How fucked are humans now that AI can easily make us interpret things subconsciously?
Not only through images like pic related, but words, anagrams, arrangement of words
I think humans are very programmable, and can be easily triggered
Now with AI any fucker can make a tweet that makes you do whatever
And all you have to do is read some words, or see an image, or a combination of both
Ok maybe in pic related case it's not programming anything, just a way to cleverly pass a message
But you get the point, or is there a point?
humans are pretty fucked, yea. the future will be very noisy and very few will know what the fuck is going on
as an aside to
most people get very angry when they learn a party has cut off another partys water, plenty suddenly know what's going on when they hear that whether in the minority or globally, they just pretend not to.
*whether locally or globally
technology is our downfall
quite a while ago, before open source image or language but at the time of open source teleconferencing (re-orienting a face as if it was looking at the camera straight-on; changing hair color, shading etc basically filters), the us presidency made a video address so flagrantly ai-generated that even the open source of the time could not have messed up that badly: the bone structure of the face was morphing through the video.
when i showed this to others, they thought it was creepy.
when i showed it to that one guy obsessed with politics, he couldn't distinguish it as artificial in origin.
i think a few extremely volatile people are easily influenced, but that has always been true. now the artificial gibberish, again of even lower quality than stuff old when sd1.4 released, is used to justify genociding pallies. however, how does this work in practice? everyone thinks it is wrong, then that one guy throws a complete toddler temper tantrum meltdown, and a few days later everyone moves from "israelites bad" to "both sides are bad".
why? this weird guy posts everything on social media. someone said, "i can't say some things explocitly; it will follow my career!" basically everyone follows the loudest, most obnoxious ingrates because otherwise they are probably screwed (sometimes), the ingrate just serves as a reminder that one way or another your inner morality will be judged against a satanic standard.
tldr; your image looks like it has a hooked nose and is rubbing its hands together, but people only care about whoever is the loudest and not offensive them so that everyone can get along. i don't think that is uniquely exploitable by ai. i have yet to meet people i've known for years who actually believe the globohomo stuff, they just don't want to trick the boat.
t. someone who has to listen to "kill the pallies israelites number one" literally every ducking day oh god it's so annoying but i can't just shut people i know out of my life
Be careful anons, there are three separate artificial engrams embedded in this post. I would suggest asking Claude or preferably a local AI to summarize it rather than reading it directly yourself.
Yeah it's a weird post
explain, are people using LLaVA to make bot posts now
The same way that OP's image has the happy merchant embedded in it, that post has several engrams embedded in the text. It's basically rewritten by an AI to present you with an idea directly as you parse the text, bypassing the normal steps of reasoning and comprehension. It's not really "mind control," but it's a sneaky way to trigger readers into certain patterns of thought. Asking another AI to rewrite it will eliminate the embedding, but occasionally you'll get weird responses where the AI interprets the embedded engram and attempts to rewrite it in words as well.
is it like how when you feed gpt to dalle to got and so on, you eventually get moire patternstexturing a gaussian height map? (cf twitter unfortunately)
can you tell me what one of the engrams is? i've stuff this for images for a while but the language stuff is pretty new.
No, it's more like the controlnet stuff people are using to make the other images in this thread. I can't help you with the specifics of the engrams, I stopped reading when I noticed the weird structure and fed it through a language model to identify the embeddings and rewrite the post without them. I'm not sure that even if I were to read it directly that I'd actually be able to put them into words; we're dealing with crystallized knowledge here, which completely bypasses language. Obviously there's a way to represent the engram with language since there was presumably a written prompt in the first place, but I probably wouldn't be able to work backwards to figure out what it was, the same way that I don't think I could guess what prompt might result in the image you posted. Maybe something to do with ancient Persian architecture? LLaVA and CogVLM have been getting a of press recently so it's possible that someone adapts it to make a similar "reverse prompt" AI for these text embedded engrams, but I wouldn't expect it to be publicized much. The technique is flying pretty low under the radar since no one legitimate wants to touch it since it's basically the exact thing that all of the "muh alignment" Yuddites are screeching about, and everyone else gatekeeps their secret tech even harder than /aicg/ does their GPT-4 keys.
thank you for answering me perhaps earnestly. in fact i discussed myself as a naive onlooker to lure you into a false sense of security. i mean nothing ill, but to steer you in the right path! in fact my posts were not ai-generated but were a literal stream of consciousness of my literal life. the embeddings you think you see are simply the consequence of not discussing with people like me, which is normal. the engram concept is silly but there are things which are real but related. you may be very delighted to investigate an obscure old article on the topic of "seeCoder" to find out more, essentially an instant-Neural Architecture Search thing, but the paper didn't say any of that. good luck and have fun, apparently there is a comfyui for language models these days so you might also enjoy investing what you can create with that.
i really do hope this interface chat between us serves as evidence to my claim that most people are not NPCs and are quite intelligent, but do just try to earnestly get along. the real engrams, so to speak, were in my manipulative steering of our question and answer session.
the image is a painting derived from one of only a very small number of surviving bronze age anatolian frescos apparently depicting mycenean presence in the region.
*disguised, not discussed, myself
Are you referring to the Biden NBC(?) interview where his ear shifts vertically, or has this sort of thing happened multiple times?
the one i saw had the ocular structure on the left eye going everywhere pan and scale wise, not an interview but some formal address
there are probably a few more, like the one your mentioning
i suspect people intentionally do this to make it super obvious something strange is ocurring
Truly a powerful technique
I envision museums centuries from now will have these
And people will watch but not actually understand the meanings
>more chaos for jannies and corporats to clean
good
Is there a guide to making top kino images like this?
This reads like one of those stupid chain mail ghost stories, albeit with a veil of sounding vaguely scientific. A quick search around the internet seems to confirm that it's just that, a story. The "ah but of course this one has a delay" thing is just plain cheesy.
not very, humans are like AI and AI isn't easily purposed without human interpretation. If you're worried about anything you should be worried about how we educate the youth, analogous to training AI.
>worried about how we educate the youth, analogous to training AI.
This is not how you train AI
well that's not how you're supposed to 🙂
it's not how you train AI
retard
computer illiterate. get the fuck out of here lol.
I'm not worried about AI, I'm worried what companies, and their marketing departments will do with it. Subliminal advertisements everywhere, like ads being pushed now?
Reading is kinda pozzed
What? I don't don't see anything unusual in this picture anon
what about this one
Nothingness
wtf is this picture, my screen flips out when i scroll with the image at full size
because we are bio machines. we have bunch of neural networks controlled by survivability scripts . ai is part of us
What do you use for these things? img2img?
For this one yes
The one in op is likely qr controlnet
qr? what is qr?
https://civitai.com/models/111006/qr-code-monster
That shit’s crazy. I can only see it in the images you posted when it’s in the smaller thumbnail size, but if I open them I can no longer see it.
Controlnet is the actual technology that guides the image gen, qr codes are just square barcodes that link a url, even boomers know that
Fucking retard.
I have been making images like the one you posted since before controlnet was even released, claiming these are all made with a qr model is admitting you're a newfag
Doesn't do text well and preserves colors which is not always desirable. In the first reply I did mention img2img though, learn to read
hello fellow picture enjoyers
Is there a psychological name for this? It is quite scary when you think about it. I suppose it’s no different than ancient symbolism used today such as the square and the compass.
there are le secret messages in my son's music when i play le vinyl backwards the demons will take him
Well, yeah. They can state a different thing in official lyrics on genius.com while speaking to your brain something similar but totally different
they can't even ban this image from anywhere
based
fuck merchants
Are you all feeling it now?
t. /x/pat
Subconscious manipulation as you describe it isn't a real thing, it doesn't work.
>source just trust me bro
Is there any Web i can do this online?
Can you put an example of the prompt?
sd controlnet
>he just discovered pareidolia
A human brain is a convergent system and what it converges to is food and pussy and some other things that lead to food or pussy.
You actually can't program a human at all, if you want it to do something you invariably have to appeal to its self-interest, otherwise it will just ignore you.