alt Hacker NewsAI;DR (AI; Didn't Read)
Comments
I think the main reason many people (including me), very often, lack the motivation to read content that is likely generated by AI is the suspicion that it comes from a place of intellectual laziness. Another reason, based on personal experience, is that AI content may suffer from too much verbosity, too much jargon and over-confidence, which makes the reading experience feel fake and border-line irritating. In many cases the content may have very little to no nuance, which is ultimately a waste of time. As an anecdote, someone posted a blogpost on Linkedin on using agents to implement a driver to access PCIe devices over TCP/IP. I was intrigued because that's not an easy task for several reasons, like handling PCIe interrupts and DMA. For exmaple, how does the remote machine map the device's PCIe BARs? And when it issues I/O to the devices registers, how are these reads and writes transferred to the remote device. In the end, this is just some virtual memory. In a local machine, this is either directly mapped to the PCIe physical addresses or some IOMMU virtual address space which is then translated by the hardware upon CPU/device/VM access.
After reading the long verbose promising article, in the end, the guy (with the help of the agent) only managed to implement access to the PCIe config space so that lspci on the remote machine works and shows the remote PCIe device, but that's all. It never addressed the issues above nor even mentioned them. The code was AI generated. The article was AI-written. The article never made a reference to DMA, interrupts, MSIX-X, IOMMU, IOTLB, virtual memory, etc, but it made big claims on next-gen datacenter disaggregated architecture, boosting GPU utilization, reducing large scale inference costs, etc.
Anyway, you get my point: big long beautiful words, but zero nuance.
The part that astonishes me is that in the year of our common era two thousand twenty-six that it's not universally offensive and reviling to post an AI-generated response to another person.
If I'm reading something on the internet, I'm either reading it to learn, or I'm reading it to be persuaded. If I wanted the LLM to teach me (thank you, no), I would ask an LLM. I'm reading your website/newsletter/email because I want to hear from you.. If you can't be bothered to put your time into writing it and teaching me what you think, why should I be bothered to read it?
I really like the point I read somewhere the other day; instead of sending me the AI output, just send me the prompt you used to generate it. That is the only part that contains only the information you are trying to convey. The rest is just guesses flowery language added, and it confuses the actual message being sent.
> Look, I get it. It’s Q3 2026, and we should expect that everyone is utilizing AI at SOME point in their process (sourcing ideas, creating outlines, refining prose, etc.).
Sounds like this isn't everyone, but rather, it's people who have nothing to say, and -- even after they've "sourced ideas" -- they don't know how to reason about it, nor how to communicate it.
This isn't everyone. This is people engaged in generating noise, for the sake of noise.
If some text is AI written as a response to a much shorter prompt, then the prompt and/or sources used to make the text should be published instead (or at the very least together with the text).
Don’t be a meat proxy [0] is another one I like
3 years ago; https://news.ycombinator.com/item?id=35638229
One of the ways I think about this is to treat AI as if it were another person. If I want to give information to Bob but I think Alice might do a better job of writing something, I'm not going to ask Alice and then copy-paste her response to Bob. I'd either tell Bob to ask Alice, or loop Alice into the conversation. Me being the middleman between Alice and Bob can cause several issues including delayed communication, miscommunication, and obscured provenance.
Alice being AI or a real person doesn't change those factors much.
I am writing excellent documentation with AI. The key success is to have rules to write documentation, PRs, commits, etc. I use ASD-STE100 and some constraints on what to write, how, and what it matters to my projects.
For me the bar is still quality. Sure it bothers me if I search for something and find a crappy AI written tutorial that answers none of my questions, but if I found a high quality well written, insightful, and informative tutorial, I would not care if it were AI written.
I guess the same thing goes for articles written by humans. I don't really read the articles that are "Shower thoughts that I think are profound" or "The way I do things is the way everyone should do things"
There is a fairly common wisdom that the way you get good at writing is by writing a lot, I'm prepared to cut people a lot of leeway at people putting all of the stuff out there before they are good at writing. It's just the path you have to take. That's OK, because I don't have to read it.
And I'm fine with people making crappy AI generated tutorials while they figure out how to make better AI generated tutorials. I am not fond of people who are creating bulk content by volume without any regard to quality or improvement. Some may be doing this to make money, and maybe that works, I suspect for the most part it doesn't. I don't appreciate it though, whether it be AI or human produced bulk content.
I'd spare a thought for those who feel that they are producing high quality content when they are not. Everyone has a poor perspective on something sometimes. Usually the harshest punishment they will face is the feelings their future selves experience when they look upon their work with a more balanced perspective.
Recently I've gotten lots of DM's from boutique marketing firms offering to promote the game i'm working on. I'm pretty sure they all find leads via scraping the Steam API. Many of these small firms are people in developing countries.
It's frustrating dealing with them since they all use AI in communications. Everything is written so flowery and with so much extra words. Any feedback I give is met with overwhelming praise about "how right I am". Its exhausting.
I understand many of them are in developing countries and aren't confident in their English skills. But I always tell them I'd rather they write in broken English to communicate their own ideas. I want to hear what they have to say, directly from their brain. I don't have issues with parsing bad english. Usually its straight and to the point.
They always end up defaulting back to AI messaging. Even when I tell them to be genuine. I'd love to give one of these firms a chance, but they fail this basic step. I think they think everyone in the developed world likes AI or something.
> Yes, there are certain situations in which we should expect 100% AI-generated copy. Customer support would be a perfect example.
hard disagree on this
I once spent two days on a pr and got an obviously AI generated review that contradicted what we agreed one before. So I had AI respond to it. The next day he asked if I'd used AI. I used the same justification he'd used for his review. Fight magic with magic. He never reviewed my pr that way again.
I used to have a rolling product satire I called "Claude Handshake" which basically explained how cold-email is dead today as people feed email contents into Claude and Claude responds with a draft, that usually comes from another person doing the same thing.
The idea was to just avoid sending email altogether and just simulate the interaction entirely using Claude.
AI comments are becoming a second documentation system nobody asked for. If the code is self-explanatory, “why this is good” is usually just noise.
Anyone else have coworkers responding to your human PR review comments with AI? I feel like I'm taking crazy pills, how can anyone find this socially acceptable?
I've asked questions to people who wrote interesting articles where I didn't know it was AI generated. But then, they couldn't really answer me because most examples in the article wasn't something they experienced themselves. So basically, we were both reading the article for the first time, with made up scenarios.
Whenever I'm tempted to base my judgment on how something's been created rather than what the end result is, I think about the scene where Sheldon throws away the french toast Penny made for him, despite smelling good, just because "Monday is oatmeal day."
The problem is that in most domains we reward effort sometimes even more than results, especially in fields where results are hard to verify. A wall of text is the classic way to signal effort and delay feedback, a signal which was costly pre-AI, but now is not costly. However norms have not shifted.
I personally don't care whether a piece of writing was AI-generated as long as it's useful or insightful.
However there's only 10% of AI-generated writing that is worth reading. If some articles convince me that there's somebody behind it contributing valuable understanding, I'm willing to overlook the telltale AI mannerisms. It's understandable that some domain experts aren't very proficient in English.
That said, statistically speaking 90% AI-generated pieces are a waste of readers' time. It's a good strategy to simply skip them.
I know this looks like a bot comment. It's not. It's a much older style of writing known as Chinglish. That's why as a non-native English speaker, I can totally understand why someone would want to use AI to help with their writing.
I have been feeling the same way for the past few months. I just want to say I will never read a paragraph that has signs of AI, because most of the time, the people didn't put much effort in it. I don't care if it is factually correct or not.
We could put comments in separate files with an annotation format. For example: ion.cpp.cmt And autoremove all comments from source files. Then use plug-ins to toggle comment files within the text editor.
Not sure how others are using AI, but for us, an AI dump is actually much better than some human editing/writing it. What we've found is that people become gatekeepers of context due to their role in the company they have privileged access to certain context, so when some question is asked, the person who's AI has the right context can look at it and generate a response making sure not to leak anything important and making sure the question is not nefarious. Also important that your AI doesn't generate a wall of text but is super concise. This works incredibly well for us.
If a coworker sends me an AI generated report to help on my task, I am more likely to ask them to just send me the prompt and model they used. Let me own the session and steer it however I want.
> I’m SUPER jealous that I didn’t think of this first...
Not to toot my own horn, but...[0]
On a more serious note, we're planning a trip with family, and my mother-in-law asked if I'd read "the planning doc" yet.
It was a shared ChatGPT thread. Thankfully I read enough LLM output to skim it easily and avoid offending her (:
"Look, I get it. It’s Q3 2026, and we should expect that everyone is utilizing AI at SOME point in their process (sourcing ideas, creating outlines, refining prose, etc.)."
Hell no. Sourcing ideas? From AI? Are you kidding? Essentially guaranteeing they won't be original? Refining prose? Making it much more likely it'll be generic as hell. Call me old fashioned, but while I can certainly understand asking LLMs concrete questions - how do I fix this, where can I unsubscribe from that - using AI to help communicate seems bass-ackwards. LLMs have no idea what you're talking about - they're free associating probabilistic sentences! What on earth value could that have in communicating. They self contradict, hallucinate, speak most frequently in a weird smug corporatese. How is this an attractive use case of the technology?
There is a sort of symmetry here with there also being things that everyone writes that no one reads.
* Cover Letters
* Most test reports
* Lots of docs required by the government
* I'm sure there are more you can think of
So now we can have stuff written that no-one ever writes and no-one ever reads? Hmm...
The fear I have is that I accuse of AI without being 100% certain. I've not found a good way around this except to talk to the person.
This has really started to hit my radar because people are filing PRs with AI generated commit messages that read like a bill being presented to congress. Holy shit, give me a summary.
I don't know if the contributors are writing any of the code they're submitting or just using AI to hop on every 'help wanted' tagged ticket in every product database but if it looks mostly like a smart human wrote it, which usually takes two rounds of reviews, then it can probably go in.
But if you make me work for it more than you did I'm just going to close the PR and do it myself.
We have a junior research student who does this on slack.
Before I think he was using AI and other tools to translate his messages since english isn't his first language. His own words, then AI translated it. Sounded kinda clunky, but I could sense the human behind the words and I gave grace since I can only imagine how hard it is to properly communicate your ideas when English isn't your first language.
But now it's gotten to a point where I can tell he's not using it for just translation. I'm gonna need to chat with him.
Redacted example below
```Hey @PERSON_WHO_ASKED_QUESTION Both good, and the retrieval one isn't written down anywhere. In order.
One corpus or four. My lean is one. Same chunks and embeddings tables, source_type column to tell them apart.
Values I'd propose, flat rather than nested: paper, dataset_description, dataset_readme, dataset_contributors, dataset_records, dataset_files. Description and readme split because their units already differ, one row versus one row per paragraph. A discriminator that can't separate those isn't doing much. The existing 440 rows would need backfilling to paper.
Chunk id in the same spirit: dataset doi, source type, ord. So EXAMPLE_DOI.
Worth checking before any of this matters: does chunks.doi carry a foreign key to papers.doi in 0001_init.sql? If it does, a metadata chunk with a dataset doi can't go in that table at all, and separate storage stops being a choice. Ten second read, I haven't done it. Shout if you get there first.
Retrieval is the one that's bigger than it looks. Search once across everything and nothing guarantees a paper chunk and a metadata chunk both land in the top k. The facts we want relate the two, and the generator can only write those if it sees both sides in the same window. So if one type systematically wins the ranking, that class of fact doesn't get worse. It becomes impossible.
Which way it goes I don't know. Two mechanisms pull opposite ways. Metadata chunks are short, tens of tokens against roughly 450 for a paper chunk, so they may just lose. But we embed context header plus text, and on a short chunk the header is most of the vector. The header is the dataset name, which is also most of the query. That points the other way.
Cheaper to measure than argue. Load one dataset's metadata, run a normal dataset level query, record the rank of the first chunk of each source type. Runnable as soon as any one of our four subtasks lands.
After that it's one pool, per source with quotas, or one pool with a floor per source type. I'd rather not pick before there's a measurement. ```
Totally agree with the core argument. I understand there must be some etiquette in the space of llm-writing where writers must respect the readers enough to edit and review. Also, pages and pages of LLM content is not something anyone should be proud about sharing. But we are in this world where condensing an argument is so easy that in the busy lives we live, sometimes some arguments are made with LLMs online. If those arguments are pages, I understand the backlash, but if it’s a terse argument, then I would urge others to address the argument made rather than categorizing them under AI;DR, it’s like ad-llminem in my view.
What do you think of AI generated infographics? Someone at my job is creating them on the daily and sending them around to leadership. I find it lazy and too busy.
Every time this comes up I have to ask, how do you know it is AI? Sure, sometimes there are strong tells, like if they forget to take off the "Do you want me to change anything" at the end.
But AI was trained on good writing. Good writing looks like AI. I've had to reduce my use of em dashes for example because they are an "AI tell". I used to like using the term "load bearing".
And the list gets longer every day. Eventually every mark of a good writer will be an AI tell.
I don't know how to resolve this. But I've been accused multiple times of copy/pasting AI when I did no such thing, and it's getting equally frustrating.
Throwing unread AI output at others is the new version of “sitting on your phone, oblivious to those spending time with you”. Every new consumer technology comes with a new way of being rude.
Valid and fair especially in the 95% of cases where the AI added a lot of unnecessary detail.
On the other hand, I suspect that we are pretty quickly going to get to the point within a few years where we realize that the best AIs are not just being overly verbose. But rather their communication is just more compressed and thorough.
So even though we still see a lot of jaggedness and weird failures in some places，in other ways I think this is actually an aspect of intelligence where AI has surpassed humans. Just in terms of being able to deal with a certain density of information.
I thought this was: I summarized with AI; didn't read.
I am pro AI, and also deeply share this sentiment.
I think sometimes people have a big long thread with Claude where they feel enthusiastic about the back and forth being very productive -- and then they genuinely want to share the summary of the conversation with their colleagues so that they can feel it too. It's not just typing a short prompt and copy/pasting the response.
But, Claude, especially Opus is so unnecessarily flowery, and smug -- and it's so identifiably characteristic.
It somehow rubs salt in the wound: "I didn't take the time to write this out in my own words, so now you get to listen to this petulant asshole mansplain it to you".
If it would get to the point, and with some humility, it would be a much different proposition.
> Customer support would be a perfect example. We’re not looking for artisanal “did you make sure to reset your phone” style dialogue.
Yeah...I'm probably dropping whatever services do this. If the company can't invest the capital to hire humans that help humans, then what does that say about their products?
We just need better specialized models for these. Writing is as much a process of thinking as of communicating the idea. Current generation of outputs eliminate the process of thinking
I find after 4 or 5 rounds of targeted feedback and requests to tighten the text AI can produce exceptionally readable documentation.
There is a light at the end of the tunnel for all the minds burned out by AI slop.
I already saw someone getting fired for only producing AI text as part of their entire output, after being unable to explain what they “wrote” in multiple situations.
And recently my company enacted a mandate that text made for humans should not be AI generated.
Sooner or later every company will start realizing that this is just people too lazy to actually work and coasting on a paycheck, at the expense of every other worker.
Related:
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/
While I don't always agree with everything Drew writes, the title of his blog post[1], "Please stop externalising your costs directly into my face," really resonated with me.
At the end of the day I don't care how much of what you're presenting is your own words vs AI generated, as long as I can see evidence that you have reviewed and understood what you've passed to me. My resentment and ire is raised when the onus is pushed on me to be the first real human reviewer, having to do all the weighing of merit, practicality and in some cases plausibility that the quote un-quote author should have done before ever hitting send. It's passing the buck and an abdication of responsibility. It is lazy and makes more work for those you foist it upon when your role should be making the effort in your communications to make it easier for the recipient. Not wasting their time having to spend orders of magnitude more effort refuting bullshit you shouldn't have put out there in the first place.
It is something that has always pissed me off, however AI has made it fair easier for lazy people to think they sound smart and externalise their costs into my face.
[1] https://drewdevault.com/blog/Stop-externalizing-your-costs-o...
> ...then I’m afraid I received a different message than you intended
Are you sure about that?
My coworkers continue to dump hundreds of lines of AI documentation in every PR and every other line of code has between one and ten lines of AI generated comments, talking about the real unlock and how things are byte for byte identical on the load bearing path or how the acceptance ladder is misleading.
Features are coming out and metrics are improving, but we’re basically in a post readability code base, with the occasional performative comment about a variable name.
I don’t really know how to address this situation or if it needs addressed. I certainly don’t read the long-winded AI comments or the AI documentation, but perhaps it’s useful for the AI on its next pass.