for destroying our planet, supporting fascists and paedophiles, and generally wanting to destroy humanity...
@Noisecolor I agree. AI is no magic, it’s a complex tool. It creates things that can be good or bad, like every tool. If a scientist does not check the tools’ output, it is a bad scientist, not a bad tool.
AI created great content and can give the user the power, to create something he wasn’t able before. Many people have excellent ideas but 99% will never be realised because of missing money, knowledge…
AI can partly fix that.
Oct 13, 2026, 10:27 AM@wacholderpolka @Noisecolor Is it a tool though? A powerdrill where you can’t be sure that the diameter is right is not a tool. Even if it tells you „you‘re right, you needed 5mm and I drilled 8mm. That’s on me. I will update my skills.md to make sure I drill what you‘ve told me. Should I order screws for 8mm holes for you?“
@Noisecolor Part 2: As you can see, people do not want to understand your point. AI is bad and there is no way to turn that belief around or have a more detailed opinion on good/bad topics. Intend to not discuss this with hardliners, waste of time. I discuss a lot with people that are open minded. AI is a great tool and a risky thing at the same time. We should use the first and try to avoid the latter.
One thing is for sure: it is here to stay.
@wacholderpolka @Noisecolor This is total nonsense, because it can be said about anything.
LLMs can only ever generate mid-grade content, the absolute average of everything consumed in training. It's designed that way. To statistically choose the next most likely word/phrase in a string. It can never "discover" anything new.
If it's "good" that's because it's regurgitating something good a human wrote, that the LLM consumed.
Why would anyone be excited by this?
Oct 13, 2026, 5:06 AM@grumpygamer Correction: "If you re using AI [...] you should'nt be doing it."
Even if it was of excellent quality, AI is still destroyng life on earth, stop using it, don't use it, period.
@grumpygamer (for the people in the back: this applies to code, too!)
@claudius Yup! I keep tapping this sign.
@grumpygamer they might be (currently) saving money but everyone else is (or will be) paying for it...
@grumpygamer genAI
@grumpygamer Please define 'time' and 'money' compared to earth resources and global warming
Oct 13, 2026, 8:39 AM@grumpygamer I would say the "saving money" part is only now because tech faschos try enforcing it and for now take the bill mostly them self, seen so many slop coders already who get overwhelmed be the token cost and cry like a baby about it.
And the "time" part is also temporary on many aspects as soon you have to go in and fix the underlying mess you don't even understand at this point, if it comes to code specifically.
And yeah for everything art related no brainer it sucks and i still have no idea how anyone can find it appealing.
@grumpygamer also you can't improve as you don't train the "making content muscle". Also AI doesn't improve by doing a task. So it will be shitty forever
@grumpygamer Okay Ron, but seriously what’s the plan when the IPs you created and lost through poor deals will be legally acquired by entities whose sole purpose is to monetize them through automated, AI-generated content pipelines? Just hope that it will be done by people who are as bourgeois, gentrified, culturally aware and well educated as you are? Will that kind of classist cultural gatekeeping suffice?
@grumpygamer ironically, while I read this my wife is watching a "documentary" on YouTube about the Italian town we visited yesterday. Images are all obvious slop - which pains me no end and makes me instantly doubt any of the "facts" presented.
At least this one seems to be original text - the first one was obviously LLM text and voice as well. All the content in YouTube about this historical town seem to be slop... So annoying!
@grumpygamer agreed 😄 but at the same time it's not as black and white(the idea, not the shitty things). I see AI as a tool that can automate some of the boring stuff, like refactoring bits of code. Trying out new concepts. And making placeholder graphics. I wouldn't release anything like that, no matter the color of it, unless it's a free perhaps.