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.
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.
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.
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.
> 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.
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).
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).
> ...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.
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.
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).
As much as I prefer shorter output, I think the example given displays too much trust that LLMs are capable of solid logical thinking.
Whether you ask it explicitly or not, what's really happening is you're using a statistical language model to predict what kinds of words usually come after the words you've provided to predict what kinds of words usually come after the words you've provided.
If you train it for a short outputs, you're prioritizing the documents where someone answered and never explained.
If you run its output through a hidden "summarize yourself" path, You're adding additional potential for error.
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 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.
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.
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.
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.
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 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.
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.
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 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.
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.
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.
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.
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.
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.
Just prompt them back: "that's a lot of detail, could you please summarise as briefly as possible what differences concern our requirements specifically?"
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?
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...
Reminds me about similar "manifestos" about netiquette, properly asking questions, searching web and answering emails. And I expect exactly the same impact - none.
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.
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.
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…
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.
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.
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 :).
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.
> 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.
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.
“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.
> 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
>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
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.
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.
>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.
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.
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.
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.
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.
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
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 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.
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.
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.
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.
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.
I'd prefer just "I don't know." I can ask Claude on my own.
> To me
> I'd prefer
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.
> 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.
Copy/pasting a question to LLM and pasting back the output isn't an attempt ar being helpful. It's the equivalent of a lmgtfy link.
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).
Good intentions don’t excuse bad behavior, and shouldn’t force one to tolerate it.
Having good intentions doesn't give a free pass to be obnoxious.
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).
> ...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.
I would have written a shorter letter, but I didn't have the time...
We intuitively think large documents show significant thought.
I don't just mean the readers.
The generators of slop often think this is useful.
Things have changed.
Our intuition has not.
"WoT"
hmm.. Wheel of Time? never got into those books personally
Wall of Text.
Both being infamously long
It is too bad that neither Unhinged or Unglued had a proper Wall of Text card.
Thanks!
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.
Just, pull your braid and smooth your skirt for a few times, and you'll get into the spirit of them.
I just want a ”report to HR” button. Someone is actively inhibiting their coworkers’ ability to work.
HR isn’t there to protect or help or serve employees.
Except often the prompt is just the previous comment. In the example, the prompt would be "Should we use Redis or Memcached?"
In that case, there is nothing beneficial about the prompt, but the answer could be boiled down to a useful recommendation (from an AI, not a person).
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).
I mean, you can just point your LLM at the wall of text and ask it to dream up with the prompt that made that. Or ask it to summarize into a TLDR.
As much as I prefer shorter output, I think the example given displays too much trust that LLMs are capable of solid logical thinking.
Whether you ask it explicitly or not, what's really happening is you're using a statistical language model to predict what kinds of words usually come after the words you've provided to predict what kinds of words usually come after the words you've provided.
If you train it for a short outputs, you're prioritizing the documents where someone answered and never explained.
If you run its output through a hidden "summarize yourself" path, You're adding additional potential for error.
> Nobody writes essays in Slack
I 100% write long texts in Slack. I always try to provide as much context as possible when reaching out to someone with a question or request.
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.
That's a great question! Ahem.
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.
Math professor here.
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.
I would probably reply with "that is a great question" only as a euphemism for "I don't know"
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.
I have only heard this phrase in american tv shows and movies
People use that too when speaking in real life when they are stalling a bit to conclude their internal thoughts before providing an answer.
work at a crown corp and you'd change your stance on above...
<context> <tutorial> <anecdata> <answer> <sumary> <funny hook>
Introducing AI made markdown tags for conversations so others can only see what the wanty
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.
> Could add a <vitriol> tag to that
Is that equivalent to the already popular <rant>?
Hah, can we do that for recipes next?
Conversation add blocker unlocked
Exception that proves the rule. You know what context that specific recipient needs from you. GenAI usually doesn't.
Came here to say this as well. I've written and read full human-written essays on slack before AI.
With that said, I don't disagree with the article. Don't use more word when few work.
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?
I made a basic reference site around similar ideas a while ago: https://quick-answers.kronis.dev/
I think some people just prefer a more conversational format.
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, a back-and-forth is not needed, and the entire response is necessary for someone to understand to interact with.
This is when I open up a text editor, draft it, and paste that into Slack.
The next step is to not talk with each other at all.
Just have a LLM that "knows you well" in all your position argue by points and values assigned to the points with the LLM of the opposition.
If value alignment exists, a actual conversation may be engaged.
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.
Then at the end, "Use AI to make things clearer". NO! STOP USING AI AND JUST TALK!
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.
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.
https://translate.kagi.com/proofread
Why would you proofread in 2026?? Punctuation and grammar errors give further credence to human authorship.
I have stolen your link, dear sir
Thank you kindly for sharing
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.
I disagree. It would be a wonderful world where every overseas contractor that I interacted with used the AI tools in this fashion.
Even among native speakers, literacy is way down. AI could help with that… if people actually do the work.
That’s the real problem, not AI: no one wants to do the work. That is purely a PEBKAC situation.
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.
1. https://news.ycombinator.com/item?id=48219992#48221497
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 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.
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.
Did you read your own post?
> Should we use Redis or Memcached?
Couldn't they have used an example aimed at a broader audience?
I'm in IT but even I barely know what Redis or Memcached is about (never used either).
90% of people here know what those are.
And with a more broadly applicable example we could share the link with friends, family and coworkers who aren't on HN.
Yes that was exactly my point :)
But then it wouldn’t have tickled the HN reader quite the same way and wouldn’t have gotten voted to the top.
This doesn’t even need to be a website at all. This is pure slop designed in a pig lab for HN trough.
> This doesn’t even need to be a website at all
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.
AI was used in writing that article. Know how I can tell? "You asked a simple question. They lobbed a document."
" It's a weapon disguised as helpfulness" hello GPT
"Use AI to make things clearer, not longer. Let it sharpen your thinking, not replace it." holy slop goddamn
I swear most executives can barely read so you're not doing your career any favors sending them more than 150 characters.
The CEO of one firm I worked at wrote emails totally in bullet point format.
Made it much easier to read and you could just reply with:
> bullet point
response
which made life much easier
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.
It's called an executive summary because that's the only thing the executive will read of your carefully put together report.
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.
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.
Just prompt them back: "that's a lot of detail, could you please summarise as briefly as possible what differences concern our requirements specifically?"
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?
Maybe these people don't understand the impact of walls of text because they're not reading in the first place?
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...
LMGTFY definitely did not soften the blow, maybe it even increased the shaming factor lmao
Reminds me about similar "manifestos" about netiquette, properly asking questions, searching web and answering emails. And I expect exactly the same impact - none.
I actually saw nohello in some bios. About 2-3 times.
Obviously you need to use an AI to summarise the wall of text generated by the AI. Duh.
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.
There’s someone in this thread unironically suggesting this.
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.
It’s a classic tragedy of the commons situation.
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.
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…
I am so tired of these people, but it’s so sad they don’t understand themselves how ridiculous they are
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.
I’ve been thinking about this one a lot! Wrote a post on it a little while ago: https://productnow.ai/blogs/write-for-human-download-time
But I really agree with use AI to make your communication sharper. I think a lot of us, especially in corporate settings could use the help
Another great one is https://dontasktoask.com/
reminds me of https://stopsloppypasta.ai/
related: https://news.ycombinator.com/item?id=47389570
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.
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:
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):
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 :).This is a variant of "Computational Kindness"
Yep indeed, if I discuss with you I want YOUR opinion.
If I wanted a generic opinion... I wouldn't bother you.
We desperately need some cultural norms and taboos to develop around AI usage.
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.
The stated problem is so context dependant that this is borderline useless and quite hostile.
> 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.
Yikes, first LLMs took my emdash away and now they are coming for my verbose qualifications.
Darn and I was hoping we would see a new invention someone could form1 with the BATFE.
That’s interesting. When I use AI to help me write chat messages it’s almost always, “make this shorter,” or “clean this up”
Why do you use an AI to write chat messages?
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.
>Lots of people lack confidence...
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.
“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.
"Sorry, I asked you, if I wanted to ask AI I'd do it myself, if you don't know just say"
That's a good one.
This is slop too though, right?
> 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
I found the writing clear, concise, and human.
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"
Eventually you just get a sense for these things.
99% confident the article denouncing slop is itself slop.
Genuine AIDS. It'd be hilarious if it wasn't so terrifying and didn't happen with such regularity.
All slopposters should get perma-lobotomized ^
Heres a sentence you will never have an ai say
What makes you think this is AI slop?...
> You asked a simple question. They lobbed a document.
> It's a weapon disguised as helpfulness.
These are particular sentences I find questionable. Would you write that way? I certainly wouldn't.
GPTZero is by no means perfect, but it agreed this was likely generated.
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
Not the parent, but yes I would/do write that way for effect.
Bro, do i got the bad news for you
The heading " Why it's wrong" and the structure:
>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
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.
My boss.
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.
I had a boss who was the opposite. He was so terse in his emails that if you just read one without context you'd have no idea what he was asking for.
>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.
I usually find out when meetings are when I get a text from someone asking "are you coming to the meeting?"
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.
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.
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.
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.
"You asked a simple question. They lobbed a document."
Oh look, another blog post that should have been a comment. No slop blogs either, loser.
Where should they have put the comment?
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
The particular question in the blogpost can just be answered by a skill. Once you ask enough questions, the solution becomes obvious at the end.
5 Claude Code skills I use every single day
https://youtu.be/EJyuu6zlQCg?t=80
ugh
Slop is not data is not information is not knowledge is not wisdom.
I like how the website matches the message. Short and Simple.
It's a matter of having good taste. But AI education will help.
Slop grenades are the new "let me google that for you"
Just post the prompt bro
Just pack the fries
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 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.
I like the naming. I tackled this same pitch with https://writelesswithai.com but a "slop grenade" is better, more memorable, a nice brand. Good work.
ps. register slopgrenade.com too
now I know what to call it, thanks
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.
"Why it's wrong"
But you can ask AI to summarize it. /s