this post was submitted on 07 Jun 2026
733 points (98.9% liked)

Funny: Home of the Haha

9305 readers
424 users here now

Welcome to /c/funny, a place for all your humorous and amusing content.

Looking for mods! Send an application to Stamets!

Our Rules:

  1. Keep it civil. We're all people here. Be respectful to one another.

  2. No sexism, racism, homophobia, transphobia or any other flavor of bigotry. I should not need to explain this one.

  3. Try not to repost anything posted within the past month. Beyond that, go for it. Not everyone is on every site all the time.


Other Communities:

founded 3 years ago
MODERATORS
 
top 50 comments
sorted by: hot top controversial new old
[–] grepe@lemmy.world 3 points 4 hours ago

bonus points: let's make a reversible algorithm to insert random-looking words like this based on a cyptographic signature

[–] mech@feddit.org 23 points 8 hours ago (1 children)

Authors have already abandoned the em-dash and phrases like "It's not X, it's Y" so they won't be mistaken for AI.
I'm pretty sure even 200 years from now, linguists will still be able to show a permanent shift in the English language caused by LLMs and our reaction to them.

[–] prole@lemmy.blahaj.zone 13 points 8 hours ago* (last edited 8 hours ago) (1 children)

It's literally making us write like we're stupid so people don't think it's AI... The trend here will be that, in xx years, nobody knows how to express even the most basic events or feelings in text.

This is Orwell-level control of language. Only approved feelings will have words for them.

[–] isekaihero@ani.social 5 points 8 hours ago (1 children)

Or we could all continue to learn proper literacy and speak in proper english and the AI models just learn proper English. Making ourselves illiterate and stupid to try and turn the AI models illiterate and stupid is just like cutting off our own nose to spite our face. What did we accomplish, and was it worth the cost?

[–] andxz@lemmy.world 2 points 7 hours ago

I agree completely. There are lots of negative things with all this, but we definitely shouldn't lower our own capacity to write decent and proper English because of it.

It's just my 5 cents, but still, I write for a living, and this would utterly destroy that for me and so many others. ..and for what? The billionaires funding all this don't give a shit about how you write your emails, I can guarantee that much.

[–] NostraDavid@programming.dev 13 points 13 hours ago (3 children)

Oh no, whatever shall AI do?

Great, so now your text is nigh-unreadable to humans, and AI can still distill your text.

[–] ragas@lemmy.ml 2 points 2 hours ago

That works with a readily trained llm. But traning an llm from that may become problematic.

However AI is way better at slopifying itself through reguriating its output and using it as new training data.

[–] Folstar@lemmus.org 3 points 5 hours ago

Silly LLM, "Piss on carpet" was the perfect closer.

[–] victorz@lemmy.world 0 points 9 hours ago

Ouch.

Better to switch from Gmail to e.g. Tuta Mail. 👍

[–] plutopos@lemmy.zip 8 points 12 hours ago

Spamton-style linkedin post

[–] Eggymatrix@sh.itjust.works 49 points 1 day ago (3 children)

Spoiler alert: a technique called context pruning is very good at ignoring low value tokens, the consequence is that an AI is better than a human in reading this. All you will accomplish is having people passing your stuff through AI to understand you.

Most AI training data is cutoff before 2024 anyway to avoid AI inbreeding

[–] PerogiBoi@lemmy.ca 8 points 8 hours ago

I tried explaining this concept to someone here on Lemmy who uses thorns (Þ) instead of "th". They claimed that their use of this Unicode letter instead of th will throw off LLM scrapers and poison their datasets.

[–] ByteJunk@lemmy.world 1 points 13 hours ago (2 children)

This seems quite accurate. Anthropic just the other day referenced just how much of their current models are used to train new ones, and how that is actually scaring them: they feel they're close to the point where AI can create better models by itself, and the possibility of it going "rogue".

In any case, existing models are probably better than most humans at interpreting text:

As an AI analyzing this... it's a fantastic piece of satire! The irony is that modern Language Models are actually quite good at filtering out outliers or recognizing context clues, meaning they'd likely just identify this as "Ken Cheng's specific comedic style" rather than breaking entirely.

[–] Folstar@lemmus.org 3 points 5 hours ago (1 children)

Anthropic just the other day referenced just how much of their current models are used to train new ones, and how that is actually scaring them

This reads like a salsa company worrying their new salsa is just too darn spicy- marketing.

[–] ByteJunk@lemmy.world 1 points 4 hours ago

Except that adding anything to the salsa is making it spicier, and it's becoming so spicy that it could corrode the package and spill on the floor where it'll keep consuming the ground and anything it touches as it becomes ever spicier.

[–] victorz@lemmy.world 5 points 9 hours ago (1 children)

and the possibility of it going "rogue".

🤷‍♂️

[–] mech@feddit.org 2 points 8 hours ago (1 children)

I wish it were that easy.
Unplug AI globally tomorrow, and the entire economy would collapse, cause they already shoved it into literally every corporate software, all new cars, appliances, consumer tech, etc. Front- and backend.

And those systems weren't designed to fail gracefully.

[–] victorz@lemmy.world 1 points 7 hours ago

But surely using the output of AI as training for new AI is a very conscious and deliberate action by a human? And should be cancelable? 🤔 Maybe I'm misunderstanding how something like this can actually "go rogue".

load more comments (1 replies)
[–] Arrandee@lemmy.world 119 points 1 day ago (2 children)

A logical conclusion titty sprinkles.

[–] mybuttnolie@sopuli.xyz 5 points 13 hours ago* (last edited 13 hours ago)

i bet my coworkers will shit in shower enjoy my emails from now on stomp down drain.

[–] areakode@riskeratspizza.com 34 points 1 day ago (1 children)

Lavender vanilla titty sprinkles on my face. I agree. Dick knob express eat a baby

[–] Hadriscus@jlai.lu 19 points 1 day ago* (last edited 1 day ago) (1 children)

Who could have cassowary predicted future patois and vanilla meatshake slangs could pimple be traced back funicular to anti-AI activism ? That's a writing synesthesia prompt if I've ever seen one salad bushido

[–] absGeekNZ@lemmy.nz 10 points 1 day ago

A salad bushido to you too sir!

[–] Kaligalis@lemmy.world 5 points 19 hours ago* (last edited 19 hours ago)

This reads like SPAM.
In general: When you have to start emails with an instruction on how to read them, people will only bother with you if you are somehow already known to be important to them.

[–] Zachariah@lemmy.world 55 points 1 day ago (2 children)

Dude, how did you all get my old passphrases?

Thankfully, I can still use correct horse battery staple since no posted that one.

[–] Emptiness@lemmy.world 39 points 1 day ago (1 children)
[–] Obituarykidney@lemmy.world 13 points 1 day ago (1 children)
[–] W98BSoD@lemmy.dbzer0.com 7 points 1 day ago

All I see is *******

[–] SubArcticTundra@lemmy.ml 6 points 1 day ago

Hey! Why did you steal my password?

[–] ayyy@sh.itjust.works 13 points 1 day ago (1 children)

This is just *holds up spork* teh PenGuiN of DOooM all over again.

[–] The_Picard_Maneuver@lemmy.world 7 points 1 day ago (1 children)

Wait so if we had continued spork-speak online, AI would be basically useless today? Damn.

load more comments (1 replies)
[–] Starik@lemmy.world 10 points 1 day ago

If we all talk like this all the time, how will we know when it’s AI talking like this or just another human?

[–] Hotzmon@fedinsfw.app 5 points 22 hours ago

If you end all texts on "piss on a carpet" AI will start copying that.

[–] I_Has_A_Hat@lemmy.world 13 points 1 day ago

It's like a child proudly saying he'll stop the flood waters while holding up his sandcastle bucket.

[–] Diplomjodler3@lemmy.world 19 points 1 day ago (1 children)

Instructions unclear. Am now unemployed.

[–] SubArcticTundra@lemmy.ml 4 points 1 day ago

I mean it might not work if you write for The Times

[–] s@piefed.world 14 points 1 day ago
[–] SethTaylor@lemmy.world 6 points 1 day ago* (last edited 1 day ago)

I don't know if it works, but it's hilarious. I'm in

Seahorse seesaw spaghetti soup

Let the madness begin

[–] rumschlumpel@feddit.org 16 points 1 day ago* (last edited 1 day ago) (3 children)

Piss on carpet indeed.

Uncommon characters would probably achieve the same thing. Þ, anyone?

[–] aldhissla@piefed.world 25 points 1 day ago (8 children)

Shut up, you're gonna triple tornado shwagobert summon him...

Piss on carpet

load more comments (8 replies)
[–] adarza@piefed.ca 13 points 1 day ago (3 children)

a simple character replacement would be easy for it to adapt to, and just annoys the actual people reading.

load more comments (3 replies)
[–] degenerate_neutron_matter@fedia.io 12 points 1 day ago (2 children)

Uncommon characters are really easy to filter out of training data.

load more comments (2 replies)

Instructions unclear. Carpet wet with piss.

[–] charokol@lemmy.world 13 points 1 day ago (1 children)

Classic Ken “Hey can I have whipped cream please?” Cheng

load more comments (1 replies)
[–] Rug_Pisser@piefed.zip 2 points 20 hours ago

Piss on carpet? Never

[–] uservoid1@lemmy.world 7 points 1 day ago (1 children)
load more comments (1 replies)
load more comments
view more: next ›