Hacker Newsnew | past | comments | ask | show | jobs | submit | penetrarthur's commentslogin

Next love in 0:19, can't wait


God damn it. Can people write interesting articles in NORMAL writing style nowadays? Why is everyone writing in these stupid short "punchline" sentences?

Seriously. This is trash. It presents no evidence, contains no original ideas, it’s just written—excuse me, generated—to be as provocative as possible.

I think I’ll just start flagging these. They’re just a new kind of spam.


It's exhausting to read.

I can't quite put my finger on it; obviously the "it's not this. It's that." is part of it, but even without the obvious tells that writing was AI-generated/improved, it's just so tiring to read?

Maybe a linguist can chime in why all these texts are so samey, cloying and annoying to read? Is it (just) the pacing?


I wonder if part of it is that we're mentally trying to get the actual meaning and thoughts out of it. It's inflated like trying to read a bad students essay that's struggling for word count? I wish people would just post their prompts directly.

It reminds me of the "overenthusiastic youtuber" presentation style, with jump cuts etc., just in written form. From its prevalence I can only assume that some audiences prefer it - I'd be more interested to know why that is.

or "reels" equivalent of an article

It's like a LinkedIn post, stretched to 3000 words.

Because it's not people doing the writing.

That's ChatGPT for you:

> The ... isn’t just ... . It’s ... .


This is just myth and faith. Even if all AI wrote like that, it doesn't follow that all writing in that style is by AI, hence the belief in the style. Focus less on the aesthetics, more on the message. After all, for this article to have been written in the form of a sonnet was just a prompt away.

These people do not put in any effort and go for defaults. I see this with images too.

It reads more like a transcript for a podcast. Somewhere along the way, we no longer favor illustrative anecdotes, logical arguments, or dialectical arguments. Everything is a podcast now.

All of a sudden everybody is a writing style critic. The only question that is pertinent is if the message of the post is relevant.

Think about it. You wouldn't give someone crap for writing in broken English because there are many really smart people that are non English speakers. So why are we giving crap for people using AI to write better posts? If the idea is relevant, what's the point in criticizing the style?

A fair question would be "is the idea in the post actually the writer's or was it entirely done by AI"? However how can one actually tell if the idea, not the style is original? You can't. So it's pointless to be angry about style. Focus on the message.


No. I have a lot of respect for people who write in a second language. You can often tell because the content is thoughtful and some of the word choices or grammar is quirky.

This is a bozo who prompted the machine for a viral essay. He did not write anything. He does not know anything.


Do you know this for a fact? How? Honest question.

If I’m wrong I’ll eat my hat.

But I’ve spent enough time with these tools and coaching people on writing over the years to recognize the extremely low signal to noise ratio and prompted style instructions. I’m equally confident the gentleman in my spam folder is not a Nigerian prince.


Understood

> So why are we giving crap for people using AI to write better posts?

Pretty sure the point is it's making the post worse, not better.


Not talking about this post anymore, because it being flagged is more reason it's just click bait.

But in all honesty, for some it's extremely hard to write concisely in English. Maybe what you call worse is actually better in comparison to what the person could write by himself, insights notwithstanding.


"is the idea in the post actually the writer's or was it entirely done by AI"

Look. If you don't want your readers to worry that your hard-written article is AI slop - just don't run it through the slopifier. Or at the very least, spend 5 minutes tweaking the output.

If you can't be assed to do that, then it's very likely that you don't have valuable insights to share.


Because they aren't writing, it's vibe blogging or whatever

Because it’s a lead generation machine

I started working on a simple Telegram bot with llm backend and .md knowledgebase that would help me organize, track and research my long-term hobby projects.

The problem that I am repeatedly facing is that I am trying to build a home server and I keep asking chatgpt questions, but it is hard to keep all the little details in one place. The way I see it is that I can just text my assistant bot and ask it something like "hey, can you research which NAS setup would be the best for me given x and y". It will offer some setup and I would say "can you add it to the plan" and "can you plan the next steps for me?". The bot will also update the knowledgebase and version control it.

You might also want to use it for something like planning a trip to Paris, where at some point you might say "hey, given my schedule, can you squeeze a tour to top5 croissant places in the center of Paris".

The whole thing sound really vague and sounds like something solved long ago, but I cannot find solutions that will be guaranteed to stick to a very precise plan that I can review at any given moment. If you happened to know existing solutions, please let me know. I really don't want to build this thing.


I ran into the same problem. Tried using ChatGPT as a scrum manager. It was really good at checking off todos through conversation, but as I fleshed out the details of the project it would forget critical information - not necessarily todos, but critical information constraining the implementation of the project. In the initial stages, it was great, but it lost usefulness over time.


By ChatGPT you mean web interface of chatgpt? You were copypasting the "context" every time?


Great article!

On a sidenote, what is this new style of writing using small sentences where each sentence is supposed to be a punchline?

"And most of those sequences? They don't fold into anything useful. They're junk. They aggregate into clumps. They get degraded by cellular quality control. Only a TINY fraction of possible sequences fold into stable, functional proteins."


> what is this new style of writing

Congratulation, you are now able to recognize an AI-generated text.

(As of December 2025 at least, who knows what they will look like next month.)


I don't mind the style but the factual errors are not good. Like "How NVIDIA..." when it was done by DeepMind with TPUs.


Sounds like TEDspeak, only in writing.


Short sentence are good. Especially when you interact with low attention individuals. Make sure they stay engaged. It's not just a style. It's a game changer for your blog.


Any information on how comfortable the strap is? I am wearing a Garmin HRM Pro for one hour a day during workouts and it is not very comfy. I know a lot of athletes are moving to way less precise optical hand straps just because of the comfort issues with chest straps. I would not wear a chest strap for longer periods of time, unless I absolutely had to.


sadly, comfort for chest straps compared to hand straps is a known issue and ours is definitely no different. Wve done a bunch of tests, tried different materials/custom solutions, and honestly we're still clueless how to make it significantly better (if anyone here works in textiles or wearable fabrics, I'd love to connect). So yeah, if wearing your Garmin for more than an hour already feels uncomfortable, ours probably won't be much better in that regard


Great idea! Great website! Terrible video. The 90 second format is great, this is how much I would like to spend learning what exactly your product does. But the whole video is just clicking some user interfaces with no result. After watching the video, I have even less idea of what it the product is for. I would love to see a video that goes through the "next, next, next" in the wizard and then shows the actual outcome.


Great feedback, I'll work on the video ASAP. I intended to immediately create a follow-up video that steps through each component of a newly created decision, got distracted, never circled back.


And Dropbox is just an FTP server with SVN.


and moisturize


Total Commander is still the first thing I install on every fresh Windows install for the last 20 or so years. Copy/move/delete etc keys are the same as in mc.


They are the same because both projects are inspired by Norton Commander for DOS which also used those keys.


The best are the youtube movies that try to explain to amateurs how to use the horrible built in windows file manager WITH A MOUSE!.... while there is total commander.


Windows Explorer? It's not that bad. Granted, I only ever use it with the keyboard. But then you can tile two windows side-by-side (or four) and it's a bit like a commander of sorts.

The only time I ever have to use the mouse in WE is when I try to move to the Quick Access side-bar: sometimes focus gets lost somewhere around all the bloody menus I never use and I can't place it where I want it. It's weird and I'm not describing it clearly because it happens only occasionally and I don't understand exactly how.


This still implies that the person who is currently paying freelance programmers is 1) good with LLMs 2) knows some html and js 3) can deploy the updated website.


You're probably right that these people still need some baseline technical skills currently, but I'm really not assuming anything here – this is something we've seen multiple of our clients do in recent months.

It's funny you say they need to be able to deploy the update to be honest because we had a client just last week email a collect of code snippets to us which they created with the help of AI.

This is the problem we have though because we're not just building simple websites which we can hand clients FTP creds for. The best we can do is advise them to learn Git and raise a PR which we can review and deploy ourselves.


Sounds liked programming with extra steps. And I don't like it when the extra steps involving mailing snippets of code


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: