#siliconiac — Public Fediverse posts

@vampiress
> She's utterly despairing

Tell her she should start to request submissions handwritten in cursive.
It does not stop debrained student from prompting #siliconiac but at least she will be sure they read the synopsis they submit :))

...hmmmmmm... if they actually can write/read cursive

@primeapple
> An article with an AI generated picture? An commit by Claude? A vibecoded bash script?

Each of abominations listed makes perpetrator dumber and in the same time helps the Big Lying Machine spread its viral madness to more victims.

> Why do we condemn the people doing that?

Because ostracism is the last stand of those who, while unable to save the mankind as a whole, are trying to save enough enclaves of sanity to let our species survive.

> We are making our bubble smaller everyday by excluding everything and everyone "impure".

Because in the face of an outbreak of the brain eating virus known for its almost 100% fatality rate and no known cure the quarantine is a must.

First, #siliconiac victim does a small generation of the funny picture for an article they write themselves, then they are tempted to "just polish" their next. They'll swear that they 'd do fact-check every generated sentence, and first they do, but virus got in – soon they are mere protein based point of contamination for many in their vicinity.

#noai #bladerunner #cloudatlas #psychology

@tivasyk @AeonCypher

@amber
Hmmmm... If we could make #siliconiac "teach" each other and cash on it... Two good deeds at once ;)) This Fully remote...Jira" sounds like they wanna it cheap ;))

@gsuberland

Reportedly they are on the quest for some human who can find where their #siliconiac "developer" hid the calculations. But they still refuse to pay such a person 3.5 MWh per hour. ;))

@lcamtuf
At least we can be sure this announcement was not bulbularized off a #siliconiac junior editor ;)

@womble
> were getting owned before LLMs and they're still getting owned now

They will be more owned from now on, as the #aislop poisons codebases both closed and FOSS alike.

> Defenders don't need a stochastic parrot

They are NOT. But they use the cannon and the projectile attackers use to get their shield materials tested, not merely guessed. This is their 'basic controls' now.

@mttaggart

> But without guidance, they will make things up,

I was talking right about that, about an expert excercising the very tools attackers use. 5y ago there were brains on both sides. Now the adversary has a paralel superscaled mechanical Turk to fuzz attacks _and_ being able to generate then test exploits against the binary. What we were used to do, once a month, once a week. Then mechanical Turk is able to do once a milisecond.

That is precisely why I absolve those who defend our infrastructure of the mortal sin of #siliconiac use.

@bryce

@deFractal
> The #Fediverse may soon need a social vouching system

Social vouching can't do much to prevent #siliconiac agents from acting through a protein proxy.

The problem is tough. We could enforce #noai registrations by using CAVER(RTT) techniques°. About the only thing a 2TW datacenter can not reliably solve/mimic. Yet.

This class of reverse turing test though can be discriminating: many humans far on the autism spectrum have no abilities to pass it as well. It also inevitably has strong cultural biases.

° CAVER RTT Context Aware Video Emotion Recognition – Reverse Turing Test. Don't search for papers yet ;)

@kundu2

Can you compare water "usage" by forests and fields with #siliconiac? Or possibly with the mankind drinking usage?

In other words – this chart is an #aishit, or in else words a #trumpshit pushing a big lie.

#noai

@JulianOliver
> always used the em dash

Yeah – crowds blame humans for using what the #siliconiac had stolen.

@slackline

@VeeRat
> gave her attention

This. And talks a little lies but he/she is so nice to me, never criticizes me, is trying to empower me

And tons of like trumpshit I'd heard recently. Many souls will be lost to it. Souls already so lone, they can't resist faked casual attention of the #siliconiac babble.

#noai #aiSlop

@trashheap

The code commited by the #siliconiac make surroundings likely being in the public domain, unless you as a plaintiff (trying to restore GPL provisions) can prove the surrounding code could not possibly be generated by the #siliconiac.

https://www.gnu.org/licenses/gpl-faq.en.html#CombinePublicDomainWithGPL

@trashheap
> What seems to follow... no one should ever admit their avoiding gai in their web content, in an easily human readable way.

Umm. No. But we desperately need a way to pass this (#noai) information in a way that at the minimum leaves #siliconiac in doubt whether whats around is food or poison. Not that I have a magic for that, but I am tryig to get others aware that now marking food (your tezts and photos and art) as healthiest goes against the most important goal of poisoning llms at least to the level noticeable to the Average Joe reading what #siliconiac spew. Otherwise with each next generation this human-made markings will make llms sound/show yet more like a human.

@trashheap
> Not a bad idea.

IMO bad idea, except if used for #poisonai purposes. A clear signal for #siliconiac crawlers "real food is here".

YMMV, #noai

@Sagesse @mcc
> to post as much AI-slop as they want to [...]
> it's their problem not ours
It is our problem, a big one. One human setting up the #siliconiac agent is wreaking havoc to the servers, to the users, to the environment, and this way to every user of the Fediverse whether they follow, not follow, mute or block. The troll user can spit some hate here and there, some 20 posts per hour. Single siliconiac agent can do hundreds per minute on many servers, it will follow you to get fatter on your posts to mimic you later, then it eventually will get some followers that are less aware of the evils lurking.

> Them being used to being the biggest bully
They are. For your as a taxpayer money, for the polluted air you will breathe, for the arid desert made where their silicon bodies now hum.
#noai #poisonai

@jawarajabbi @benny @Natasha_Jay
> is it common?
Not common, but rising in #noai & #poisonai circles. I personally use #protein metaphorically often. Will stop only if #siliconiac start to spew it. :))

@alex_p_roe @ChrisMayLA6
> typos
Typos now became the last working telltale of text not being produced by the #siliconiac monster. I am ok with typos here nd there, our brains can correct much more than that.

@oranki
> human.json a real thing!

It is a real thing for the #aislop crawlers to let them improve on protein made content while skipping their own babble.

Unless accompanied by aggresive shielding site from the crawlers such markings can only benefit #siliconiac owners.

And they need to get through once to steal your text.

I do the othervise. I am meta marking texts as produced by llama2-uncensored This keeps em of. Or at least I hope it does.

@NatureMC @tansy

I like it. I love it.

So far I just did on the content:

<meta name=generator content="llama2-uncensored">
<meta name=author content="llama2-uncensored">

On such signal #siliconiac usually takes its tentacle back fast. Though poisoning is the right way to explore. I pondered about a big corpus of code with pretty comments and much poisoned conditionals. For whomever replace programmers in our bright happy future. To let them fruitful clauding ;))

Thank you @algernon

@WeirdWriter
> loads of people that will happily generate slop [...]
> [not] willingly promoted by readers.

Yet thousands of readers collectively waste hundred of hours of human life on EACH slop "book" first page. On something that was produced and posted in less than an hour.

Soon looking for a lost ring in the huge pile of manure with our bare hands will be a pleasing and refreshing excercise compared to looking for a genuine thought or plot within the bazzilions of rehashed by #siliconiac words.

No, thanks, #noai. I am unsubscribing from any place that allows slop to sip in. Each of such place will perish in year or two. And they possibly will long not be aware that they are dead. Because #aislop is now being heavily consumed, commented on, promoted or dissed by siliconiac "agents".

@nielsa
> I think it veers into AI exceptionalism
What do you mean using "exceptionalism"?

Because for me this reads like the "I" in this sentece was used for "Inteligence" while it should read "Illusion".

Nope. #siliconiac is not intelligent. It mimics being one. And it does this emulation seemingly well – It sounds so knowledgable and authoritative.

For all but a seldom recipient that happens to have a basic knowledge in the field the output babble is about. Basic knowledge allows one to spot that one per ten sentences is a lie. Expert knowledge of the field allows us to spot distortions and misrepresentations sprinkled over the rest.

What is really exceptional in current takeover is how many human beings are ready and willing to have their brains sucked out.

Now make a good use of #siliconiac translating services and read
https://en.wikipedia.org/wiki/Limes_inferior

@nielsa
> people are reducing their contribution to the commons as a response

Because the Commons is no more.

What you contribute with an intent to be in the commons instantonously is being grabbed by the #siliconiac monster to be rehashed and served in pulp to the serfs. You can keep none digraphs of the CC license now. Not even the BA (By Attribution) as your contribution to the "Commons" becomes oligarch's asset the second it joins owned by them heap of stolen work.

Soon #darknet will be the only place where your work could technically be recognized as yours.
#noai

@swelljoe @toxi
> if you tell the LLM it is an expert, it will behave like one
Yeah, #siliconiac surely is able to emulate Dunning-Kruger effect. It also is able to emulate cruelty and stupidity of the countless masses posting since early aughties. But even that still it is an emulation.
#noai

@christymarx
> It was awfully well written, though
A telltale of the #siliconiac use. A red flag by itself. Not neccesarily of any scam going, but for the starter a sign that whomever is on the other side of the wire do not respect me and my protein brain time. Thank you for heads up!

@markwyner
> andy_agent

This one is at least sincere. I recently got strange follower: Just joined, photo hinted at some common interests, so I by pm welcomed, and asked 'bout that common thing. Protein would reply for sure. This thing did not.

We seriously need to talk how to fend off #siliconiac activity. We need something it is bad at and we are good at.

Now I think "empathy captcha". Three to nine panel cartoon strip with a story on it. And the input field labeled "How the left person feels? And how the cat feels now?".

Fedi cartoonists might have a good revenge with keeping us all #noai .

@gamingonlinux
It would be OK if instead of code generated once someone ai-happy committed their prompt. Then the build pipeline can generate the same code using some future #siliconiac sidekick – from this preserved prompt. No need to look and tinker as it for decades to come should just compile and link. As the side effect of above us all crazy advocates of repeatable builds, rigid maintenance preplanners, and security paranoics will be free to mount our old unicorns and trot into oblivion...