Týden na ScienceMag.cz: Budeme soudit i umělou inteligenci? (diskuse)

[xy]

Investice? To jako vážně?

[LarryL]

Nechápu kam ti lidi na ty blbostě o AI chodí. Vypadá to, že už i právníci z University of Brighton si myslí, že za pár let bude mít AI svá práva, přitom je to jen obyčený software, který se lidské inteligenci a samostatnosti neblíží ani náhodou. A ještě dlóóóóóóóóuho blížit nebude.

[xyc]

Uděláme programátorskou konferenci o právu. Třeba jim to dojde.

[Petr]

Někdo by si měl sednout s paní Alžbětou Krausovou, a podobnými lidmi a vysvětlit jim, co to ta umělá inteligence skutečně je... Oni to sice myslí dobře, ale o umělé inteligenci mají představy na úrovni Sci-Fi literatury....

[Fluttershy, yay!]

ACKSHUALLY

It is not to recognize a robot or a humanoid robot or some form of artificial intelligence as something close to us, to give it the same rights such as the right to life, the right to a language, the right to privacy and so on. It is a construct, like the term ‘legal person’.

“Legal person is an artificial concept of our society. We created the concept of a legal person and applied it to companies because we needed something that transcended humans. With electronic persons or robots, there are similarities and differences. Imagine a robot who is buying pills for you or other items to help care for elderly people. We need rules that make it so it is actually legal for the robot to purchase the items.

…

With drones, even though there still much debate, we still presume that the one who is in control of the drone is the one who is legally responsible if anything goes wrong. However, with robots that are equipped with artificial intelligence, it is very difficult to determine who bears ultimate responsibility due to the fact that the robot is learning all of the time.

You have a manufacturer who makes it and programs it and may have somebody teach it basic information, and then the robot is in contact with other people and society. If the robot causes any damage, it is very hard to prove who is responsible.

The manufacturer will say that it is the person who initially instructed the robot poorly, and the person may say it is a fault in the programming or construction that caused to robot to inflict harm. So, we need clear rules on how to deal with that, and an electronic person is one of the possible solutions.