Which will be the main topics in AI - ethical or technical?

Question

Recently AI had its first victim due to a missing sensor in an autonom driving car.
But AI seems to be our future - with new challenges.

Answer 1

i think it should be both, i didn't have any experiences with AI, but it is very interesting 👍😉

Answer 2

Last year I attended a conference on machine learning and AI and one of the topics was explainability and interpretability of DL models, especially of neural networks with many hidden layers.
We cast votes - if you had to choose - what would you prefer:
* a model predicting with 99% accuracy but explainable in 80%
or
* the other way round?

It touches ethics and compliance issues - you'd probably answer differently when it's about you getting a credit and about you being medically diagnosed...

Answer 3

Like tooselfish said, it's both.

Answer 4

Ai and the idea of it in my opinion are still in Pre infancy stage. when we start creating Hardware specifically for the purpose of running AI and handling its methods successfully the world will change overnight. what I wonder is whether that change will be positive or negative.

Answer 5

Kuba Siekierzyński Interesting point. Nothing is rarely 99% accurate. We need to work with thresholds that create false negatives and false positives.

Algorithm / AI / human system for predicting who's a terrorist. Sensitivity of algorithm too high? Innocent people caught. Too low? Terrorists not caught! This is true of almost all systems. Even big data doesn't necessarily help the problem, as poor quality data, even (and especially) lots of it, can also make a mess!