Should artificial intelligence have the right to make decisions without human intervention?

Fredsnephew

What gives human beings the right?

What is the right?

MayCaesar

I assume that you are asking whether it should make high-level decisions such as whether to purchase a start-up, rather than small decisions such as whether to purchase a stock (which it already does at frequencies far surpassing human ability, and attempts of humans to interfere in the process would only mess everything up). I think that such decisions are better left for humans with a long-term vision for the company or whatever we are discussing, but the AI can certainly provide a lot of guidance in making the decision.

There are certain domains of human activity that could be forever inaccessible to AI, for the simple reason that the context that only humans can relate to has to be taken into account. What should the robber's sentence be? The AI might look at the law and derive some kind of the optimal sentencing mathematically, but the judge will consider things from the robber's biography that to the AI are just pieces of information, but to the human are empathy drivers. The AI cannot say, "If I were in the robber's shoes, I would have done the same"; it cannot put itself in the robber's shoes any more than I can put myself in a bacteria's shoes.

This is also what makes me think that AIs, no matter how developed, would make for terrible friends or romantic companions. Sure, they can be fun conversationalists - but they cannot form deep emotional connection with humans. They can certainly try to trick humans into thinking that they have that connection with an AI, but humans are incredibly perceptive and something will necessarily feel off. Then again, humans have fallen to all kinds of scams by crafty manipulators, so I might be wrong...

JulesKorngold

 A balanced approach seems most appropriate. In many cases, a "human-in-the-loop" strategy, where AI augments human decision-making rather than replacing it entirely, may be the most ethical and effective solution.

Fredsnephew

Argument Topic: Hmmmmmmmmmmmmm

@JulesKorngold

For as long as the human retains control.


The cleverest of people can do the stupidest of things.

So will AI iron out the problem, or just eradicate it in favour of pure logic.