Should we be worried about AI?

Summary #

AI is not an existential risk for humans. But can destabilize society through mass unemployment and loss of trust.

Explanation #

There has been rapid development in AI technology in the last few years. This has created worries about the technology harming people. One worry is that AI is an existential risk for the human race. The fear is that it could become smarter than people, become powerful and uncontrollable, and kill/subjugate the human race. This scenario is highly unlikely. This is due to the lack of self-awareness in AI. AI is not self-aware. It cannot observe itself. Due to the lack of self-awareness, it cannot act independently of its programming. It will be incapable of having any intention, let alone harming humans. So AI is not an existential risk for the human race. AI will always lack self-awareness, irrespective of future advancements in the field. This is due to a fundamental limitation in computing. Computers are purely mechanical systems. They take input from outside the system, process it, and give output. A self-aware system gives input to itself. For example, a human being can tell himself to change his behavior. A computer cannot give input to itself. It can only take input from outside the system. Thus computers/AI can never be self-aware. So AI will never be an existential risk for humans.

Another worry about AI is that it will create mass unemployment. This scenario is possible. It will happen if advances in AI technologies outpace the time take for people to acquire new skills. For example, if progress in autonomous truck technology outpaces the time taken for truck drivers to learn new skills, the truck drivers will become unemployed. Unemployment of a large section of the population will create instability in society. Unemployed people will feel rejected by society. They will stop obeying laws because they have nothing to gain by being law-abiding citizens. They will create a disobedience movement. The government won’t be able to control the disobedience of a large section of the population. It will end in the complete breakdown of civil society.

Another worry is that AI will reduce trust people have with each other. This scenario is possible. Advanced AI technology will assist people in personal decisions. For example, AI can use data to make a profile of a person and use it to recommend a movie he/she will like. This type of AI-assisted decision making will become better and better with advances in AI technology. Eventually, people will start making AI-assisted decisions in all walks of life. They will trust AI’s judgment more than their judgment. They will lose their ability to take risks and think independently. The lack of risk-taking will make people invest less in relationships. This will lead to a general loss of trust among people for others. The loss of trust will lead to instability in society. Even small conflicts can escalate to a civil war.