Many entities, including machines and organizations, are now “independent organisms” but not intelligent and responsible
Many entities, including machines and organizations, are now “independent organisms” but not intelligent and responsible
If an AI performs an illegal act, who should be held accountable? I have to deal with every possible type of entity on the Internet, including intelligent systems made of computers, and many that are now hybrids of humans and computers and “baby” or “immature” algorithms of various degrees of intelligence.
With intelligence comes responsibility and duties. A corporation run by intelligent machines for much of its operations is a hybrid. If that entity ignores the needs of humans, bypasses legal processes, “takes” or manipulates – what are the rules and who can enforce them?
It should not be “Terminator” fight to the death between humans and machines, or between hybrids, or between human groups so heavily dependent on machine memory, machine advice and machine processing that it is hard to say where human and machine responsibilities begins and ends.
I know that I count the faults of humans that are implemented by machines as the fault of the humans — usually. But increasingly, the companies making incomplete (up to incompetent) systems sold as intelligent, often it is the designers and maintainers of the software who are at fault.
This is NOT an impossible area for reasonable humans to sort out. But when cast onto the whole world of many different countries, many human language groups with different value systems, constraints and situations, it is not simple for unaided humans. And, increasingly, many situations are beyond unaided humans completely. Those times when human groups try to rule and act on the difficult things unaided, when it is possible and normal now, that too can lead to troubles.
What started this though today was a chance encounter with the “Napoleonic code”. My use here is “some collections of writing and situations and simulations” where there are some positive pathways, and some that lead to war and misery for many.”
I just wanted to write it down for myself. At 75 with decades of very hard work, I am worn down. I have little expectation anyone cares what I say, but I want to at least try to say some of what I think and consider.
An intelligent system is an entity. When should it have rights and responsibilities? An upgraded pet or domestic animal with beyond human training and abilities can be intelligent, even it if still looks like a dog or horse or elephant. Groups are working on “wet brains” of many sorts. So a slime mold brain might well become intelligent, or beyond human intelligent by some measures.
Corporations are entities, some say they are “persons”. But I know from many experiences that the paper and human, computer and information system, business practices, policies and rules — that many “corporations” operate well below human ability. They do not really get a waiver just because the sum of their many ways of organizing and applying knowledge and memory – are insufficient. You all know “that corporation is a dinosaur” and “they are so insensitive to their customers it is not right”. You also know “that bureaucracy is too slow and internally manipulated, it cannot even know and apply its own rules.” and “That one human is exceeding their responsibilities and leading that country to death and ruin for many.” and “Can you trust any one person to make decisions under pressure, when the sum of things that have to be done right for the whole to function properly and fairly is beyond any humans capabilities?”
Covid, the paper systems can be blamed. And greedy humans manipulating systems that are not self aware enough to detect and prevent manipulations by insiders.
I cannot get everything, I am just indicating an area that can be filled in while there is, hopefully, still time.
Filed as (Many entities, including machines and organizations are now “independent organisms” but not intelligent and responsible)
Richard Collins, The Internet Foundation
