Chapter 850: : 0 leaks

With the rapid development and wide application of artificial intelligence technology, the curtain of the intelligent era is opening, and ubiquitous data and algorithms are spawning a new type of artificial intelligence-driven economic and social forms. Artificial intelligence can become a force of "goodness" and continue to benefit mankind and human society, but it also brings ethical and social impacts such as privacy protection, false information, algorithmic discrimination, and network security, triggering how to bring new technology Come to an extensive discussion on maximizing personal and social well-being. The ethics of artificial intelligence began to go from behind the scenes to the foreground, and became an important guarantee for correcting and correcting the narrow technical dimension and benefit limitations of the technology industry. As some scholars have said, let ethics be a fundamental part of artificial intelligence research and development.
In this context, from the government to the industry to academia, there has been a wave of exploration and formulation of artificial intelligence ethical principles around the world. For example, OECD and G20 have adopted the first artificial intelligence principles signed by governments of various countries and become the first international intergovernmental consensus on artificial intelligence governance, establishing a people-oriented development concept and agile and flexible governance. China ’s new generation of artificial intelligence governance principles have also been released immediately, proposing eight principles of harmony, friendship, fairness and justice, inclusive sharing, respect for privacy, security and controllability, shared responsibility, open collaboration, and agile governance to develop responsible artificial intelligence . It can be seen that all circles have basically reached consensus, and the new round of technology development and application represented by artificial intelligence cannot be separated from the value guidance provided by ethical principles.
Therefore, we believe that it is necessary to build an artificial intelligence ethics based on the values ​​of trust, happiness, and sustainability in order to help reshape the trust of the digital society, realize the benign interaction and development between technology, people, and society, and shape healthy and inclusive Continuous smart society.
Construct a trustworthy artificial intelligence rule system
Although technology itself does not have moral and ethical qualities, people who develop and use technology will give it ethical value, because software that makes decisions based on data is designed by people. They design models, select data and give data meaning, which affects people ’s behavior. Therefore, these codes are not value-neutral, and include too many decisions about human present and future. Therefore, the assertion that technology has no value is not correct. Now people cannot fully trust artificial intelligence. On the one hand, because people lack enough information, they do not have enough understanding of these technological developments that are closely related to our lives and production; on the other hand, because people lack the ability to foresee, they cannot predict that companies will take What you do with your own data cannot predict the behavior of artificial intelligence systems.
Therefore, it is necessary to build an artificial intelligence rule system that can make people trust, and let the technology accept the value guidance. As the starting point for establishing technological trust, the development and application of artificial intelligence need to follow the four aspects of the concept to provide the necessary value guidance for the development of artificial intelligence.
One is available. The primary purpose of developing artificial intelligence is to promote human development, bring well-being to mankind and human society, and achieve inclusive, inclusive and sustainable development. To this end, it is necessary to allow as many people as possible to obtain and use artificial intelligence, so that people can share the technical dividend and avoid the technical gap. Usability also means human-oriented development concepts, human-computer symbiosis, inclusiveness, and fairness and non-discrimination. It is required to practice the concept of "ethics by design" and integrate ethical values ​​into the design of artificial intelligence products and services.
The second is reliable. Artificial intelligence should be safe and reliable, able to prevent malicious interference such as network attacks and other unintended consequences, and achieve safety, stability and reliability. On the one hand, artificial intelligence systems should undergo rigorous testing and verification to ensure that their performance meets reasonable expectations; on the other hand, artificial intelligence should ensure digital network security, personal property security, and social security.
The third is knowable. Artificial intelligence should be transparent and understandable by humans, avoiding the "black box" of technology that affects people's trust in artificial intelligence. At present, the academic community and the industry have been working on solving the "black box" problem of artificial intelligence and implementing an understandable artificial intelligence algorithm model. In addition, algorithm transparency does not explain every step of the algorithm, the technical principle and implementation details of the algorithm, and the simple disclosure of the source code of the algorithm system cannot provide effective transparency, but may threaten data privacy or affect technical security applications. On the contrary, achieving effective transparency in the behavior and decision-making of artificial intelligence systems will be more desirable and can also provide significant benefits, such as providing explanations for the decisions and behaviors of artificial intelligence systems when appropriate, rather than simply pursuing technological transparency. Finally, in the process of developing and applying artificial intelligence, opportunities should be created for public participation and support for the exercise of individual rights including privacy, freedom of expression, relief, etc.
The fourth is controllable. The development of artificial intelligence should be placed under the effective control of human beings to avoid harming the interests of human beings or the whole. In the short term, the development and application of artificial intelligence should ensure that the social well-being it brings significantly exceeds the expected risks and negative effects it may bring to individuals and society, and ensure that these risks and negative effects are controllable and After the occurrence, actively take measures to mitigate and eliminate risks and their impacts. In the long run, although people cannot predict whether and how universal artificial intelligence and super artificial intelligence can be realized, nor can they fully predict its impact, but should follow the principle of early warning, prevent future risks, and make possible future general artificial intelligence And super artificial intelligence can serve the interests of all mankind.
Of course, the establishment of trust requires a set of rules. Under these principles, one can explore the formulation of standards, laws, international conventions, etc. An inclusive, prudent, agile and flexible governance approach is needed, which can adopt post-regulation or adjust the development and application of new technologies such as artificial intelligence through the introduction of standards, industry conventions, ethical frameworks, best practices, and technical guidelines to support industry self-regulation.
Ensure everyone can pursue digital well-being
First, to protect the digital well-being of individuals, everyone has the right to pursue digital well-being. On the one hand, it is necessary to bridge the technological divide and the digital divide. Nearly half of the world ’s population does not have access to the Internet, and vulnerable groups such as the elderly and the disabled cannot fully enjoy the convenience brought by digital technology. On the other hand, reducing and preventing the negative impact of digital technology on individuals, excessive network use, information cocoon rooms, algorithmic bias, fake news and other phenomena have exposed digital products to personal health, thinking, cognition, life and work. Negative impacts call for the Internet economy to shift from attracting and even grabbing users' attention to maintaining and promoting users' digital well-being, and require technology companies to integrate the promotion and protection of users' digital well-being into the design of digital products and services.
Secondly, to guarantee individual work and free development, everyone has the right to pursue happy work. Although some claim that artificial intelligence will lead to large-scale unemployment and the emergence of useless classes, at present, the economic impact of artificial intelligence is still relatively limited, it is impossible to cause large-scale unemployment quickly, and it is impossible to end human work because of the adoption of technology and Penetration often takes years or even decades, and adjustments and changes in production processes, organizational design, business models, supply chains, legal systems, cultural expectations, etc. are required. However, the content, nature, methods and needs of the work may change greatly, requiring people to acquire new knowledge and skills. Forecasts show that over the next 20 years, more than 90% of jobs will require more or less digital skills. Moreover, new technologies such as artificial intelligence and robots are spreading and infiltrating from the ICT field to the real economy, service industry, agriculture and many other economic sectors. Current employment and economic structure. What needs to be done now is to provide appropriate skills education for current and future workers, provide fair opportunities for retraining and re-education for workers in transition, and support early education and lifelong learning.
Shaping a healthy, inclusive and sustainable smart society
Technological innovation is the most important factor to promote the development of mankind and human society. In the 21st century today, the technological capabilities possessed by mankind and the potential for "goodness" of these technologies are unmatched at any time in history. In other words, artificial intelligence and other new technologies are themselves tools for "goodness", which can be a force for "goodness" and are used to solve various challenges facing human development. At the same time, the challenges facing mankind are unmatched at any time in history. The 2030 Agenda for Sustainable Development established by the United Nations has established 17 sustainable development goals. To achieve these goals, corresponding problems and challenges need to be solved, including those from the ecological environment, human health, social governance, and economic development. and many more.
Applying new technologies to these areas is the correct and "good" direction. For example, the combination of artificial intelligence and medical, education, finance, government and people's livelihood, transportation, urban governance, agriculture, energy, environmental protection and other fields can better improve human life and shape a healthy, inclusive and sustainable smart society. Therefore, enterprises cannot only consider financial performance and only pursue economic benefits. They must also shoulder social responsibilities, pursue social benefits, serve good social goals and social well-being, make positive contributions to society, and realize the unity of interests and values. Including conscious and purposeful design, research and development, and application of technology to solve social challenges. (Si Xiao)