#One could read patterns in searching. keystrokes. clicks. eye movements. what you say not just online but anywhere on a computer | Explore Tumblr posts and blogs

anarcho-masochist · 1 year ago

Text

Don't they already do AI-controlled drone strikes? I remember news about it 1-2 years ago. Give me a second to fact check myself.

Here's an article from npr posted June 1, 2021

There are several others from around that time frame.

So after looking up your article, what the US is doing is arguing against new laws to regulate AI weapons. They're already using them, they just don't want their usage to become restricted at all, so, as you say, a very bad sign.

This and multiple other articles in the same search mention continued development and plans to increase numbers so that's...great (sarcastic).

america is an evil-fucked-up-beyond-words-dystopia like actually this is the premise of psycho pass more or less what the actual fuck we need a revolution

#I will say in response to your psycho-pass comment and tagging:#at the very least this is currently restricted to military use and does not purport in any way to 'read minds'#but it's still bad #and it does seem like we get increasingly close to that dystopian future #especially on a technological level #since we are now capable of surveillance on an easily comparable level sans mind reading (sort of)#I wasn't planning to initially but let me explain:#currently it's used for advertising. every electronic device including those ridiculous screens instead of windows on supermarket freezers #we have the technology - those screens i mentioned in particular track your eyes and read your facial expression to tell whether you like #the products you're looking at #a crucial aspect of psycho-pass' premise is that the Sybil System doesn't *actually* work #It can't tell if someone actually *is* a criminal from their psycho-pass. It can tell that they seem like one and thus have a 'higher risk'#It is everyone believing in the system that makes it 'justifiable'#In some ways we've already surpassed the technology level in canon #spoilers for the end of season 1:#Since in 2014 when it was written the human brain was still far more powerful than AI #This is no longer the case and it would be ludicrous to use human brains instead of AI which is cheaper and runs on computers with a higher #processing power than human brains #With AI as integrated as it is becoming in our daily lifes and automated surveillance already here #One could easily train AI to do things that would probably be more effective than just reading faces #One could read patterns in searching. keystrokes. clicks. eye movements. what you say not just online but anywhere on a computer #One can track every location you go to. Combined with digital behavior this creates an image of your intentions #'it would be far too expensive to do this for every person'#well...they serve you custom ads #a lower threshold for sure. but my point is we're not that far away from a serious risk of automated justice for everyone #what they would do realistically is use your own devices' processing power plus some very eco-unfriendly server farms #and it would simply detect patterns in your behavior and flag it if it gets above a certain threshold of 'reason for concern'#ex: your phone tracks your location and you went to a protest...ah hold on that's already used to arrest protestors...#you get the point #the main thing is that as far as we know no one has invested enough *money* into the idea or else it would be happening

9 notes · View notes