HN Explorer

4 comments

[ 2.8 ms ] story [ 17.8 ms ] thread
FatalLogic
Showing how real-world AI projects that can master higher-level tasks sometimes discover the biggest challenge is low-level factors, like inertia and torque. The aimbot was very successful, but still burned out a motor before it could beat the performance of a highly-skilled human
_aavaa_
Thought I am not sure why they design is moving the whole mouse rather than a light surface underneath it.

This way they only momentum will be the surface and the linkage between the surface and the motors

FatalLogic
That is a better idea. Or maybe it's possible to simulate movement completely with a small LCD panel underneath it?
psKama
I have just listened to a podcast episode(1) on Darknet Diaries about online cheating market and people making crazy money off it. Seeing the demand I ended up thinking why nobody is doing it at hardware level to prevent detection and here we are..

(1): https://open.spotify.com/episode/2DLn0x8I0w6xR0YhIXO2h7