Hacker News Reader

Comment by coffeecoders | Hacker News Reader

I actually think we’re overestimating how much of "losing our voice" is caused by LLMs. Even before LLMs, we were doing the same tweet-sized takes, the same medium-style blog posts and the same corporate tone.

Ironically, LLMs might end up forcing us back toward more distinct voices because sameness has become the default background.

View on HN

My theory is that LLMs are accelerating [online] radicalization by commoditizing bland, HR-approved opinions. If you want to sound like a human on the internet, for better or for worse the easiest way is to say something that would make Anthropic’s safety team have a heart attack.

I mean there's still Grok... surely that gives may safety teams heartburn.

But I find this take interesting. The brewing of a new kind of counter culture that forces humans to express themselves creatively. Hopefully it doesn't get too radical.

Grok still has that annoying tone, it just uses it to say weird things.

Grok sounds like the "how do you do fellow kids" meme. It just sounds cringe.

So like a 54 year old rich white man trying to post cool edgy takes?

> commoditizing bland, HR-approved opinions. If you want to sound like a human on the internet, for better or for worse the easiest way is to say something that would make Anthropic’s safety team have a heart attack.

I agree.

LLMs are like blackface for dumbfucks: LLMs let the profoundly retarded put on the makeup and airs of the literati so they can parade around self-identifying as if they have a clue.

If you don't like the barbs in this kind of writing prepare for more anodyne corporate slop. Every downvote signals to the algorithm that you prefer mediocrity.

"Every downvote signals to the algorithm that you prefer mediocrity."

Not on hn it doesn't.

I'm all about using the rainbow of language to it's full breadth, but if you're going to go for shock jock you should maybe have something better to say than "hurdurr I smart and world bad 'cause dumb people". Makes you sound a little like you're trying too hard to be part of the 'literati', whatever that's supposed to mean.

Bullshit. You can just use ANY abliterated model or non-safety aligned model. People have popularized ‘hard r or you’re hardware’ for a while now, but any self hosted LLM that’s de-aligned will gladly say whatever.

People believe this and continue to get fooled by LLMs all day.

I don't think I would want to hang out in any online space where the n-word is a shibboleth. The practice of asking members to do something that would ostracize you from the rest of society is cult behavior—even if it's not intended as such.

Censoring the n-word is an even worse shibboleth.

https://iowacapitaldispatch.com/2025/11/13/school-fired-work...

I agree, and it’s also a stupid measure that doesn’t stop anything.

I've never heard of this, and I'm not seeing much on Google.

What spaces are you in that you see this??

Twitter

> If you want to sound like a human on the internet, for better or for worse the easiest way is to say something that would make Anthropic’s safety team have a heart attack.

That's always been the somethingawful crowd's stance since, what, 2000ish?

Also ironic is how the post about having a unique voice is written in one-sentence-paragraph LinkedIn clickbait style.

You're absolutely right! Tell me more about how ironic is how the post about having a unique voice is written in one-sentence-paragraph LinkedIn clickbait style.

The other day I interviewed a candidate who had lost their unique voice...

Spoken like an Ai bot.

woosh

You're absolutely right that I'm absolutely right! So how may I serve you today?

Content recycling has become so cheap, effort-wise, it’s killed the business. Thank god.

Yes. That particular content-farm business model (rewrite 10 articles -> add SEO slop -> profit) is effectively dead now that the marginal cost is zero.

I’m not mourning it.

That particular content-farm never interacted with me.

Tangentially, we've reached to the point that we can't trust google/internet anymore.

It doesn't it just makes it cheaper by not requiring human effort.

Yes, fully agreed. Most people producing content were always doing it to get quick clicks and engagement. People always had to filter things anyhow and you had to choose where you get your content from.

People were posting Medium posts rewriting someone else's content, wrongly, etc.

You are right. An observation, even before the rise of LLMs, was that computers were becoming more human-like and humans were acting more like machines

The sculpting force of algorithms is bite sized zingers, hot takes, ragebait, and playing to the analytics

Bingo. Your message is only what the medium promotes.

The greatest thing "AI slop" achieved was highlighting how much "human slop" content there was.

I mean, if you typed something by your own hand it is in your voice. The fact that everyone tried to EMULATE the same corporate tone does not at all remove peoples individual ways of communicating.

I’m not sure I agree with this sentiment. You can type something "by hand" and still have almost no voice in it if the incentives push you to flatten it out.

A lot of us spent years optimizing for clarity, SEO, professionalism etc. But that did shape how we wrote, maybe even more than our natural cadence. The result wasn’t voice, it was everyone converging on the safe and optimized template.

If you chose to trade your soul to 'incentives', and replace incisive thought with bland SEO and professionalism -- you chose this. Your voice has become the bland language of business.

So in that case, would someone willing to publish LLM-speak under their name be similarly adopting that "voice".

Does that entail that LLMs are not in fact erasing our societal voices, only making it easier to adopt bland-corporate en-mass?

That's a reasonable interpretation. People are choosing to mute their voices, and replace their identity with ChatGPT.

It's not a passive loss of voice. Their voice didn't fall off and slip between the couch cushions.

Reply on news.ycombinator.com