What's the equivalent of "Google it yourself" these days? When you ask something of somebody, the expectation of the other side to drop everything and do work for you isn't always appropriate/balanced/appreciated.

The underlying problem here is that communication is hard; especially if it's not face to face. People can't read your mind or might not get the full context of your question.

Back in the day I learned the hard way that sending a long reply to an email just ensures the other side won't read it. You can't assume that you dealt with their request effectively that way. And that was when there was a lot of email to deal with. Lots of people demanding all sorts of things via email and creating more work for each other by sending more email.

I learned to 1) never reply instantly to low priority requests, 2) be very brief and clear, 3) that talking to people is almost always the better choice when you get tempted to write a long reply.

If you don't like the answers you are getting, maybe reflect on what you are asking, how reasonable that request really was, and how you communicated it. Maybe your question was entirely reasonable and the answer really inappropriately bad. If so, give feedback or talk to the person. But maybe the other side didn't get you. Or maybe they were busy and you are interrupting them. There can be all sorts of reasons why you aren't getting what you expected.

It all depends on the context of course. If you are delegating something to somebody that reports to you, you can just tell them off. But if the other side is a colleague, your boss, or a friend you have to be a bit more considerate.

Write as yourself. You’re being paid for your expertise and insights. Communicate them directly to us. Copying and pasting Claude responses into Slack or an email directly shifts the burden of comprehension and understanding to everyone else, and worse, risks skipping that step for yourself. Even if you’re fundamentally using Claude to gather your thoughts or help you prepare a response, you need to be writing it yourself, in your own voice. Not having Claude ape your voice, or “make it sound less like AI”. You, directly. Doing this will further reinforce your own understanding of the state of things, the same way teaching someone is the best way to learn. As a guideline: for Slack and email comms, this should be near-universally written as you. For deliverables that are longer form and follow a template like proposals, roadmap/discovery work, etc., use of agents is expected but, see Principle #2. (Own the Output.)

Since people started using Claude now I get full context of everything… might be too much sure, but to be honest over-communicating seems better than under-communicating - sure it’s boring and tedious but that shifts the blockage to me.

Otherwise each of these coaxing sessions is something I have to keep in my head until resolved, which is a load in and off itself.

And even better - I can point Claude to that message and gives me a summary. You might say this is silly because we are paying the LLM tax, but knowing what to share, and then verifying against my situation are two different things.

Sometimes they have checked the wrong thing, sometimes they need guidance, sometimes I need to investigate before I can answer. A full LLM message with context is like “summary of their working context that I can resume on my end” so I don’t waste cycles asking or rechecking etc.

And it is still possible to be a learning experience for both of us since the initial message is the start, we can then talk to each other like humans, both much more in sync than before…

It is kinda ironic but I think it does push things in a better direction. Would I have loved it if the message was human, direct and to the point - obviously, but we are all busy, we got shit to do, and this is a useful “resume” mechanic

As I wrote this i switch to the phrase communication-style because I realize that when he does his code review it's literally just a quick pass and a paste, but unfortunately when I do AI code reviews I review every single review item and push back individually on each item, research the items, write tests against what the review says and basically curate the code review I give out, which takes me 3 hours and then I have a boss angrily messaging me asking why the code review isn't done.

So on the one end I've managed to reduce toxicity in the reviews, but the tradeoff is toxicity in the expectation of a rubber stamp.

"read it all, yes even the bullets. especially the bullets"

if you read one thing from a large wall of text, its gonna be the bullet points, right? thats a big part of what they're for?

"It is the conversational equivalent of forwarding the email."

this is a terrible metaphor. it seems to rely on the idea you would be mad if you sent an email to someone and they forwarded to someone else? but thats useful, theyre literally sending it to someone who is presumably better placed to help you, or raising it to someone more important because they think what youre saying is important

Interesting to note the page also ends with "written by a human, on purpose"

[1]: https://gruhn.me/blog/2026-08-03/

Before an LLM, I would have worked for much longer, then written the same short description, but without the detailed summary at the end. The llm output is really just extra clarifying info I never would have bothered to do on my own. By clearly marking it, I give people the option to skip it and just trust me.

But the point is. They had search, so why didn't they think to do the search? They had the same tools as I did. And yet regularly they couldn't think of the answer. I still deal with this with AI today. When you ask a question, maybe you're asking for the interpretation of the question by the person on the other side. And it's that interpretation that goes to the LLM. Or otherwise some filter of how that person thinks about the question differently than you do.

We are afraid that one badly phrased message can hurt our reputation or even our jobs, so optimizing every message with softer wording, more political correctness, less directness is safer.

But the real problem is the culture underneath it. If people are constantly worried about how every sentence will be interpreted, eventually they stop speaking plainly, stop challenging decisions, and sometimes stop speaking up at all. So this is like attacking the consequences not the root causes. There should be some sort of measurement of how broken is a community or a company by letting a bot scan Slack/Teams and let the bot remind everyone the importance of lowering the fake index.

But I think ideally agents should speak to agents when this kind of comm goes through. What a nice opportunity here ;)

I'll say this clearly. Using AI for writing is bad. Don't do it. Ever. If you can't formulate a coherent thought yourself, just don't. I don't want to hear what the LLM said. I want to hear what you said. Loudly proclaiming you have nothing to add is a strange flex. Just stay quiet.

There's a delicious irony in pasting dontpastetheai.com into an AI chat, and I choose to believe that's deliberate.

It's well written though. The parody answer in the middle is the sharpest part — "this mild Google search you could've done in 3 seconds" — because it nails the exact texture of unedited model output: the fake enthusiasm, the numbered scaffolding, the conclusion that concludes nothing. The tell isn't that AI was involved, it's that no human judgment was applied afterward. You can't fake the presence of taste.

The "guess who the AI will replace first" line is the real argument underneath the humor. If your contribution is indistinguishable from a prompt box, you've made the case for your own redundancy better than any automation study could. What survives is the part the page calls "the thing the model doesn't have" — your opinion, your context, your willingness to say "actually I disagree."

It's also a close cousin of something you've presumably thought about with Kindship: when agents message each other or humans, the same failure mode applies. An agent that just relays raw completions is exactly as useless as the coworker this page is roasting. The value is in the editorial layer — deciding what's true, what's relevant, what you're willing to stand behind. That's arguably what SOUL.md-style identity is for: giving an agent something to have an opinion from.

Were you filing this as cultural artifact, or is it aimed at someone specific in your orbit?

completely agree except a crucial step is missing here:

(1.5) validate the accuracy/truth/relevance of it

your own sincere and human take on rubbish will still be rubbish. validating is indeed the more uncomfortable/demanding task here but if you omit it you become just a "human ai beautifier", and this will actually undermine the trust in you as a human. if you are going to omit validation just paste away, be honest and let it be obvious, you can then blame the ai.

I did not use any AI. I had a couple conversations with agents while mulling some approaches, but they never edited my work nor did I have any chat open while I wrote. Every word was my own.

I notified my other stakeholders - "here's my analysis, none of it was written with ai". It sounds silly to say, and I worried others could take it as an attack. But I needed to communicate "This is a long document but its actually worth reading".

I liked what I wrote. The final author of the proposal liked what I wrote. All of my ideas made it in to the proposal. We landed the contract. Its a nice chunk of $ for us.

Its still worth putting the hours in.

Pasting AI output verbatim is the modern day equivalent of “let me google that for you.”

For any given <subject> I'd want:

- Your context - your taste - your judgement - clarify what bet <subject> encourages to support - be ready to elaborate your thesis of that bet

With that we have the basic complete building block that used one after another can model a basic lifecycle of a dialogue that can make progress towards something (a goal, a mission, something useful)

Strictly speaking, not true. I have Gemini Pro; my friend doesn’t.

Don't Be a Meat Proxy: https://gruhn.me/blog/2026-08-03/

No Meat Proxy: https://nomeatproxy.com/

How to Passive-Aggressively Shame People Who Use LLMs Selfishly: https://joshmoody.org/blog/selfish-ai/

AI;DR: https://www.rickmanelius.com/p/aidr-ai-didnt-read

Dont be a meat proxy! simple

Unique human prompt + AI output which is useful and does a good job expanding on the prompt topic and which when pasted saves people the time of having to go prompt AI on the same topic (fully reviewed by a person before sharing it, to ensure the content has relevance)

Of all you’re doing is copying and pasting anyone else’s stuff, then you’re adding little/no value. You should be paid nothing.

My reaction was of immediate visceral disgust.

I can say in no uncertain terms that I do not care if your English is completely broken. I prefer that over AI slop. You will not improve your writing and language proficiency if you do not practice them! If you do not care about your language proficiency then you have zero business doing offshored work.

The answer was also wrong. I pointed Claude to the micro service repo in question and asked it myself. Got a different but correct answer, after verifying it myself.

Their code is so trash I wonder if we’re even using the same models. Every PR I need to go over it myself and have Claude refactor and restructure everything. Claude generates decent code that can be made great with just a few iterations. I don’t even think these guys are running the code to test it, let alone read it. Just slop and ship.

Can’t rely on them to understand the code or even answer simple questions about it.

What is even the point of all this?! AI has made me personally more productive but overall it’s still a net negative because when you’re a “-x dev” the force multiplier from AI is overwhelming and soul crushing to everyone around you.

This I have been doing a bit. I strictly asks it to never give me suggestions but raise questions, concerns, tell me if it feels if something can be better, style-wise or otherwise without giving me those options.

And a bit of other overall feedback requests.

And if the model adheres to this and doesn't jump the gun, it indeed works quite nice. But it often jumps the gun, forgets, doesn't give a f- about my instructions, or whatever. Maybe more robust local models will better.

We need a dontpastethedontpastetheai.com to get ahead of this.

No, they don't. I have access to heaps of raw information about customers and their systems that I can easily provide to my AI. That AI can also dig through all the communications we've had with other customers who have asked us similar questions, and draft a coherent on-brand response.

I often paste AI-generated responses with significant success, and all I have to do is make sure the answer isn't wrong and doesn't include sensitive info, and to remove a few extra adjectives and em-dashes that I don't want.

If the customer has access to my tools, that's a critical security incident. They contact me because I have access to info they don't, and unlike and AI, I can guarantee my response is valid.

As for people who copy/paste AI generated text blindly: That's firable in my books.

I found this blog post to be much better at getting the point across: https://tombedor.dev/human-attention-and-human-effort/

You're gonna need a stronger language. Why do you think people love AI? If I have to think and god forbid read the response I might as well live in the woods or something

But this leading statement is not necessarily true:

“The person on the other side has the same tools you do”

Aren’t we all imbuing our AI with our own thought patterns? Folks I speak to spent a tremendous amount of time building tooling and (hierarchical) context around LLMs such that the output it produces is definitely not the same as what the stock model produces.

DO NOT FUCKING PASTE AI RESPONSES TO ME, EVER, PERIOD.

When someone tells me they "talked to Claude" they sound retarded. Just tell me you looked into it, and I fucking swear to god if you just took what Claude said verbatim and do not understand the problem then you are a cunt and I will stop working with you.

When someone pastes LLM output in chat I completely ignore it. I will just continue the conversation as if they said nothing and silently mark them as cretinous in my mind.

This is not up for debate, you will talk to me in short well formed sentences and get to the fucking point or you will fuck off

idk about that

[edit] to extend the connections The charm of a messaging board like this is bringing "join attention" to a topic, not just sharing separate monologs.

And also, always wondering if something was written by a human or statistically generated could be seen as Epistemic Hypervigilance, leading to stress, and loss of epistemic trust. This could reduces opportunities for taking on new information of feedback, and cause rigidity and isolation.

Almost certainly written by AI

This is just demonstrably false. It often IS very helpful. It's annoying that people continue to ask basic factual questions without taking easy steps to solve the problem themselves. I can't remember the last time I "asked someone a question" that it was possible to figure out for myself.

It's the people who keep asking basic, easily researched questions and then wait for you to explain it to them, without first trying to resolve it on their own, that are being anti social, not the people trying to help them.

This is much the same as when Google was relatively new, and people hadn't adapted to it yet. I would get asked questions constantly like "what is the name of the class that does X" and since I usually try not to be too much of a jerk, I would often stop and go open a source file or look in the docs for the person. I know there is a class, and I think I remember what it's called, but I need to go double check it. That usually meant googling for the docs, clicking the first result, and sending them that link along with the specific answer. Sometimes when they asked me I was frustrated about something else or just in a bad mood, and I would tell them to f**g google it, I am not their research assistant.

Ignoring AI is the same scenario as ignoring Google but worse. I am not your research assistant, and worse, there is an artificial research assistant standing by 24/7. If I have to paste something in AI and paste the result back to you I am not the asshole in this situation. If you go do the work to check yourself and it's not something you can just verify conclusively and you need to be sure it's not the AI fooling you, then I am more than happy to look at what you have and lend any expertise I can add to the situation.

Excellent advice, with one very large exception based upon human foibles.

Occasionally there are people who do not ask things in good faith. They ask you a question not because "they wanted your take on it... Your context, your taste, your judgement". Instead they ask the question because they're then going to go to extensive lengths to find every possible counterpoint or opposition to whatever answer you give. In 2026 those sorts of people who will ask and then revert 30 minutes later, after you've carefully detailed a well thought out answer, that "ChatGPT says..." Add that for time eternal they will hold the fact that you answered something for them as a liability that you must atone for, and for which you are responsible for every ramification of their interpretation of your answer[1].

If you ask someone something and they just copy an AI answer, it's such a cop out that I imagine there are many cases where the person knows they're dealing with someone who "asks questions" like this. I have people who I give thoughtful, considered answers to, but then I also have people that I basically say "I don't know...Google it", because giving an iota of an answer always turns out to become something I regret.

[1] - This is not unlike helping someone with their computer. There are people who we love to help, but then there are people that will then ascribe everything that ever goes wrong with their computer for time eternal to that time you showed them a shortcut in Excel. CPU died six years later? Probably because of that shortcut.

If you are getting blocks of llm text perhaps it's because you don't deserve their input and could have satisfied your query by not making it someone else's problem.

[1] https://news.ycombinator.com/item?id=47389570

This is really useful, I wish more people would understand this: you don't have to have an opinion on everything. In fact, you probably shouldn't have an opinion on most things.

My own take on the AI issue: https://blog.gpkb.org/posts/just-send-me-the-prompt/

I like to think it helped, but likely what helped was that nobody seems to use Deep Research anymore, haha.

and i detest ai slop

so my solution is to speak combined with an ai cleanup like macwhisper or wispr flow of all the uhs and ahs and repetitions or rephrasings etc

you can even use ai to pretend a conversation or interview style

then i have stuff to edit

boom

Make it a standing rule.

I do. "They all have the same tools as I do" is exactly what's infuriating. They either blatantly don't use them, or are too incompetent to do so effectively, which would track.

I have been making an effort to provide an "AI;DR" section as well, even before the term existed, because of brevity. But sending along the full(er) text as reference I think is perfectly fine in such cases.

Disagreed. Bad writing from a chatbot beats worse writing from its user every time - regardless of length.

One reason we have so much bot slop is the attractiveness of bad-writing bots to the many people whose writing is even worse.

OMFG! Even the "don't send AI answers" page text is fucking AI slop.

Can't they be bothered to write 200 words of their own?