alt Hacker NewsNo Slop Grenade
Comments
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.
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 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.
> 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.
Then at the end, "Use AI to make things clearer". NO! STOP USING AI AND JUST TALK!
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.
> 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).
I swear most executives can barely read so you're not doing your career any favors sending them more than 150 characters.
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 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?"
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...
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?
Reminds me about similar "manifestos" about netiquette, properly asking questions, searching web and answering emails. And I expect exactly the same impact - none.
Obviously you need to use an AI to summarise the wall of text generated by the AI. Duh.
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.
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 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/
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.
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."
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.
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.
Darn and I was hoping we would see a new invention someone could form1 with the BATFE.
Yikes, first LLMs took my emdash away and now they are coming for my verbose qualifications.
That’s interesting. When I use AI to help me write chat messages it’s almost always, “make this shorter,” or “clean this up”
“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.
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 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.
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.
"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.
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"
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.
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
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).