3 ms·
> rules engine with an RGB camera I dont think its true? They use convolutional networks for image recognition and those things can certainly halucinate - e.g.
by _visgean 4y ago
> rules engine with an RGB camera
I dont think its true? They use convolutional networks for image recognition and those things can certainly halucinate - e.g. detecting things that are not there.
- perth 4y agoLike in this image super resolution AI: https://twitter.com/maxhkw/status/1373063086282739715?s=52&t=7DtsZPxff37nJLSeXIuVNw https://twitter.com/maxhkw/status/1373063086282739715?s=52&t...
- stefan_ 4y agoI guess what the grandparent means is that there is some good old "discrete logic" on top of the various sensor inputs that ultimately turns things like a detected red light into the car stopping. But of course, as you say, that system does not consume actual raw (camera) sensor data, instead there are lots of intermediate networks that turn the camera images (and other sensors) into red lights, lane curvatures, objects, ... and those are all very vulnerable to making up things that aren't there or not seeing what is plain to see, with no one quite able to explain why.
- John23832 4y agoYou were correct in the first half. My ultimate point was that hallucinations in this sense are just computational anomalies. There is no human "purpose" to them as the post that I was responding to was trying to infer.
- 2muchcoffeeman 4y agoWe need to stop anthropomorphising machines. Sensor errors and bugs aren’t chemicals messing with brain chemistry even if it may seem analogous. Or maybe when I get a bug report today I’m going to tell them the software is just hallucinating.