The future has arrived. How should we embrace the era of artificial intelligence?

Stills from "Terminator"

Machines do not have advanced intelligence to cause fatal damage to humans. Like the plot in Terminator, they have no compassion and aim to chase humans. In science fiction movies, we feel this threat from the future more than once, lurking in the laboratory of some evil scientist, or undercurrents in that system called "Skynet".

There is no good or evil in technology, it depends on who it is in the hands of. The one who uses technology to do evil should not be the technology itself, but the "people" who use technology.

With the widespread application of artificial intelligence (, AI), more AI products have entered people's lives. Before machines have real intelligence, we need to prepare for some philosophical and ethical issues, such as whether a self-aware machine should be developed? Is human-machine marriage legal? How should machine crime be defined? Does machine decisions contain discrimination? etc.

The AI ​​era has quietly arrived, and it has posed many challenges to law, psychology, politics, business, infrastructure, community, services, etc. in human-computer interaction. Artificial intelligence ethics is still in its infancy, and scientists and sociologists are aware of the existence of many problems, but fail to give convincing answers.

For example, if an autonomous driving ( ) has a car accident, who should we blame? General algorithm or programming errors will not cause any life worries, but driverless driving has this risk. Consumers are unwilling to spend money to buy a potentially dangerous product. If the safety of life and property is not guaranteed, and if the relevant legal responsibilities are unclear, autonomous driving will cause serious social problems.

Creating Artificial Intelligence (AI)

In terms of scientific research, intelligent machines have extended human exploration capabilities. For example, after the Chang'e-3 landed on the moon, the Yutu lunar rover began to excavate and analyze soil samples, video transmission and other work. "Yutu" has automatic navigation, radar lunar measurement, energy collection, astronomical lunar-based optical telescope, extreme ultraviolet camera, spectrometer, stereo camera and other equipment to help humans complete a large number of scientific research tasks on the harsh moon. In the future, more and smarter machines will participate.

On December 14, 2013, my country's lunar probe "Chang'e 3" soft landed on the surface of the moon

Before the arrival of self-aware machine intelligence, how to protect AI technology from developing in a healthy direction, rather than being exploited by the evil of human nature and becoming the endorsement of human civilization is a problem that we must face and need to be solved urgently. Technology itself has no good or evil, but the humans who apply it have a distinction between good or evil. How to implant more good things into unconscious machines? How to use ethics to regulate artificial intelligence? This is worthy of deep thought by every reader who cares about the future destiny of mankind.

Animation "Iron Arm Astro Boy"

There is a branch of philosophy, ethics (), also known as moral philosophy, which studies the code of conduct recognized by human society and its value, and then constructs some moral guiding principles to judge the right and wrong of behavior, good and evil, beauty and ugliness, wisdom and ignorance, justice and crime. Ethics is roughly divided into four fields: metaethics, normative ethics, descriptive ethics and applied ethics. Artificial intelligence ethics basically belong to the category of applied ethics.

More than two thousand years ago, the ancient Chinese philosophers Lao Tzu, Zhuangzi, Confucius and Mencius had systematic research on moral ethics. Their ideological influence has been deeply integrated into the blood of Chinese civilization.

Nobel Prize winner in Economics (1978), Turing Prize winner (1975), and pioneer of artificial intelligence Herbert Simon (1916-2001) once said: "With social consequences, I believe that every researcher has the responsibility to evaluate and strive to inform others of the possible social consequences of the research products he is trying to create." The purpose of artificial intelligence research is to benefit the entire human civilization, so "do not do evil because it is small, and do not do good because it is small." Rather than using ethics to restrain machines, it is better to use them to restrain humans themselves. Therefore, artificial intelligence ethics is the code of conduct for human development and utilization of AI.

Many existing AI technologies are not explanatory, and they do not know the real reason for the problem with the model or product. Some technical defects are fatal to humans, and users should have the right to know about possible adverse consequences. Therefore, the evaluation of the quality and performance of AI products requires a more scientific, rigorous and systematic evaluation method.

Humans need to carefully develop AI technology to ensure that it promotes the evolution of civilization, rather than becoming a new weapon for humans to kill each other. Human beings are like mayflies in heaven and earth, and only the glory of reason is eternal. Fair and open cooperation, rather than malicious competition, is the AI ​​path we are going to take.

In the United Nations Headquarters Square, a sculpture "knotted gun" stands, symbolizing "peace, tolerance, understanding and non-violence". It is a work by Swedish sculptor Carl Frederick Lutwald (1934-2016), presented to the United Nations by Luxembourg in 1988

Technology is a double-edged sword. Humans use it to overcome obstacles and are at risk of self-harm. More and more AI products have the ability to make decisions, from weak to strong. Do they accurately represent our wishes while facilitating human life? Will the smart assistant know his master too well and become a hidden danger of leaking secrets? How to protect personal safety as much as possible in autonomous driving? Should autonomous killing machines be banned? Many AI ethical issues await people's rational thinking.

"Ethics of Artificial Intelligence" is a popular science book that has attracted the attention of the art of science. It has almost no mathematical formulas and preliminary knowledge, and is aimed at all readers who are interested in artificial intelligence. Everyone can talk about AI ethics and pay attention to the fate of humans and AI. These topics are not limited to rigorous scholars, they are the reality that humans need to face in the future. For example, intelligent manufacturing has the risk of large-scale unemployment. How should education in the AI ​​era cultivate new types of skilled workers? Will robots intervene in human life bring some structural changes to society? In the era of big data, how to better protect personal privacy? Before the complexity becomes uncontrollable, we must anticipate and analyze the worst situation. The purpose of writing this book is to reveal the problems, explain some ideas at all, and inspire more readers to participate in thinking and discussion.

"Artificial Intelligence Ethics"

ISBN: