alt Hacker NewsAI-generated videos to maximally drive a target brain region
Comments
I’ll leave it to other comments to discuss the societal and moral implications of being able to do this (which, I agree, ick…). On a practical level:
We train an encoding model, a “digital twin”, that predicts how each visual region responds to any video. Now we can ask: which video would make a chosen region light up the most? NEvo searches for that video automatically, using the twin’s prediction as its reward.
I only scanned the paper, so maybe I missed it, but is there any confirmation that this ‘digital twin’ works? Like, do the generated videos actually cause the same patterns as in the ‘digital twin’ brain model in real humans in an MRI machine? My instinct is to be skeptical that it’s possible to reliably create a video -> brain activation prediction model.
This is very similar to last week with that mind reading startup thing. Please read the paper before commenting.
This is a tool to help researchers in figuring out what different parts of the brain are actually for with less experimenter bias contamination of “well we think maybe it’s about this so let’s show it video of x to see”.
The essence runs on having someone sit in a scanner for a couple hours watching all sorts of things, and then feeding that to a model that will then build its own representation of said data and try different things on it until it’s found what makes a certain part sing in the model.
The purpose is a generalized understanding of brain function, more or less the same way we’ve been doing it all these years. Expose brain to something, record it somehow, see if brains reaction in the recording helps you understand more about who we are and what cognition is.
An anecdote, but some reason to think that overworking part of your brain is a bad idea:
I had an Aunt who had dementia. This is obviously a terrible outlook. But her and my Uncle seemed to be doing okay. My Uncle was an ultra-competent guy, highly stable, the person you could rely on; and had been his whole adult life. So it was shocking when he had a mental breakdown and became manic. There was probably something physical going on, but what was also going on was that he had wanted my aunt to be able to live normally for as long as possible, so had been covering everything - for a year or more he'd had to be alert 7 days a week in case my aunt tried to cook on the gas stove or something like that, at which she was no longer safe. So the risk-alert part of his brain had been constantly overworked.
I appreciate that this is scientific research, but there are definitely companies out there that will try to row-hammer everyone's brain if this sort of thing is not heavily controlled.
As others on Telegram have said: automated search for visual superstimuli likely leads to bad outcomes.
https://en.wikipedia.org/wiki/Supernormal_stimulus
https://en.wikipedia.org/wiki/BLIT_(short_story)
Also: one of the V3A animations reminds me loosely of things I saw when I was a kid, at night, shortly before I slept (though my experience then was more circular).
I’m skeptical.
I’m always skeptical of these papers that claim to read minds, because my understanding is current brain reading technology is very coarse (Neuralink is SOTA, very invasive, and only 1-3 thousand electrodes for 55-70 billion of neurons; fMRI is much worse) and neural networks generate plausible images and videos from noise. Indeed, the example videos don’t seem particularly stimulating, just random.
I can imagine truly horrifying images coming out of this. Many people who have spent a lot of time online have come across unfortunate gore images that they may have seen for only a fraction of a second that they wish they could unsee. One can imagine the brain is able to see much worse.
Someone should use this to make the "Infinite Jest" film from the book "Infinite Jest"
Among my secret research ideas, this is the most dangerous and morally wrong one.
If future generations of researchers will wonder why IRB reviews became mandatory for computer science, studies like this will be the answer.
Seriously, some people don't seem to realize the point at which they are becoming Fritz Haber.
Aee they trying to create snow crash for real??? [1] https://scifi.stackexchange.com/questions/184341/in-the-nove...
What's interesting is that the process of identifying what gets human attention is nothing new.
e.g. cartoons from the 1950s were designed to have:
- cute animals
- short and fast scene cuts
- catchy music
b/c all three of the above were known to trick the human brain into "sit still and observe" mode.
Focus groups did the same thing for advertising which then morphed into A/B testing to determine revealed preferences.
Same thing with the whole fat/salt/sugar research that led to near addictive fast food.
This is just the next logical step.
This is already happening at scale by the social media feed algorithms. We don't need generated content to accomplish this. In a sea of user created content, plenty of it is already at peak activation.
Let's stop pretending that we can separate the good from the bad here. This industry and everyone in it is morally bankrupt.
The title reads a lot like the lab logs you'd find in a horror game.
But for the paper itself, it seems they're using genetic optimization over predefined keywords. Wonder what would happen if they did gradient descent on the latent space directly. Is brain stimulation just not a good domain for GD?
We are really getting to the point where the tech industry must be stopped if humanity is to continue at all, let alone thrive.
Ah great, work on the Torment Nexus is progressing well. I can head into the weekend with peace of mind.
I actually had a similar idea that I cam across in my own research in brain-to-video applications. You could theoretically train the model in real-time by using brain data as a reward signal, such that the model would find the image or video that most strongly produces the desired response (e.g. an increase in certain brainwave bands associated with pleasure, or alertness, or fear).
I can imagine this has some very undesirable usecases if used in the wrong way.
Reminds me of the parrot https://www.sfsfss.com/stories2/BLIT.htm
WOW! Cant wait to tell to future generations that we had voluntarily made these algorithms to manipulate and influence our own brains
Apart from ethically bad and evil use cases of this application, can we use it to massage the parts of brain like we do it to our bones and muscles with the help of physiotherapists?
reason I am asking it could be some relief to our brains after tedious working day, especially after heavy AI usage
I'm pretty sure by blindly following this path to commercial success in marketing; we're just gonna end up with content that is as implicitly adjacent to porn as possible without being explicit.
I can see lot of potential in this, but I somehow doubt it's going to be used for good.
Very interesting. We have an organic experiment converging to maximum stimulation in short form videos (which will become the majority of training data for future video gen models) Already approaching the capabilities of a “mood organ” from blade runner. Except usually most people don’t even make the choice to change their mood anymore. https://en.wikipedia.org/wiki/Do_Androids_Dream_of_Electric_...
What are these videos supposed to do? I watched few of them and it does nothing for me?
if it is targetting visual regions of brain and I have aphantasia (I cannot visualize anything in my mind) is that connected?
What's next, a movie you watch and die in 7 days?
For a video to _literally_ maximally drive a target brain region there would be fatal consequences - e.g. the plot of David Foster Wallace's Infinite Jest (1996).
I wonder if the end-game of this field of research will be to run these simulations at scale using neuron-on-a-chip services such as [0] Cortical Cloud.
I don't think it's a matter of if but when. Grim.
So, AI images to maximally drive reward centers in the brain?
What could possibly go wrong?
I wonder what Meta could do with a similar technology…
But here we can start also the usual discussion about technology research for the sake of it vs calibration of possible side effects of new research
Personally i think we haven’t solve this problem and thus it’s just a matter of time until we’ll get in a non-going-back point
A portion of young men already are more or less addicted to pornographic content.
Tolls like this have the potential to make this so much worse.
Absolutely horifying.
The whole site looks AI generated. Those 3D brains are... smooth.
The sci-fi hand wringing is embarrassing. A huge chunk of the population is afflicted with mental health issues and the best psychitrists can seem to do is "Oh bro, wanna try some ketamine? Anti-psychotics?? Maybe some SSRIs, how about psychdelics?!" - this is a joke. Im all for more neuroscience research in every form because our contemporary understanding is embarassingly poor. This is a research tool for understanding what visual brain regions encode, not a mind-control device. Better models of brain function may eventually improve diagnosis and treatment of mental illness. Calling every neuroscience advance "Black Mirror" is not ethical insight. It saddens me the reaction so many commentors have here, it is just lazy technological illiteracy imo.
Wait, this is with a digital twin only? Not fMRI or webcam based?
How can one work on that and not consider that they're pure evil?
the logical endpoint of this research is an AI that generates videos so compelling your brain forgets it was supposed to be working
Ludovico Technique as a service
Fuck this and whoever is driving this type of research. We’re increasingly living in a world designed by sociopaths. What do these people think this research will be used for?
I wish we had the Hippocratic oath for STEM, or at least that they would take ethics seriously rather than an afterthought against the god of Progress at all costs.
What in the zuck is this?
Straight out of Echopraxia.
Will be interesting to see how strong the controlling forces can be - enough to make you miss things in direct perception like in the book, or only softer effects further up the cognition layer stack
I can't wait until I see AI-generated gambling ads that are specifically created to stimulate my brain the most
My brain never liked vertical video, shortform content and AI slop.
Is my brain different or am I just a grumpy millenial hipster?
"Prime Intellect, I would like you to begin stimulating the neurons of the pleasure center of my brain, one at a time, and remember the ones I report to you as being favorable."
Future looks bright....
I am not sure if we should want to understand how our brain works. This is why I don't like libertarian free will people. We need to accept that we don't have free will. Otherwise we will study the brain thinking that we are above it. No. We are soon going to find out exactly how the brain works and I am not sure people understand what that means.
This is the absolutely horrific next stage for social media platforms:
- They're already well able to surface the most addictive short video for a specific user out of millions of real videos.
- But these millions of real videos are just darts thrown into the space of "videos that could hook the user", in the end even the best-selected of them is not perfect.
- Now, behold! AI allows to generate the perfect video to surgically hit all the switches in the viewer's brain and turn it into a zombie hooked for days on end.
Let's hope our regulations hit these "social networks" hard enough so that never dare deploy this kind of technology.