ZH version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
63% Positive
Analyzed from 1461 words in the discussion.
Trending Topics
#don#something#modern#read#reading#claude#llm#writing#ftp#article
Discussion (39 Comments)Read Original on HackerNews
But I do care about not having to read the word „genuinely“ a hundred times just because Claude likes it so much.
I read the headlines, skimmed the prose that has no purpose other than 'it was what the AI put out', then read the conclusion.
Similarly, I skip video content where the title is something like "How to get better fuel economy for your car" and the video starts "Let's investigate the entire history of wheels", and the whole video could just be a guy sat in a chair saying "keep your tyres at the correct pressure"
And once you notice the bait (via telltale signs) you feel betrayed. Betrayed for your attention, thoughts and time.
> Now a beginner's tutorial opens with sixteen tools you've never heard of, half of them named after Japanese words for "fast," and the first command downloads more code than the Apollo guidance computer ran on, just to render a contact form.
After finishing the post, yes OP was funny, at the beginning only unfortunately. The rest of text is so LLMish that it hurts.
The beginning was promising…
https://www.newyorker.com/culture/infinite-scroll/the-ai-des...
I was GENUINELY (ha, see what I did there?) interested and started reading, only to immediately understand that no human sweat was spent in writing what I was reading.
The result is that within 2 minutes I went from "fantastic, just what I needed" to "fuck, no way I'm spending my time reading this, considering it's probably hallucinating parts that could not work as described". It doesn't matter if text is perfect, it's the "vibe of the vibe" that just puts me off. Plus, the design SCREAMS Claude Code.
So, to everyone who does or thinks of doing what the author of this text did, a pledge:
PLEASE, take time to implement your ideas. That's what gives them value. Nobody forces you to produce this slop nor to implement your ideas, even if you think it's the best in the world.
If you do, and you think that shortcutting like this equals being productive, you are sorely delusional. You are producing slop, and you are only contributing to noise. And if you do that in the hope that something produced this way might eventually take off in a game of large numbers, well you're just throwing shit at the wall to see what sticks. Some of it might stick. All of it will still be shit.
When I was first learning web development in the 90s, after a couple of sites, I switched to a hosting provider that offered a bit more control over the hosting environment. I discovered that they supported SCP and SFTP but not FTP. When I asked them why, they told me it was because FTP was old and insecure. I dig a bit of digging to learn about the situation, and sure enough, FTP was 25+ years old already and literally a decade older than the Internet (FTP: 1971; TCP/IP: 1981) and was obsolete and much less secure than the modern alternatives available in the late 90s.
This was a full decade before “the olden times” of 2008 that the article talks about, so using FTP even back then was a massively outdated way of working.
It's AI slop.
I don’t seem to have cracked the Hacker News writing formula yet. A few years ago I got similar feedback, but for the opposite reason: people complained that my English felt awkward because I’m not a native speaker. This time I asked an LLM to smooth out the wording, and it seems I accidentally outsourced my personality too.
The content and research are mine. I spent a few hours trying to collect all the moving pieces into something I wish I’d had as someone returning to frontend after many years away. It was just meant to save a few fellow dinosaurs from opening 47 tabs and wondering why they suddenly need seven package managers.
Can you put the pre-LLM version of your post on your site somewhere? I (and presumably many others) would be much more interested in reading that, even if the English is a bit awkward.
Many sites could be built on the stack espoused by Alex Petros, "The Hundred-Year Web Service", <https://www.youtube.com/watch?v=lASLZ9TgXyc>: SQLite, Express.js, Nunjucks, and HTMX.
But if your app needs islands of interactivity, my library does it in a way that I haven't seen elsewhere.
A lot of people on this site complain about modern tooling but this stuff wasn't done for kicks. React (and npm, I think) was (were) _so good_ at solving the problems it solved that we were willing to deal with the fallout.
Now that we're back to consolidating all of the gains through server side rendering, I think we're in a much better place overall.
Contrary to popular belief, even big corporate web dev projects for high profile clients can still be, and often are, just plain HTML and CSS. The design does most of the heavy lifting. This is especially true for anything related to marketing.
I have noticed myself having a visceral reaction to the AI tells now, somewhere between disgust and anger. And I am not an angry man and I like AI! There is just something about the voice that is incredibly off-putting, like a know-it-all friend who just ripped a giant fart. It is annoyingly correct and also smells really bad.
https://rebeccahazelton.com/
"the generated code quietly assumes you know everything in the eight layers above"
These AI tells are getting really easy to notice. A negative that absolutely isn't needed in a sentence, and you know it's AI that wrote it (, not a human ;) ).