Introduction

The difference between artificial intelligence ethics compared to other technical ethics: it is not about the relationship between people and the relationship between people and nature, but about the relationship between people and the machines they create. This human-like and even superhuman machine has stronger social interaction and behavioral initiative, and many people propose to give artificial intelligence the identity of legal subjects to regulate it.

Li Huajing/ Text

Start: Artificial intelligence should be good in nature

Machine ethics is not a new topic, and the nature of goodness is determined from the very beginning. In 1942, American writer Isaac Asimov proposed three laws of robots with ethical attributes in his short science fiction novel "The Ring Dance": First, robots must not harm humans, or they cannot ignore the harmed humans; Second, without violating the first law, the robot must obey human commands; third, without violating the first and second laws, the robot has the obligation to protect itself.

Although these three laws come from science fiction writers nearly 80 years ago, they have always accompanied and profoundly influenced the development of artificial intelligence technology. They are like the "Sword of Damocles" hanging in the air, always warning people to prevent artificial intelligence from happening like It blooms beautifully like a poppy and bears bad fruits.

Although it is not yet certain that AI is good or bad, what is certain is that AI should be good in nature. Human beings must do their best to ensure that the development of artificial intelligence is beneficial to humans and the environment. Especially today when artificial intelligence is exploded again, the nature of artificial intelligence should be paid attention and reiterated.

More importantly, artificial intelligence ethics have their complexity and uniqueness. To borrow the judgment of Tesla CEO Elon Musk, artificial intelligence is an "immortal dictator" created by humans without realizing it. This image reflects the difference between artificial intelligence ethics compared to other technical ethics: it is not about the relationship between people and the relationship between people and nature, but about the relationship between people and the machines they create. This human-like and even superhuman machine has stronger social interaction and behavioral initiative, and many people propose to give artificial intelligence the identity of legal subjects to regulate it.

Inherited: Difficulties in making decisions about artificial intelligence ethical

How to make artificial intelligence with strong overflow and driving force "leader effect" a good person and do good things is very necessary but also extremely challenging.

A typical case is the Notre Dame fire in April 2019. In order to curb fake news on its own website, Google's video service company has added links to Wikipedia and encyclopedia entries to news videos since 2018, which was originally a "good thing". However, after the Notre Dame fire in April 2019, the link to the 9/11 terrorist attack was automatically linked under a live video about the fire. This means that netizens will be recommended when browsing the live reports of Notre Dame fire. Pay attention to the 9/11 event entry in the Encyclopedia of Britannia. This connection immediately aroused strong dissatisfaction among netizens, and was even criticized for confusing the public and creating conspiracies, which is really a "bad thing".

How to judge "good things" and "bad things" is difficult for people and machines.

Therefore, the current standards for making ethical decisions in artificial intelligence are difficult to unify, and they should not even be uniform. The Ethics Committee of the German Federal Department of Transportation and Digital Infrastructure, in 2017, jointly proposed about 20 driverless rules with scientists and legal experts. For example: driverless cars should first consider protecting human life, even if doing so will harm animals or lose property; if an accident is inevitable, driverless cars cannot choose to save whom, and should not be based on age, gender, or race. All human life is equal if you make decisions. And, they promise to enforce these rules in some way. However, the US Department of Transportation has a different attitude in this regard. For example, the annual guideline issued is voluntary rather than mandatory, and advocates that artificial intelligence, a new thing that is still developing rapidly, is to eliminate unnecessary action. Obstacles, flexible and technologically neutral policies can safeguard and improve safety than specific technical solutions.

From the above, it can be seen that artificial intelligence ethical decisions are not black or white or either one or the other. So, who will control this direction? The eyes are focused on people.

Translated: Multiple Subjects of Artificial Intelligence Ethics

The ethical subject of artificial intelligence involves all aspects, and the first thing to bear is the three pillars leading the development of artificial intelligence technology - the people behind algorithms, computing power and data calculation. How can they not make mistakes and drive artificial intelligence to be positive and good? To put it simply, they must achieve algorithmic ethics - brain power for good; computing power ethics - mind power for good; computing material ethics - motivation for good. Therefore, the programmers, scientists, engineers and consumers behind artificial intelligence algorithms, computing power, and computing materials are all important subjects that affect artificial intelligence ethics, but they are also difficult to stand alone. To set the direction and rhythm for artificial intelligence ethics, it is necessary to integrate the power of multiple subjects so that everyone does not go astray. And those who own and are good at integrating are only entrepreneurs.

Artificial intelligence entrepreneurs control teams, opportunities and resources, integrate hard technology and soft ideas, and dominate the ethical direction of artificial intelligence. Entrepreneurs will also fall into a dilemma between pursuing personal interests and following ethical norms. However, entrepreneurs do not make choices in the "light zone" or "dark zone", but "draw maps" through innovation, risk taking and advanced actions. In other words, ethical oriented entrepreneurial actions help integrate forces from all parties to creatively solve the ethical problems of artificial intelligence.

Combined: Shangshan Ruoshui's artificial intelligence entrepreneurship ethics

Speaking of the way of following good deeds, I believe many people will think of Chapter 8 of the Tao Te Ching, "The Supreme Good Deeds are Like Water". Although the "good" here does not only mean kindness, the seven virtues summarized by Lao Tzu - living, being good at the land; being good at the heart; being kind to be kind; speaking, being trustworthy; being good at governance; doing things , good at being able; move, good at being good at being good at being - there are ways to refer to the ethical practice of artificial intelligence entrepreneurship.

In April 2019, the EU issued the "Ethical Code for Trusted Artificial Intelligence" (hereinafter referred to as the EU Code), which listed seven key conditions: human autonomy and supervision, technology robustness and security, privacy and Data governance, transparency (traceability), diversity, non-discrimination and fairness, social and environmental well-being, accountability.

To this end, it is better to combine the seven-point water virtue of the Supreme Goods and the seven EU guidelines with the entrepreneurial ethics actions of Chinese artificial intelligence enterprises, and extract the similarities between theory and practice, classics and exploration, and the interweaving of the East and the West, and explore the entrepreneurial ethics of artificial intelligence. direction of action.

Residence is a good place, which originally means choosing a place. The implication for the entrepreneurial ethics of artificial intelligence is that goal positioning should focus on social value. The EU norm “social and environmental welfare” also emphasizes that the application of artificial intelligence systems should promote positive social change and enhance sustainability and ecological responsibility.

The heart is a good place, the original meaning is a deep mind. The inspiration to the entrepreneurial ethics of artificial intelligence is that market development should be towards inclusive innovation. EU norms “diversity, non-discrimination and fairness” mean that AI systems should inclusively consider the overall scope of people’s abilities, skills and requirements to ensure accessibility.

Be kind and kind, originally meant to be friendly to others. The inspiration to the entrepreneurial ethics of artificial intelligence is that user services should focus on friendly experiences. EU standard “Ruleability and Security of Technology” also advocates that artificial intelligence algorithms should be robust enough and reassuring to deal with errors or deviations in the evolution cycle of artificial intelligence systems.

Words are good faith, and the original meaning is to abide by trust. The inspiration to the entrepreneurial ethics of artificial intelligence is that value realization should be based on integrity. EU norms “Privacy and Data Governance” mean that citizens’ data should be fully controlled by themselves, and the application of citizen-related data cannot harm or discriminate against citizens.

Government, good governance, originally means good governance. The inspiration to the entrepreneurial ethics of artificial intelligence is that the governance system should strengthen institutional construction.

Things are good at abilities, and the original meaning is to give full play to the strengths. The inspiration to the entrepreneurial ethics of artificial intelligence is that technological development should be based on innovation capabilities.

Movement, good time, originally means to seize the opportunity. The inspiration to the entrepreneurial ethics of artificial intelligence is that sustainable growth should seize entrepreneurial opportunities.

Epilogue: Where does the water flow

The positive mechanism of ethics in the innovation and entrepreneurship development of artificial intelligence technology still needs to be explored in depth. Harvard psychologist Steven Pinker once questioned the positive significance of ethical review to the scientific process. He published an article in the Boston Globe, proposing that excessive ethical supervision will interfere with the speed of innovation and have little positive effect. It can be seen that ethics is not a simple way to unblock technology, and it cannot be as simple as controlling switch valves, but a complex social system project.

In particular, the latest developments in the combination of blockchain and artificial intelligence highlight the importance and urgency of the entrepreneurship ethics of artificial intelligence. When artificial intelligence meets blockchain, it will help push data resource management and information technology applications to a higher level, but it will also bring more challenges to public awareness and thinking paradigms, thereby aggravating ethical conflicts.

Optimistic or pessimistic about this? Optimists believe that because of the distributed network of blockchain, new technologies such as artificial intelligence can be integrated together to form a new governance structure, thereby raising the threshold for "evil doing" and in education, employment, pension, targeted poverty alleviation, medical and health , product anti-counterfeiting, food safety, public welfare, social assistance and other areas of people's livelihood have broad application prospects. Pessimists have suggested that applications related to blockchain technology, such as issuing digital currencies and recording data and privacy, have huge risks. Don’t you see that issuing coins has become a cut of leeks, and privacy is still difficult to protect.

The new force in exploring the answer is still the entrepreneur. In June 2019, a new cryptocurrency, Libra, was officially released. Immediately, Ma Huateng commented on WeChat Moments: "The technology is very mature and not difficult. It depends on whether the supervision allows it. .”

But how can we make the supervision reliance on mountains and the flood of technological innovation become an ecosystem that benefits mankind? How can we realize the ethical value of artificial intelligence "benefits all things without fighting, and dealing with what everyone hates"? We still need to continue to observe. Just like Confucius' reminder to Zigong, "A gentleman must observe it when he sees a big flood." We look forward to the full flow of scientific knowledge in artificial intelligence and the healthy development of technological innovation, so as to benefit the society and the public better and faster.

This article is a phased result of the postgraduate teaching reform project (number) and the National Natural Science Foundation of China project (number) for scientific research at Beijing Forestry University.

About the author | Li Huajing: Professor and doctoral supervisor at Beijing Forestry University, with the main research direction being innovation and entrepreneurship management