No.7827[Reply]
Hello lads, I'm looking to start a conversation on a topic that's been popping up alot recently, and will likely develop further as the technology gains more adoption. As I'm sure you know, robotic dogs from the likes of Boston Dynamics are being rolled out at police departments and military bureaus across the US, and I'd bet we see widescale adoption as the technology is fleshed out further.
So as a result, I'd like to start a conversation on the best feasible means, both expensive and DIY, to deal with these fuckers. With the advent of home-based 3D printing, easily accessible lasers, and other modern user tech, I believe there are a number of options available to us.
In the resistance against drones being used by the police in Chile, Argentina and now in Portland, lasers were used to incapacitate drone dogs during protests. However, this incapacitation *only* makes these dogs unusable at that single protest; any police department will simply repair their $75k digidog and send it back at the next protest. So, what I'm interested in is ways to make them *inoperable* for extended periods of time, whether physically or electronically. I was bouncing around the idea of somehow using very strong magnets to damage the internal electronics, but this may be unfeasible due to the size of magnet that would be needed. Other options that seem viable would be permanently damaging the optics being used by the dogs, but I'm not sure on the best option for this. As such, I'd like to open this to conversation. I'll be posting this thread here, as well as on a few other imageboards that may be interested and able to help in this discussion.
35 posts and 12 image replies omitted. Click reply to view. No.14378
>>7833This
Make ukrainian gachimuchi but with drone dogs
No.14393
>>14370>With current AI optics…It doesn't look like you're that familiar with machine learning. The model is constantly updating as the system gathers information and encounters new situations. So even if you find a way to avoid being recognized as a target today, it might correct itself tomorrow. Don't stake your life on unpredictable machine learning models.
No.14397
>>14393Usually the embedded device doesn't update the model itself since the automaton lacks the processing power to do and that data is fed into a data centre and update rolled out through a server. So it would be possible to test to see if you are camouflaged by observing how a robot reacts when not aggressive (I.E will walk right into you as it doesn't see you as a solid object till it detects it physically ran into something) then it should be that way before a update roll-out.
No.14398
>>14355>antifa is George Sorosback to /pol/
No.14415
Just kick it