• Question: Can artificial intelligence be ethical?

    Asked by AliceRose to Rumman, Pooja, Petros, Greg on 6 Nov 2017.
    • Photo: Greg Chance

      Greg Chance answered on 6 Nov 2017:


      I guess it depends on your stance or application. Nobody would argue that a calculator is unethical but it could be used for unethical things. I think the same applies to AI as long as it is always used as a tool.

      If we were absolutely certain that developing AI would lead to a new form of life i’m sure the debate would be much more of an ethical dilemma. At the moment it doesn’t look like we will produce new life! So i don’t think there is much ethical consideration for now. It may be revised in the future though.

    • Photo: Pooja Narayan

      Pooja Narayan answered on 7 Nov 2017:


      Hi Alice, that’s a great question.

      AI can be every bit ethical. AI was born because humans thought of it. In simple words, if a human wants to create havoc and use the technology for bad things, he could by all means. Technology is “capable” of doing good as well as bad things. It’s up to us how we want to use it. So it’s us who have the power and not AI.

      Cheers!

    • Photo: Petros Papapanagiotou

      Petros Papapanagiotou answered on 7 Nov 2017:


      That’s a good question! There is a whole area of scientific research dedicated to ethical and unbiased AI!

      In most cases, as smart as AI can be, it doesn’t really know what it is doing. It is up to the engineer to set it up so that it avoids making unethical decisions. For example, you have to program your AI to make human life a priority over lifeless objects. So we could say that an AI is as ethical as its engineer can make it be.

      As AI becomes a part of our lives more and more, it becomes harder and harder to “teach” it to be ethical. Some decisions are very hard to make, and even as humans we often have to make decisions that are not clearly ethical, more like a “grey area”. For example, there is a whole debate on making self-driving cars “ethical” and what that really means. If an accident is about to happen and you have to decide between saving 2 children or saving 5 adults, what is the ethical decision?

      tl;dr: AI can be ethical if its engineers put some effort to make sure it is. Sometimes this can be hard.

    • Photo: Rumman Chowdhury

      Rumman Chowdhury answered on 7 Nov 2017:


      What a great question! This is what I do at my job. The hard part, frankly, is defining what it means to be ethical. Sure, we can all say ‘be nice to people’ or ‘don’t hurt someone’ – but NONE of us can say we’ve successfully done that all our lives. We also have pretty complex ways of figuring out what is good or bad, and it can change in every situation. Think about how you talk to your friends – would you act or talk the same way to your teacher? Probably not.

Comments