Microsoft CEO: “killswitch” should be set to AI to prevent it from doing wrong

microsofts-ceo-sent-out-a-giant-manifesto-to-employees-about-the-future-of-the-company

According to a report from Quartz, Satya Nadella, CEO of Microsoft has recently talked about his opinions on artificial intelligence. He said in the meantime of helping people to maximize efficiency, AI also need to take care of mankind’s self-esteem and protect their privacy. However, in order to prevent robot from doing wrong, we need to set a “killswitch” for AI.

The sign of robot revolution has been vague, Elon Musk as well as some famous scientists have been paying great attention to the effect that AI development might bring to mankind. According to their opinions, we must think better of what does it mean for us to live with AI, and they believe AI will easily exceed human intelligence.

Microsoft CEO Satya Nadella had published an article on Slate in which he stated his 10 opinions on AI.

Here’s what he believes:

1. AI must be used to assist mankind. Nadella has quoted Three Laws of Robotics from Isaac Asimov, the famous science-fiction writer. The core of the laws is that robot should never hurt mankind. While Nadella’s suggestion is more profound: the program settings of robot should be actively assisting human beings. He wrote, robot should be “cooperation robot or co-bot.”

2. AI must be “transparent”. Mankind must have a thorough understanding of robots. He said, “Robots will acquire some understanding to mankind, and mankind must understand robots.”

3. AI must fully improve its efficiency and meanwhile not hurting the self-respect of mankind. Technology and AI should promote the diversify and interconnection among people, instead of leading them to estranged from each other- which unfortunately is what people are good at throughout history.

4. AI should protect privacy of mankind and never let it out.

5. Accountability algorithm should be set to AI so that mankind could avoid unexpected injury that robot might bring. In other words, AI needs to be set a “killswitch” in order to be prevented from doing something wrong.

6. AI should never have prejudice. AI should represent the will of everyone rather than a particular group.

AsimovRobots101510

7. In order to promote the normal development of AI, mankind would have to possess empathic ability. As was pointed out by Nadella, to create a robot that could truly understand humans, mankind first need to ensure they could understand the “thoughts and feelings” of each other.

8. We need to invest in education, develop and fund training programs to make sure the potential benefits of AI could easily be enjoyed by mankind. Nadella has given an example, ” engine driven weaving machine was invented in 1810, but it took 35 years to make a transformational impact to the clothing industry due to lack of well-trained workers at that time, the requirement can’t be satisfied.”

9. Mankind must remain creative which is one of the qualities of human beings. However, robot will be able to deepen the application of mankind’s creative result, this is also a “hot” area that many AI researchers are researching.

10. Mankind should reserve judgement and take responsibilities. As is written by Nadella:” We might be willing to accept the diagnosis or legal decisions made by the computer, but we will still rely on humans to take ultimate responsibility for the result.”

Leave a Reply

Social media & sharing icons powered by UltimatelySocial