AI conversations are like dreams: everyone has one they like and wants to share it with others ... but no on gives a crap about your dream/chat session, because it was uniquely appealing to you, and not them.
Don't bore your co-workers (or others) with descriptions of your dreams, and don't throw a computer's dreams (AI chat logs) at them either.
I don't know, I don't mind someone telling me about a dream, if it's not too long. It can be interesting or funny.
As always, when one says something, they need to check that the public they have is interested and the public must give cues about how they receive the thing. Of course it doesn't always work well, but this could concern any topic.
Definitely don't share your AI chat with me though, I can sustain a poker face and politeness only for so long, after that I will probably need to complain, vent or practice sarcasm.
I remember reading a book a long time back, titled something like "non-linguistic analysis of call center conversations".
One main takeaway from the book was that "you can just look at the ratio of turn-taking duration, and which speaker/participant is spending how much time" to decide "what happened in the conversation".
The same goes for AI-generated conversations; verbose responses are the default behavior, and models are incentivised to keep that output-token ratio. Too easy to catch/notice, pretty annoying.
PS - I work in the Conversational AI space, and it is quite an effort to keep the ratio right for people to stay long enough on the phone with AI agents.
I legitimately would rather hear about my coworkers dream than their AI chatlog. At least the dream might be an interesting bit of their subconscious. The unwanted log is just a more annoying way to convey what would otherwise be a single sentence answer.
> AI conversations are like dreams: everyone has one they like and wants to share it with others ... but no on gives a crap about your dream/chat session, because it was uniquely appealing to you, and not them.
> Don't bore your co-workers (or others) with descriptions of your dreams, and don't throw a computer's dreams (AI chat logs) at them either.
I was writing a comment after some code I wrote and Copilot autocompleted saying "This is hacky, but..."
I thought it was funny so I put a comment above saying "This is what Copilot said about my code:" and it autocompleted a line after it saying "Copilot was correct, but..."
I've decided that I'm done being pissy about this kind of response, or thinking that it's something that can be coached away. I choose to look at it like any other cultural communication difference - something that you learn about, try to give some grace to, and work a little harder to bridge (unless you're defusing a bomb, performing surgery, flying an airplane etc.).
In this person's communication culture, they are saying "I don't know, but here's my attempt to help."
For me, it really comes down to is whether or not I believe the responder is acting in good faith. If you can't assume good faith, the shape of the response isn't the actual problem.
Of course, my opinion of them is also related to how often their interpreted answer or conversational contribution is "I don't know", and how often they choose to interject with that when it's not necessary. I suppose the latter is cultural too; perhaps I should be clearer in open forums whether I expect them to answer.
> I choose to look at it like any other cultural communication difference
There's a way to tell if this is actually the case: can you find the members of this culture that like getting slopgrenaded? A communication culture needs both speakers and listeners. I see the speakers, I have yet to see the listeners. I could just be missing them though.
In the case that the listeners are greatly outnumbered by the speakers what you have isn't just a culture you exist outside of, you have a common faux pas. One of many, many rude and inconsiderate acts that are nonetheless common, like sticking gum under a table or catcalling. Something about the environment prompts antisocial behavior, and shame is one way to shift the environment and reduce the behavior. There are almost certainly more effective ways to change the environment and curb the behavior, but establishing a consensus on whether the behavior is acceptable or not for those in the know is an important step.
In case it was unclear I'm on the "it's not acceptable" side. Do share if you've observed someone appreciative of being grenaded in the wild, though.
Anecdotally, LLMs are apparently good enough at writing to routinely make the front page of Hacker News. Twitter has a plague of bot comments. From all these up-votes, we can infer there is an audience and this is good faith.
Partly, I admit, I just tend to be wary of shame as a tool. "Someone said something I dislike" has a long history of being abused. Conversely, even something like your gum example, one can point at concrete harms like littering and forcing others to clean up after you.
And the "We should shame them!" crowd has a long standing inability to tell fiction and facts apart, even if previously informed. There was this example of a polish writer who killed one of his characters off- and the people held mournings and church services for her.
If you take the scripture for holy, then the holy scripture talking back to you making you a prophet, will condemn other fiction as blasphemy.
Using LLM as a writing assistant to properly format your ideas, while lazy, is not the same thing as a slop bomb. They're highly rated because the original ideas are good, even if the writing quality is sloppish, and it genuinely contains human input that you can't get simply by asking LLMs a question.
If the AI has enough editing input that your article comes off stylistically as AI-written, it is lazy and your writing is not as good as you think it is.
> can you find the members of this culture that like getting slopgrenaded? A communication culture needs both speakers and listeners. I see the speakers, I have yet to see the listeners. I could just be missing them though.
Head over to Reddit and look for the obvious AI-generated engagement bait. There are usually a bunch of people earnestly responding without even realising they aren’t talking to a real person. Sure, some of those are also bots, but a huge number of them are real people too.
Average people aren’t good at noticing this stuff. They might not deliberately seek it out, but when they are exposed to it they like it enough to voluntarily spend their time on it.
> In the case that the listeners are greatly outnumbered by the speakers
I think it’s quite likely that the people who appreciate receiving slop outnumber the people who don’t.
> In this person's communication culture, they are saying "I don't know, but here's my attempt to help."
You know all those support sites for various products (Microsoft, etc) where people ask questions, and lots of folks eagerly give answers that really aren't helpful?
Throwing AI grenades generally falls into the same category.
I don't mind an AI generated answer as long as it's on point, concise, and answers my problem. If I have to read a wall of text in search of the answer, it's useless.
There's a reason I've blocked those types of "answer" sites from my Kagi search results. Kagi is awesome that way.
To me, acting in good faith means saying something like "I'm not sure, but Claude says this, which sounds right: [short informative clip from Claude's wall of text]". Don't pretend it's your response, make sure it has info you think is useful, and edit it down.
Yeah, you’re spot on, and I hope people catch up on this soon (it’s already taking too long). I think one thing that most people haven’t realized yet is that reading AI slop bombs is really time consuming and stressful. We’re still operating on the assumption that reading is “free”, which came from the fact that writing took time and effort and written words were mostly meaningful before (at least you read something that you know took someone some time to write). But now that “writing” slop is so easy, reading can’t keep up. We can’t just keep reading all the meaningless crap people dump everywhere. It’s not scalable or even doable.
Some years ago, answering "just google it!" was considered rude in some forums. It was asumed that some people didn't have enough googlefu, so googling for them was considered a good help. Also, almost all StackOverflow could be boiled down to a "RTFM" and close the questions.
Today there might be people who can't extract enough juice from LLMs, so it is not entirely useless to say "I was able to extract this info from a LLM, because I am good at it and you seem to struggle", instead of throwing "just ask Claude!".
This is not my experience at all. Getting a LMGTFY or an RTFM response really did mean that your question was easily answered by Google or the "fine" manual. It wasn't rude, it was an education, even a confidence boost saying, "you can figure this out, I believe in you!"
I was once downvoted to oblivion in reddit for answering a question with "If you asked google instead of reddit, you will already have the answer". I remember the question being something like "Who is the CEO of Facebook?". The logic of the downvoting was that some people don't trust Google answers, or don't know how to ask so they can't tell in advance if some question is for Google or needs a human.
IMO they were right, I changed my approach to those kind of questions, and since that I try to answer like "A quick search in Google says that the CEO is Mark Zuckerberg (link to the search)". In StackOverflow I tried to go "As it says in the <a href='manual.html#section'>manual</a>, the params for that function are A, B and C, blah, blah...", so a mild RTFM. And now I do the same quoting the LLM paragraph that gave me the key info. It is like you say "this is how you can figure it out on your own the next time", and feels less aggressive than "go figure that on your own".
In my experience slopbombers are inevitably the worst llm users-- returning a pile of hallucinations because they use some polluted context, don't know how to ask the LLM to double check it (aren't running in an environment where the llm can tool call to check facts or run a test), haven't directed it to write a simple and to the point response, and/or are using some ulraglzing sycophantic chat interface.
Except 99% of the time they are asking it's because they explicitly need a real opinion or the info couldn't be found via LLMs. But instead of giving an "I don't know", they paste back an wall of text with an incorrect answer that the sender hasn't even read or verified to be true.
At least with "I don't know" the asker can move on to someone who might know faster.
It reminds me of how LLM hallucination is attributed to "I don't know" being underrepresented in training data, and it being a better strategy to guess on evaluations rather than admit not knowing.
Different reward function, but the same behaviour emerges.
We'll see that improve as people move onto synthetic training data-- something now possible that we have sufficiently smart LLMs to create enough of it.
The idea is that you generate fake llm transcripts using your classical training data. E.g. look at some training data, generate q/a transcripts. Generate radom questions, RAG against your whole dataset and look for relevant stuff, if there is nothing there, train a "I don't know." reply.
A moderately sized LLM operating some tools to access more information behind the scenes, perform tests and correct its own errors can write transcripts simulating a much larger and smarter llm.
No. "I don't know" may be interpreted as "I don't care". Adding additional info may be a way to say "I do care". Similar to sometimes we say "I don't know, but I googled it, and I got this".
Also, LLM's answers can be good if the prompts are good, so can still be helpful.
There is whole range of requests or questions that I don’t care but I don’t want to reply „I don’t care” especially when someone asks for something easily googlable/llmable.
Turning up to a potluck with no food says "I don't care".
Turning up to a potluck with a bag of garbage you got for free in the dumpster on the way says "I don't care, and I'm so dumb I think you're too dumb to know this is trash".
This is more like bringing fast food burgers. Yeah, it's a bit lazy, but some people do actually appreciate the effort. And post of the people complaining aren't even part of the potluck.
That's how I'm finding I feel, too. It's not that I don't use LLMs myself, but I trust myself enough to know how to filter that data that comes back from them and compare it to actual facts. I don't necessarily know the processing of this information from the other person, to trust that they vet the information properly. So instead of it being helpful to me, it becomes incredibly irritating. Especially because I just know they're going to expect me to take that data at face value quite often. And then it's going to be on me to vet the data. How about you just let me get the data myself and cut out the middleman?
This is exactly my point. To some people, direct communication, especially "no", is extremely rude. To some people, a head bob (easily confused for a "yes" in other cultures) merely means acknowledgement, or "maybe". To some people, extended silence indicates deep consideration or respect.
Globalization resulted in a need to tolerate these differences, and in my experience, trying to "fix" them is considered rude (I suppose that's also a cultural norm!). I just think it's interesting to observe that there is such immediate intolerance of this new behavior. Of course I understand it, and I don't even entirely disagree, I just think it's worth reflecting on, there are probably so many ways of considering it.
What's culture got to do with it? If I wanted LLM responses, I'd have prompted for them myself. This is true in every single culture. The very fact that I asked a person means that I want the information they can provide, not LLM information. There is no culture in which that is false. So there is no culture in which providing LLM content is a useful response.
How you politely communicate that you don't have anything to add? That varies by culture. That LLM content is not a useful response? That does not vary by culture.
If in my culture I can't tell the boss 'I don't know' 'I can't do it' and I am handed a machine and told it's an answer machine, I can understand this conflict coming up. OP specifically said " I don't even entirely disagree, I just think it's worth reflecting on". OP is saying there might be dynamics that can be resolved with understanding why you are getting AI.
I understood the point, and I think it's a flimsy one. Hiding behind "in my culture" is not a sufficient excuse to hide antisocial behavior.
There are cultures that believe the number of chromosomes you are born with determines your autonomy in society. I think most people on here would reject that framing, even if they're met with the same excuse.
When I started managing international teams (long ago) the first months were painful for us all because I didn't understand some culturally didn't feel they could tell me 'no'. I was used to American devs that would happily flat out tell me 'I'm not doing that' or calling me out when I was being too phb style oblivious. Made me change my initial one on ones with new team leads to focus on 'how should we communicate'. I didn't realize how much I was requiring co-workers under me in the org chart to meet me where I was. Huge eye opener.
I realized the same obliviousness on my part made some of my people feel like I was 'good old boys club' because I was more relatable to other white guys into sports (I used American sports analogies up to that point because that was how the management I rose up with talked). I felt awful for making people feeling bad/stressed/in an out group.
If you were acting in good faith, no reason to feel awful. Some people are stressed out in a group or in the spotlight, no matter what. Some people thrive there and love it.
And while different regional cultures may tend one way or the other, one will find direct and indirect personality types everywhere.
Common, if you are in indirect culture, you will HINT that you dont know or that the answer is no and the other person will get it.
These stories are not about people who are from indirect cultures being frustrating to the direct person. They are about people who paste stuff into claude and unnecessary large wall of text - written in direct style.
Or just "I don't know, but researching quickly right now maybe 'xyz'"? Agreed that you need to at least read the response and summarize... Not just copy and paste
I think you’re wrong in one thing: Reading through meaningless walls of text takes time and effort that I’m not willing to waste. So it’s not just accepting a different culture; it’s not “free”. Reading AI grenades is really stressful (at least for me).
>The historical implication: because there is a lot of it, I put a lot of thought and effort into writing it
Really? That must be some ancient history, because I've seen rambling walls of text on the internet derided for decades. I always appreciated the Feynmanian respect for economy of time over traditional formats, where if authors said their piece but still had space they'd damn well fill it.
(Of course, slide all the way down that slippery slope and you'll just hit Twitter.)
I don't think it's a new normal. In my experience a large volume of text rarely has a lot of information to convey. Instead, it's intended to convey the sense that a lot of work has been done.
The consultant mindset set this trend before AI made it all worse.
> In this person's communication culture, they are saying "I don't know, but here's my attempt to help."
It's still a bad attempt at help. Objectively net-zero utility at best.
If it's really just "culture" but they genuinely want to help, then they can in fact be coached. If they're only interested in appearances, well, I agree training isn't going to help.
Sharing an LLM output is simultaneously pretentious and unhelpful. To the OP's point, some might not consider it rude, but I'd rather someone be rude and helpful, rather than pretentious and unhelpful.
tl;dr -- No, it's much, much worse than an lmgtfy link
But, if done properly (e.g. vetting that the top link is actually useful, after basically using the same words in the query that you were being asked), it's a genuine reminder that google exists, can be useful, and should often be consulted before you bother someone else.
It's a teaching moment.
Prompting AI and regurgitating the slop is more akin to googling and cutting and pasting the answer. Instead of making the statement "you can do better next time" it's pretending you are an oracle with supreme super-secret knowledge.
I suppose if it fools the querent, it's all good. But, for example, if I ask you a question because I want your opinion, and I receive AI slop in return, I have just been informed that you don't even value your own opinion.
If I really asked a stupid enough question to merit lmgtfy snark, I'd appreciate the snark. If you're not even confident enough in your own opinion to snark at me, you've just taught me something, but it's probably not a lesson that will do you any long-term good.
But the point OP is making is that it's entirely possible that the person doing this _does_ see it as them being as helpful as possible. That doesn't mean it doesn't suck, or that it isn't annoying, though. I dunno, just seems like a coin toss to me: was this backed by good intentions or not? Without other "evidence", assuming that it was well-meaning but misguided feels better for _both_ of us (at least in my experience).
I had a colleague I really enjoyed talking to. Until all the AI hype and them getting into that bandwagon. Now whenever I ask something I get a huge markdown response with unaligned ascii table or being told to ask <llm name here> instead (with much enthusiasm). I am sure they are not doing it with bad intentions or ignoring me. That said I still can’t help but find it cold though. I would rather prefer if they just ghosted my messages
No, but they obligate you to be kind in your response despite your annoyance. The other person was trying to help, but failed. Keep that in mind if you feel the need to correct them.
This is the sort of thing that only someone on the spectrum would consider doing in a professional setting, and they will end up getting coached by HR.
Saying "I don't know" is 100x less shameful than asking a clanker for an answer and pretending it's yours. If they don't realise this they won't last long
I wish that was the case for people with insincere communication practices. On the contrary, they seem to embody the sociopath stereotype that ultimately climbs ladders.
With lmgtfy, the point was to show that you can do that with Google, how you can do it, and you shouldn’t ask (not in the nicest way for sure). With replying with an LLM answer, you pretend that I cannot do the same. The equivalent would be a link to an LLM chat. There is a clear intent difference. The LLM answer version doesn’t want to teach the how.
>In this person's communication culture, they are saying
Wittgenstein termed precisely this sort of practice a “language game”.
With the rest of his work (Philosophical Investigations in particular) it is a hugely useful lens through which to view and significantly better understand the use and function of language in all its forms.
Given that language is becoming a ubiquitous UI and UX and glue for just about everything, I highly recommend it, and as a work of philosophy it is much more accessible to laypeople not of the field than some other important works in the area.
A language game assumes mutually understood "rules" in the form of shared context that makes language that would otherwise be unclear meaningful. If the intended meaning is not conveyed, we're failing at playing a language game.
In this case, the notion that the slop grenade is intended to communicate "I don't know, but here's my attempt to help" is as far as we know entirely a fiction, possibly a favorable interpretation GP uses not to let themselves be annoyed. In reality, the intent could be any number of things or nothing in particular.
It seems more likely to me that these slop grenades are attempts at appearing to a third party to have weighed in on important decisions. It's a deliberate technique to become more visible in order to win favor with someone who can't tell reasoned responses from useless bullshit.
This self-aggrandizing trash was a normal part of professional culture before the current LLM sewage torrent, but tactlessly used LLMs are so awful at writing that it's never been so transparent before.
If I am explicitly asking for someone's opinion and they respond with a slop grenade that has nothing to do with their opinion, I'm indifferent to the moral question whether they're acting in good faith. The practical effects are the same: my time is wasted, I'm annoyed, I lose respect for them, I am less likely to consider their opinion in the future.
I'm the opposite. Trust can only be given so long as it isn't used to facilitate rampant abuse. Open abuse is rampant now. I'll slam the door in your face if you try to slop on me because that's the only way anyone is ever going to learn at the scale a lesson needs to be learned.
And of course I notice the people posting the massive slop screed don't seem to have thought critically about it at all. After all, if they had time to read it thinking critically, they probably would have had time to write it.
I usually go the other way, sometimes ignoring the wall of text completely or just exiting the conversation. I find it incredibly distasteful when people do this (i also have access to all the same tools), specially because its a lazy response where they copy the conversation, paste it at the tool and just vomit the response back.
Its also usually the dickhead nobody likes that does it as well.
I try not to be pissy about it, but the problem is this from the source article:
> Worse: it's a conversation killer. There's nothing to respond to. Your wall of text suppresses dialogue. They can't reply, can't push back, can't clarify. It's a weapon disguised as helpfulness.
Conversational norms simply do not allow me to give grace or bridge the gap. It would be so, so rude for me to infer that they meant to say "I don't know" if it turns out that it really was a detailed answer they expected me to read in full. I respect that the people who do this probably mean well, but there's some conversations I just can't have with people who respond this way, in the same way that there's some conversations I can't have with people who aren't comfortable exploring contested hypotheticals.
“I don’t know” would be so much more productive and respectful than “here, this may be useful or completely useless, figure it out on your own time. Took me 30 seconds to produce…”
A bit asymmetrical by time, signal to noise, and frankly, bs metrics.
When I'm encountering some WoT like that, I'd like to have a button like "view source", but for "view prompt".
Most ai generated messages or docs are unnecessarily verbose and just reading the prompt would suffice. I don't really get why some people seem to think that it's somehow better to have their bullet point prompt as a huge text.
It just wastes my time. And probably only makes it look like it took more effort than it actually did (it may be the exact opposite).
> I don't really get why some people seem to think that it's somehow better to have their bullet point prompt as a huge text.
Simple: It looks like you did more work.
Before everyone had ChatGPT, a long document meant that someone sat at their computer and invested more effort than someone rattling off a list of partially formed bullet points. In the process of writing the doc they usually refined the idea.
Now anyone can dump the bullet points into ChatGPT and get and expand them into a document which gives the illusion of being well thought out. They can now occupy the same space as everyone who was doing a lot of work in the past, but without having to do the work.
That's only going to work until people start absorb the fact that you can now generate unlimited amounts of grammatical text for free. Shouldn't be long now.
Next, the recipient of the wall-of-text throws it into another LLM to "summarize" it to bullet-points, adding another hop of the game of telephone as it mutates what was already a fake artifact of thinking that was never done. *sigh*
I very much dislike this behavior, personally dont do this and want it stopped.
people that I like to be nice to have done this to me. I dont have a good response that wont trigger them.
They chose to do this cz its easy. I now have to choose : read that garbage and make sense and ask questions or reject it asking for your opinions not claude's.
Former leads to more walls of text. Latter makes me come across as not-nice and get broad stroke painted as an AI hater (which I'm not). Not many where I work will voice any of their discontent about anything AI. Its the hotness.
The irony of that article having an AI smell is not lost on me
> ...probably only makes it look like it took more effort than it actually did
I think you've hit on why people would do this in a work environment. It's a low-effort way of looking like they're engaged at work and know what they're talking about.
> I don't really get why some people seem to think that it's somehow better to have their bullet point prompt as a huge text
Probably people who have never wanted to put the required thinking effort in a simple, structured response to a question, and now think that "a lot of words" magically solves that skill issue.
While Blaise Pascal, he of triangle and foundations of evil, may have indeed written something like that--though probably in 1600s French--it is almost certain that he did have the time. he just so not want to use it to edit his letter to be shorter.
The question before is now is now is whether the letter contains any quotable quotes that have survived 400 years other than his critique of style.
Wheel of Time is still some of the finest fantasy you will ever read. I'm personally tired of the digs at the last few books that Robert Jordan wrote before passing the baton to Brandon Sanderson.
Wide open Throttle. Aka puttin the pedal to the metal, or twisting the throttle to the max on a bike, or pressing the lever as far is it will go on a jetski/quad.
I wish the copy button for a response included the prompt (unless it's code). So many times I'm reading a response that starts running assuming the context of the prompt and I'm left wondering what it's about for the first half.
In that case I still want to see exactly this prompt. Then I know that the person didn't even think about my question thoug I asked _them_ for their opinion and I could have asked ChatGPT myself (and already probably have).
Do you start every response off with "that is a great question"? I don't know any human who does. "that is a great question" is reserved either for really hard questions, or sarcasm. The majority of questions are not great, they are just things the asker needs a simple answer.
I remember listening to this Freakonomics episode [1] on the phrase.
It covers a variety of times people might use it. Sometimes it's genuine, other times it's flattery. Some people use it all the time (and in the episode they talk about calling someone on it). One guest says it's a bridge, to go from the question asked to the question you want to answer instead.
I don't see it in this transcript, but I also thought I remembered hearing/reading that it can be a sign the speaker doesn't know the answer offhand, and needs to consider the question to formulate their answer. I think I'd classify that as a genuine usage: the question is something the speaker hasn't considered before, and thinks is worthy of consideration.
I remember being advised to do this ~20 years ago when I was going to be answering questions from a group of people. I was told that it's good practice to say something like "that's a great question" every time someone asks anything, as a form of social lubrication, to encourage others to ask questions. I can't say whether it works, and it was advice for a spoken context rather than written, but I don't know how to finish this sentence.
When I go to research lectures, I sometimes hear that in response to audience questions, although not especially consistently. Some speakers do this more than others, I don't think anyone does it all the time.
It was so long ago that the specifics have faded, but I remember I was coached to use a variety of positive responses. "That's a great question," yes, but also things like "I'm glad you brought that up," and "I was hoping someone would ask about that!" It wasn't my cup of tea, too artificial, but the advice was contemplated.
The next question (which is a great one, from what I understand) is: Why do LLMs use these phrases so much if humans rarely use them in written form? Maybe a fair portion of training data comes from lecture transcripts, where such responses are common when responding to direct questions? And/or system prompts are just instructed to be like that?
> Why do LLMs use these phrases so much if humans rarely use them in written form?
As far as I understand, it's due to RLHF. The reviewers the AI companies use don't necessarily know what kind of question is a good one, so when the LLM answers "That's a good question!", they tend to rate the answer higher because they like being flattered. Proxy models that are themselves trained on RLHF inherit this pattern. Similar effects contribute to sycophancy.[1]
I've always seen and used "That's a great question!" and similar vacuous phrases when speaking as a polite way to buy some time while formulating an actual response to the questioner.
In a spoken Q&A setting opening every single response with "that's a great question" or "thanks for asking that" or whatever is pretty common as a way to fill a few seconds while you think about your response. This is obviously unnecessary on slack.
It might be a difference in culture, but I definitely remember people saying you should say stuffs like that to encourage others to ask more questions. The rationale is if you are in a more powerful or senior position, like a college professor vs a student, other may be not feel comfortable asking questions because they fear appearing dumb or inappropriate and it hampers communication.
Well before the LLM explosion I would often preface my answers with some form of praise for the question. It depends a lot on audience of course, but it’s amazing how many people tend to perceive direct answers to their questions as negative… and just as amazing how far a little strategic sycophancy goes to temper that. Even though everyone knows it’s half-sincere dead weight.
And in parallel: a non-trivial number of people in class or at work are asking questions for no reason other to establish their own social credibility as smart and or knowledgeable.
Validating their ego and effort and social position can fulfill their social desires regardless of an answer that dismisses, say, buzzword soup as both inappropriate to the current context and incoherent to people who know what those words mean.
I didn't, but then I watched SICP and the guy kept saying it to utterly dumb questions.
And then, somehow, he kept turning the dumb question into a deep and insightful one.
So I've adopted that technique but if I say "that's a great question" think for a while and come up with nothing, it's probably a moronic question. I'm not perfect.
Could add a <vitriol> tag to that - but yes, if that was auto assigned by LLM - i could see that.
Could even add a "Autistism" filter, preventing conversation digressing, filtering out only points that stay on topic and only the <summary>, that way.
There's a difference between a long text and an essay. You wouldn't spend hours typing a message and formatting it with headlines for example. You wouldn't insert loads of unessorasy creative writing techniques in to a message asking for help.
When a person expands bullet points they add extra information from their own knowledge and research in the process of writing. When AI does it, it adds filler and repetition.
Long form writing itself isn't a problem, it's the empty fake long form we have now.
Honestly, speaking as a friend, and as someone who's been at this a very long time, maybe stop doing that?
It doesn't foster conversion and I personally find it kind of a hostile/disrespectful communication style. It's much harder to have a proper back and forth with a firehouse than it is a few sentences at a time.
It declares authority "these are the facts" rather than "let's discuss ideas" and if you haven't fully earned that authority it honestly just kind of smells of insecurity.
If there's something in the middle of a wall of text that invalidates something much further down, trying to communicate the problem becomes a pain in the butt. It's just not a good method for discovery.
Speaking as a random internet stranger, it depends entirely on context.
Sending me a message saying "Hi, I'm getting a Frobnizzle not found error" is a waste of both our time. Explain what you're doing so that I can reproduce it, even if it takes a few paragraphs. Maybe send me your user ID so I can check our logs. I don't care if you're declaring "these are the facts" because the facts are what I need to help you.
If it's a massive wall of text with a defensive tone during a discussion, yeah, sure, that's bad. Do you work somewhere where that's common?
Some people, like me, have developed this communication style because it turned out that when they didn't they were very often misunderstood. When properly applied (i.e., not excessively, no actual walls of text), giving appropriate context helps focus the thinking of the receiver in the right direction.
> It's much harder to have a proper back and forth with a firehouse than it is a few sentences at a time.
Sometimes just providing the context about an issue is itself enough to warrant a few paragraphs, let alone a few sentences.
Obviously, it's best to express a problem, its options, and its recommended solution(s) in as few words as possible, but it's unwarranted to hold an absolute position that every discussion must be an inefficient exchange of mere sentences at a time.
I'd rather have an outline-style essay given to me in one go for me to digest and research async rather than be subjected to a barrage of back-and-forth pings. That's the real disrespect of other people's time.
> It declares authority "these are the facts" rather than "let's discuss ideas" and if you haven't fully earned that authority it honestly just kind of smells of insecurity.
Not at all.
1. Someone is coming to me with the question. They're doing so because either the question is about my area of ownership and I have that authority or because I'm a subject-matter expert and I have that authority.
2. I don't know what the other person knows around the edges of the domain that the question is in, I don't know if they understand the constraints of the domain nor do I know what constraints their specific problem has.
3. Often the answers to any actually decent question at work are fairly nuanced, and to understand the nuance you need at least a level set of context.
It's a lot more dismissive and rude to answer with excessive brevity if you treat the question in good faith. For simple questions, sure I don't need to write an essay. Some questions I answer with "Got 10 minutes so we can chat about this?" because it just needs a conversation. Or I answer with questions of my own. But if the question is well-formed, answering it starts with providing the necessary context to understand my answer so we're on an even playing field and we can effectively communicate ideas.
How long have you been at it? Because some of us grew up writing letters with pen and paper, sending them to people in the mail, and getting something back a week or two later. You just have to actually sit down and READ closely what people are saying, sometimes multiple times, to make sure you are clearly understanding what they’re saying rather than skimming everything you encounter for information to extract.
It is actually quite easy to communicate a problem in the middle of a wall of text. You simply refer to the phrase and then explain why it doesn’t hold. It is also fine to simply present your perspective to people without invitations to “discuss ideas.” You can open a discussion if you want, but if I’m telling you something then you can rest assured that those are the things I believe to be true, and if I am uncertain about any conclusions I will include caveats to indicate uncertainty. You have free will and are perfectly capable of taking or leaving anything being said to you.
I think what is interesting is that we keep needing these pages to teach people how not being an asshole works. I don't really understand why it is so hard to understand not to do (what I consider to be) impolite stupid shit.
And the effect is always near zero because the people who need to read and learn from those pages the most are those who are the least likely to do so.
> I don't really understand why it is so hard to understand not to do (what I consider to be) impolite stupid shit.
Tip 1: Starting off by denigrating folks who do this is not a way to make people change. More often than not, it amplifies the behavior you don't like.
Tip 2: It helps to indicate you understand the perspective of why people do it before asking others to change. Reading your comment, I really doubt you understand their perspective.
People listen to those willing to understand them.
Lots of stupid folks out there. All the technology in the world can't make them smart. Even if you strapped meta glasses onto them and they read the AI's output verbatim, the morons would probably stumble over the words. We'd get a society of stutterers.
Outside of social media, I never saw a slop grenade except in places where it already existed without AI, like SAV responses or other scripted marketing/HR stuff.
Even a real person calling me on my phone to talk 5min about its company without allowing me to interrupt feels like a kind of grenade. Obviously I could interrupt the impolite way but that's beside the point.
I have a coworker whose first language isn't English. She uses AI to polish up her writing, particularly long documents. She puts a ton of effort into making sure that it still reads well. Because of this effort her writing is strong and precise. Before AI she made all the obvious mistakes you'd expect from someone who's not a native English speaker. It's very hard to tell that she used AI because she puts so much effort into post-AI copy editing, it's just clear and useful writing. Sure, the occasional non-idiomatic phrase creeps in but those are hard to find.
That's AI writing done right, and it's very different from this other guy I work with who does the whole slop grenade thing.
Then a better recommendation should be to use specialized AI proofreading tools, such as Kagi Translate's proofread feature.
Yeah, it uses AI, but the "harness" around it forces you to use it only to improve your text, not sloppify it.
You do realize that when you have to find a special use case to defend something you are really giving an argument AGAINST casual widespread use of it.
If you can’t explain it to a human; you can’t explain it to a bot. I’d rather have people just send messages in their native language than correct their broken English with LLMs or send machine translations. At least that way I’d have a record of exactly what they intended to say.
Science says the opposite, sorry. People lose their language skills when they outsource their thinking to AI. You can believ what you want want but that doesn't change the facts.
Pretty sure the references you are referring to are about losing coding skills, which is a very different set of skills than language skills. The kagi link (1) a sibling comment left was an example of an AI that can improve writing while also informing about better writing style. As opposed to the slop grenade, which is outsourcing thinking to AI.
The point is, you don't have to outsource your thinking to use AI, if it's a good AI tool. But most AI companies are coming from a hyper-scaling mindset where addicting the user to the product is the same as substituting hard thinking for easy dopamine hits. The most ridiculous benchmark I have seen in AI is the tendency to say the longer an agent can work and some minimal accuracy is a good thing. That just means you have 30 steps to find an error in instead of 3, and you are much more likely to just abdicate thought instead of the hard work of proofing 30 steps yourself. AI companies and evaluators are losing the point.
I have had experiences where customers use AI to communicate and express their issues. Sometimes they produce walls of text like the website exemplifies, but overall it's a better alternative to not be able to explain the issue because you don't know the specific terminology and you are just a layman trying to do things.
Show some love for the layman, we are all laymen in areas we don't know about.
The problem with this logic, no matter the context where it’s deployed, is that you can always default to “you’re doing it wrong” no matter what case or situation is brought up. It’s an argument that is unfalsifiable no matter what because you can simply gesture to the person as the problem in literally any scenario.
If I build a car and it consistently gets into wrecks at a rate 500x that of other cars, you can’t just keep saying “operator error.” At some point you have to ask, ”why do operators keep having errors?”
The common denominator seems to be your phrasing, not everybody else’s “sarcasm detectors.” And unfortunately a lot of people write comments like your joke but actually mean it.
If we’re going to go down that route, then we should also acknowledge that it’s a nerd forum that is typically reserved for serious discussion. Being sarcastic and cracking wise is generally discouraged
People make jokes here plenty. And being sarcastic? Only all the time. You can post 300-word matter-of-fact, spelling out all the premises comments. Or twenty-word quips that say the same thing. Both will be accepted depending on the mood.
But stubbornly trying to win arguments (imagined or real)? Oy vey. That’s a reservation that takes up a private jumbo jet.
Edit: And I saw your pot/kettle reply that you deleted. Of course you couldn’t help yourself. Mate.
I removed it - apparently not fast enough - because it felt mean-spirited and I was trying to let this go but apparently you are unable to do the same.
Also, making fun of the way somebody talks is incredibly rude and shitty. I don’t know why you feel the need to mock me like that in an attempt to make me feel self conscious about my speech patterns. That’s not cool and it is incredibly low, to say the least.
Either way I think it best if we both just moved on.
I'm verbose, and no one was ever interested in anything I ever had to say before. And, they certainly wouldn't be now if I used AI to do it.
These days I only bother commenting if I really feel I need to get something off my chest, but I don't assume anyone gives a shit. I frequently just delete my responses. I don't generally feel anyone needs what I'm saying.
It takes a bit of self-awareness to realize that no one is interested in what you have to say, so don't bother. And, I think it takes twice as much self-awareness to realize that they certainly don't care about what a bot has to say, unless they're the ones asking.
I have a friend/coworker that is verbose af (14min audios discussing something that could be compressed in 100 words) and is not that we’re not interested, but it gets annoying after years and years of text walls.
Being concise is an art by itself and it’s a really good way to express your thinking and convince people, it just needs practice.
I love terse text communications most of the time (Slack and email at least). So much clearer. And easier to respond to.
I think we've all worked with someone who (I imagine subconsciously) feels the need to make things longer without actually adding more information in there, and it just makes everyone's day a little harder.
I developed this habit when I worked in software support. It made me substantially more effective. Sometimes people are reading for the joy of the prose. Sometimes they’re reading because they want to get to the answer. In the latter case, writing that is easily visually parsable is quite valuable.
In instances where context is important, I have been including a summary with call to action at the start of the message, then include details below to hopefully eliminate back and forth. It helps me be more clear with my point, and most people once they have an action only use the context for reference later.
So if somebody wrote the same lengthy answer without AI you would not interfere?
Shouldn't we criticize the issues of the text instead of it's creation process?
I view this use of AI as a type of denial of service attack. Alice sends Bob a wall of text (low effort) and Bob must parse it (high effort, yet probably no value received for the effort).
It does, because it allows for quickly sharing a prepared response instead of saying the same thing over and over. It also works because the kind of person this link gets sent to is already used to trusting random websites over their human interactions.
And if we don't (like me), I think it can be assumed that we can find out.
A basic search tells me they're both ways to speed up applications/projects through memory management and storing of the memory/data in RAM. So the person in the article is asking a co-worker their opinion on which tool they should use for optimization/performance in their project, which is why an LLM response is less helpful than his co-worker's actual thoughts: comparisons of technical tools at a high or feature level are pretty easy to find and the decision of which to use often boils down to project specific variables, so what the author is asking of the co-worker relies on the co-worker's knowledge of their specific project.
I could also probably explain this just fine to non-tech people. It wouldn't be the most complete or technically accurate explanation, but it would be fine.
Most people write based on their experience and culture.
I mean, they're not exactly obscure. And if you're curious you can just read the wikipedia pages (or ask a magic robot, I suppose, if you _must_). What universal example that everyone would know were you thinking of?
Just prompt them back: "that's a lot of detail, could you please summarise as briefly as possible what differences concern our requirements specifically?"
This is unfair to those of us who’ve been writing walls of text to simple questions all our lives. I demand AI be the ones who start writing shorter things. Then I can go back to people ignoring my accidental wall building— accident because lots of times you’re a page in still going on things you see as relevant and— well, you see how things can get away from a person. But I want people to ignore me for what I write, not for thinking i don’t have anything worth ignoring in the first place.
“Imagine that you encounter a signal. It is structured, and dense with information. It meets all the criteria of an intelligent transmission. Evolution and experience offer a variety of paths to follow, branch-points in the flowcharts that handle such input. Sometimes these signals come from conspecifics who have useful information to share, whose lives you'll defend according to the rules of kin selection. Sometimes they come from competitors or predators or other inimical entities that must be avoided or destroyed; in those cases, the information may prove of significant tactical value. Some signals may even arise from entities which, while not kin, can still serve as allies or symbionts in mutually beneficial pursuits. You can derive appropriate responses for any of these eventualities, and many others.
You decode the signals, and stumble:
...
They contain no usable intelligence, yet they are structured intelligently; there is no chance they could have arisen by chance.
The only explanation is that something has coded nonsense in a way that poses as a useful message; only after wasting time and effort does the deception becomes apparent. The signal functions to consume the resources of a recipient for zero payoff and reduced fitness. The signal is a virus.
Viruses do not arise from kin, symbionts, or other allies.
I was getting a lot of overly verbose annoying emails at my one workplace and I swear they probably had started to use AI to generate a lot of it. Outlook had added AI summarization buttons built right into the email client and I must admit that I did make use of it for these kind of emails.
Sadly the answer to save time in these cases sometimes is to use AI. Sometimes the emails being sent weren't even AI generated. Just emails from managers/directors who don't respect peoples time enough and think we care about their long essays of nothing. You kinda have to check the emails to make sure this isn't the one time something important was being said, so that's where the quick AI summarization comes in.
Pretty sure there's an implicit dynamic where the more someone uses AI, the more you require AI to understand and work with what it produced. If everyone around you is using AI, you are pressured to use AI to keep up with their level of "productivity". Like a cultural virus it multiplies in the space between people, I guess meme-like but far more virulent. Sure it empowers us, but at what cost.
A lot of organizations right now are dealing with the "Claude says..." problem.
It's not that you should never include any AI generated text in a conversation with other engineers, but you mostly shouldn't.
AI is a tool to help you understand/debug/review yourself, but you shouldn't be conferring authority to your AI (because it doesn't understand anything and is often wrong in the text it spews), and you shouldn't be running an AI for other people, they should run it themselves.
I think this touches on the core difference between good and bad use of AI; using AI as part of the process vs cutting and pasting LLM output.
Use AI as part of the research process, to help understand a concept or problem. Use it to format data, or as a part of the design or brainstorming process. Use it to build manageable portions of code that you can read and understand before committing. But if the output doesn't go through your brain somehow before you unleash it on the world, that's really no different from a seventh-grader Googling the subject of his homework and then cutting and pasting the entire text of the first result, headers and all, and turning it in.
Replace "Them" with "Coworker" and the point of linking to the site is instantly understood (a LMGTFY-style shaming with a dash of humor to soften the blow)
With "Them" I wasn't sure if you meant the AI companies, some dude I didn't recognize in the avatar, scammers, coworkers, etc...
If not…well I don't know what to call this style of writing exactly but I see it all the time on LinkedIn (or some annoying startup landing page) and it's very upsetting to me.
"You asked a simple question. They lobbed a document."
"Your wall of text suppresses dialogue. They can't reply, can't push back, can't clarify. It's a weapon disguised as helpfulness."
"Let it sharpen your thinking, not replace it."
Why structure sentences like this? Maybe go read some well-written novels, a solid essay or two…or at the very least, write in a normal conversational style?
Do people actually do this in things like slack? (One of the best things about being a professor in a non lab field is that I don't have to use things like slack.) This seems like open contempt for the reader.
I've never seen it, but I have a buddy who had a coworker like this. Would basically treat his slack as a manual copy-paste bridge to an LLM and it's was incredibly unhelpful because most questions were heavily context dependent.
I imagine this is the kind of thing you see at a large company where a good chunk of people are just coasting by doing nothing, Nelson Big Head style.
Yes all the time. Healthy conversations are shut down immediately with the "AI research" slop-wall.
Meetings are 55 minutes of speculating about what AI will be able to do in a few years. Then if you are lucky the last 5 minutes can be used to discuss a real issue.
Reminds me about similar "manifestos" about netiquette, properly asking questions, searching web and answering emails. And I expect exactly the same impact - none.
Generates entire websites with AI Slop. Instead of sending a single text mail with three links and the words please make that certificate.
No. He wastes the time of all personnel. Wastes energy. And hides the important message in a wall of text (I was the only person which recognized, that he requires the certificate…it was hidden in a side box).
Right now we re-implementing every frogging tool which was ever developed by more experienced people.
Excuse the long letter, I hadn’t the time to write a short one.
The best are the Jira tickets with a huge wall of AI slop requirements. Usually full of nonsense of course including implementation recommendations in the wrong language or framework. Questions for clarification met with blank stares from the author. Ah well, copy/paste into claude code and say “do this. make no mistakes” and get back to browsing HN…
That "weapon" quip from TFA really pisses me off. It implies actual malice on the part of the person copypasting GPT. But people aren't being malicious, they're mostly just clueless. "Weapon", "weaponized" etc sound punchy so they're a safe choice in the AI's slopertoire.
And on the subject of weapons, consider the example of a delayed-action fuze. They're designed to penetrate through the outer defenses of a target, before the warhead detonates deep inside. What might this look like in written form?
Well, you'd start by getting on the reader's side, with shared loathing against some outgroup. Get them to let their guard down. And then at the end:
>Use AI to make things clearer, not longer. Let it sharpen your thinking, not replace it.
You, yes you, the very clever and sophisticated person reading this who is exasperated by slop-grenades, should use AI more. You should use AI to "sharpen your thinking". You use AI, but in a good way. Not like those rubes, who use AI in a bad way.
Huh, just strikes me how we're all on the Internet seemingly inhabiting the same world, but really this never happens to me and I'm in a pretty AI-forward world.
The other day I found the worst podcast I think I've ever tried to listen to. AgentStack Daily, which apparently sums up AI stories (mostly focused on OpenClaw and the like), using computerized voices.
I don't even have an issue with it being AI-generated. However, the content is delivered so fast and monotone that it's impossible to listen to, and every episode is 40 minutes or more, every day.
A brief daily summary, perhaps using the creator's real voice (via ElevenLabs or similar; the creator has a real podcast on the same site), would be so much more valuable.
AI-generated podcasts are terrible. I came across one that was supposed to be summarizing FAA training material and it was unearthly and fast-paced, with pointless "humanization" umms and aahs, and in a weird conversation style. It's doubly bad because it's important to get things right, and if you just yolo the AI summary, you don't really know if it's getting everything right.
If the poster cannot do the work of writing, they should not expect others to do the work of reading. People read based on trust and respond to invest in relationships. LLM verbiage breaks both dynamics.
I prefer the brevity and branding of slopgrenade, to be honest sloppypasta itself has a slop energy to me. Maybe it's the UX and the density of the website though, it does have a similar messaging.
The point of OP could be right but I decide to provide a counter example.
This morning I asked Claude to find why locale = empty string slipped into some of our customers records. Of course it's something I used to do myself but it did it in less than a minute. I couldn't match that speed, ever. Then I verified the analysis and the suggested fix. Finally I pasted Claude's analysis in a slack conversation, with attribution. I could have summarized what claude wrote to me, but it would be a waste of time because it was already pretty terse.
We decided to solve the problem in a different way but I think that starting with Claude's analysis was the right thing to do.
That's a best case scenario. My experience is that it's more common for an LLM to confidently predict incorrect solutions to the problem, and then you wind up wasting time chasing down wild geese. At those times, it winds up being more of a hindrance than a help.
Do people actually do this? I’ve been on a sabbatical when the whole AI thing happened and I might start working again soon and I’m not sure what to expect in the Slack mines.
Had AI help me write a blog post last week. Most of the process was deleting verbosity. I guess it solves the blank page problem but once you get going the noise is worse than doing it yourself.
At work they recently added AI note taking to meetings. In interviews we have a panel of interviewers and everybody publishes feedback. I continue to take candidate notes by hand in the interview and write my own feedback. During the review process we read each other’s notes. Some people just take the summary of the conversation and use AI to convert that in to feedback, so now my coworkers are sharing sometimes paragraph after paragraph of slop when what I want is their own thoughts on the candidate! Very annoying.
I've noticed this happening here as well. The instance I realize it's not another human I lose all interest in argueing or conversing. If this happens too often I leave those sites.
Because nothing feels more like wasting my time than talking to an answering machine that is working against me. It's exhausting and demotivating.
I love asking someone who sent me a Slack wall of AI text to join a huddle, then ask them deep questions about said wall of text while they struggle because they have no idea what they’re talking about. It seems to encourage folks to be a little more careful about their wall of texts in the future.
Most of my friends are starting to talk like models (myself included), it's actually concerning to an extent, because we spend most of the time interacting with AI instead of humans, we are starting to mimick their behavior and speech.
Either you have to give the AI the points you want to convey, then just put those points in a message. Or you don't have anything to convey, then don't post a message.
I don't see why anyone would want a slopified version of whatever it was that I had to say.
> I don't see why anyone would want a slopified version of whatever it was that I had to say.
Lots of people lack confidence around their writing, and many people (particularly in tech) are not english native speakers. I can definitely see both of those groups getting use out of AI assistance in writing.
That being said, I sometimes use AI to see if I've missed anything, but the last thing I'll give up to our future AI overloads is writing text, as I enjoy it.
Not to single out OP or anything, but the more we do things on our own, the more likely we are to build our confidence. Relying on something or someone to hold our hand risks slowing down personal growth.
I agree, but there's lots of stuff I'm pretty bad at (javascript, for instance) that AI is super helpful for. I feel like I'm learning a bunch of new stuff quicker than I otherwise would've.
So I probably wouldn't argue against this in all cases, except where someone is just outsourcing all their thought to the model(s), that feels much worse to me.
Lots of people lack confidence around their writing, and many people (particularly in tech) are not english native speakers.
I am neither (a native speaker), but you learn and gain confidence by doing. Also, I think flattening the beauty of language (however imperfect) into LLM-speak is a huge loss. Everyone sounds the same and boring through an LLM.
I said in my comment: shortening, cleaning things up.
You’re foaming at the mouth a bit about AI here rather than reading what I said. I don’t see how shortened or fixed grammar is slop. The article argues the opposite.
This is very reminiscence of the whole LMGTFY (let me google that for you) phase of things. At a job in a while back, when front-level support reached out to senior staff for help the two golden rules were:
1. Do NOT answer right away. If they wait, there is a good chance the next message is "Oh wait, I figured it out" (e.g. they googled it finally)
2. Send them a google link w/ the search term showing the first result.
Granted, this was a bit tongue-in-cheek and we did a LOT of trainings to help facilitate actual learning. Still, it was far too easy for senior staff time to get burned up by folks making minimal effort to think for themselves so friction remained.
While the site makes a good point, they miss the most important point, IMO, which is inferable by the example of a good response. The good response is better principally because it contains business-contextual information, which AI can never provide without proper prompting (and if you know to provide that, you prob don't need the AI answer):
"We need pub/sub for the notifications feature."
I'm not anti-AI, but good answers include historical business context to explain decision making. Sometimes if you're lucky, code comments contain this in relevant sections :).
I’m still holding on to my hope that LLMs will destroy verbosity and usher in an era of concise and objective human-written works. Especially in academia. In my research area people write 10-page articles that could easily have been 2 or 3, for no reason other than that 10 page is the publication limit most of the time and if you write less it’s seen as lazy or something. Not to mention 20-page journal articles with humongous literature reviews that everyone skips, and so many meaningless diagrams and sections. I still hope that LLMs will make people so tired of reading slop that they’ll beg you to please submit shorter articles.
> Pasting a massive AI-generated response into a chat or email where a human would write one sentence. It destroys the medium itself. Nobody writes essays in Slack. It's only possible because of AI copy-paste.
> It's like calling someone and asking "What time is the meeting?" and they read you a 10-page analysis of calendar management best practices. You asked a simple question. They lobbed a document.
It’s hard to take the site seriously if the author themself isn’t able to write
It's certainly concise but I still remain unconvinced a human wrote it.
> It's a weapon disguised as helpfulness.
The source code is without a doubt AI (it's got a comment for the "<!-- Canonical URL -->"), so I guess one would have to assume they prepared the entire document beforehand, then fed it to Claude and instructed it to use that copy exactly.
...or they prompted "make me a site which tersely criticizes people who post AI slop on Slack, use the term slop grenade and style the site like nohello.net"
We should instaban anyone who uses these patterns on ycombinator. If you talk like that for realsies, well too bad, try being more human and less linkedin next time
They haven't only been trained on the common author. They have also been trained to prefer quality and to dismiss hallucinations. So if a human actually does some quality writing, people might think it's AI, like has already happened.
>Worse: it's a conversation killer. There's nothing to respond to. Your wall of text suppresses dialogue. They can't reply, can't push back, can't clarify. It's a weapon disguised as helpfulness.
This is slop. What it's saying is not even true, it's just punchy.
Right, recent youtube solo lecture vlogs ripe with these clever punchy finishers. Annoying. I know people aren’t smart enough to actually be that witty ad hoc. Makes the whole thing feel fake af
Annoys me when somebody thinks helping is sharing an LLM dump
I get that the solution is in there but it's like you didn't really help me if I have to go through and read something myself. Just tell me your conclusion if I'm asking you the human. Context is a 7 page word doc output or pdf.
I agree that between humans the example given is laziness and disrespect.
That said, I want to make it clear that if people are going to regurgitate LLM results either way, I'd rather get the longer slop than trust a concise "Use Redis" conclusion from a system that doesn't think the way we wish/assume it does.
Ultimately we're using a statistical language algorithm to predict what kinds of words usually come next in a short story we've constructed.
* If you train it for short outputs (or stories where a fictional computer character has short dialogue) you're prioritizing text from places where someone answered without explaining.
* If you run its output through a hidden "summarize yourself" path, you're adding additional potential for error and dropping details you could have used to detect it.
I have begun using the acronym TL;DP (Too long didn't prompt) For when someone sends a wall of text and I didn't want to waste tokens having an agent summarize it for me when the sender could have done that for me with their own agent.
Nowhere since gpt wrote it. Normal humans dont talk like that. Normal actual humans with veins, arteries and a fully functioning endocrine system talk like ME
> Use AI to make things clearer, not longer. Let it sharpen your thinking, not replace it.
If someone sends me an AI generated email, chat message, or message substantially influenced by AI[1], one of two not mutually exclusive things will happen:
1. I ask them not to use AI as I want to hear from a human colleague about their human thoughts, not a robot;
2. The message gets deleted.
I try as best I can to teach and mentor others. I am more than happy to work through spelling mistakes, poor grammar, and misused words because at the end of the day I'm talking to a human colleague.
Sometimes my messages get pretty long and detailed I will admit, though it's for a reason: context, nuance and technical details are important. If you're just going to offload your brain to a robot, I'm not going to waste my time feeding that robot with you in the middle as a conduit.
[1] It is very easy to tell in in-person conversations: the authority with which a person talks about a particular topic via text communication, does not propagate into a verbal in-person conversation.
I think the bigger issue is the motivations for posting AI slop.
To me, a lot of these responses aren't made in good faith; instead, they come from bots that are some kind of training experiment. (Like when a bot responds to one of my HN posts.)
Even if the response isn't a bot, if it's just someone copying and pasting AI, how can someone reasonably think that just shuffling a comment into AI is adding any value?
Not everything can be answered with yes or no. I used to give thorough answers even pre-LLMs. When someone asks about a project and its details, I am supposed to give a one sentence answer because LLMs give lengthy answers and that sucks?
This is like people hating on em-dash because LLMs use them a lot.
I find that the people who are the worst at their jobs, write the largest blocks of absolutely useless texts. In all disciplines. So yes, I see humans writing 2 A4 docs in slack; they have no clue what the question was about and just insert drivel.
Speak for yourself. I do write essays in Slack. Just because you, the author, are too dumb or too lazy to put some effort into written communication, doesn't mean that we can't do it either.
A while ago I worked with a guy that did not put effort into written communication and he was doing the marketing for the company. I would have words with the boss about this and, initially, I struggled to explain why spelling, grammar, punctuation and use of paragraphs was important.
After some thought I was better able to explain the not-so-obvious. It is all about respecting the time of the reader.
Some people don't respect the reader since they think they are important. Notably with the Ep*tein files, we have this clique of wealthy people, with few of them able to string two sentences together. Writing standards were those of an eight year old at best.
Personally I write fairly long form because I don't have ideas that I can express in glorified grunts. However, this LLM stuff is encroaching on my turf, since the likes of the guy I used to struggle with can now churn out better English than I can. That is the problem for those of us from the pre-internet, pre-grammar-checking age when written communication mattered.
I actually don't care about the length so much. Short AI slop answers are also offensive, and people who act like reading a paragraph or two is a huge burden are frankly get no respect from me either. Idk why professionals should tolerate subliteracy among their colleagues.
>If they wanted an AI essay, they would have asked ChatGPT themselves.
This is not true in the least bit. The page even included an example of calling someone to ask when a meeting was instead of asking an AI assistant to check their calendar. There is a reason why so much of company support can be done using AI or via people following a flowchart. People do not know how to solve problems by themselves.
No no, let's just stop thinking entirely and paste conversations from LLMs back and forth to each other. Then we'll use an LLM to summarize the conversation to tell us what was said. Then we'll use an LLM to do what was said. Then we can ask an LLM if what was done works.
“Worse: it's a conversation killer. There's nothing to respond to. Your wall of text suppresses dialogue. They can't reply, can't push back, can't clarify. It's a weapon disguised as helpfulness.”
I can reply.
I can push back.
I can clarify.
I am not helpless.
When real people use AI slop to spam me down, I instantly know that this person does not want to communicate with me. So I stop all communication with that person.
What is interesting is that some people don't understand this - even some clever devs.
For instance, on the ffmpeg mailing list a few weeks ago, one of the lead devs from Germany, spammed a proposal with AI slop. Someone else asked the question why he expects others to read the slop and "engage" with this or that developer. That was a great question. The interesting thing is that the original developer who succumbed to slop, did not even understand why AI slop spam is problematic to other people. AI already changes how people work and also think. That is a big problem. I used to semi-jokingly say that AI slop is the beginning of skynet, but as I watch real people succumb to the AI slop, they actively (!) become dumber and don't understand why AI slop wastes the time of other people.
I am not at all saying that AI is completely useless, though the current hype is annoying to no ends. But some individual humans don't understand the problem at all anymore. Personally I do not want to "interact" with AI slop at all. It just wastes my time.
What did he say? I'm very curious about his perspective. Presumably he wouldn't knowingly be harming his own project. So he must think it's actually good.
The sad part of our reality is now any well written text is assumed an AI. The other day I wrote a funny long text in some group chat, took me some time to do and proofread, only to be ask if it was AI generated! I like details and I am sure a lot do, now people will be inclined to write shorter “dumber” text to counter signal the AI slop, which in the long run might actually make people dumber.
The only way to defeat a grenade is to toss it right back where it came from. Slop replies get 2x the slop in response. Most effective way I've seen to get people to stop doing it.
Easy solve: "ChatGPT, generate an explanation that's short and concise and won't be interpreted as a slop grenade. Also write it in a way where nobody can tell it's written by AI because everyone thinks they can spot AI, but really if you tell the AI to write it in a way where no one can spot it... it can."
The sheer audacity of using generated slop like this is something that always amazes me in a bad way. You can always tell.
Every time someone uses answer like this it shows that he doesn't even want to discuss something with you and possibly knows nothing about the question asked. So the answer it self could potentionally be bogus or straightforward lie. It's just rude. It's even more rude that when someone tells you to google answer instead.
Don't bore your co-workers (or others) with descriptions of your dreams, and don't throw a computer's dreams (AI chat logs) at them either.
reply