After three development waves，Artificial intelligence has penetrated into human production、All aspects of life，In manufacturing、Traffic、Education、Medical、Home、Finance and other fields are widely used。The development and application of artificial intelligence brings many benefits，But it also caused many problems。For example，Security、unemployment、Privacy Invasion、Algorithm discrimination, etc.。Facing possible risks，"Responsibility" has become an inevitable requirement for the development of artificial intelligence。June 2019，China National New Generation Artificial Intelligence Governance Professional Committee released the "New Generation of Artificial Intelligence Governance Principles -Artificial Intelligence to Develop Responsible"，Determine "responsibility" as an important principle of artificial intelligence governance。Documents on artificial intelligence released from the current domestic and international scope、Study and practical research in specifications and related fields，Artificial intelligence with responsible development has become a wide range of consensus。Responsible artificial intelligence faces two practical problems: Who is responsible? How to be responsible?，Who will bear the responsibility of artificial intelligence; the second question points to the way to develop responsible artificial intelligence，How to fulfill the responsible subject。

　　Bet365 lotto review

Multiple applications in artificial intelligence，At present, there have been actual responsibility issues。For example，In the field of autonomous driving，Too many traffic accidents caused by artificial intelligence。March 2018，A car accident caused by an Uber -driving car in Arizona, USA，Caused a passerby to die。In the after -after investigation and report，There is a certain controversy on the determination of responsibility。bet365 Play online games Reports to propose，The accident should be attributed to software defects，It is also reported that it is the responsibility of pedestrians rather than the responsibility of the vehicle。The problem led to this case is: Who is responsible for the responsibility of traffic accidents in the field of autonomous driving? Is the government in the autonomous driving bill、Autonomous driving program programmer、Auto Manufacturing Company，Still Uber、User、Pedestrian?

Another example，In the medical field，There are also responsibility issues caused by artificial intelligence applications。Sarah Jabri, a schoral scholar at Constantz University in Germany, introduced a related case in the study: Optical Coherence Tomography，OCT) technology can realize the living three -dimensional high -resolution imaging of eye tissues such as retina，The diagnosis and treatment of eye diseases such as retinopathy。In clinical practice，Few experts can correctly evaluate OCT，Therefore, the delay of treatment often causes treatment。For this situation，Artificial intelligence experts have developed a deep learning algorithm that uses computer -assisted diagnosis，Can be able to evaluate the diagnosis made by OCT，and classify patients based on the urgency of treatment，to reduce the waiting time of patients with severe illnesses。Test indicates that the artificial intelligence technology has reached or even exceeds the performance of human experts，and show clinical applicability。In this case，When artificial intelligence is used for the determination and actual diagnosis of patient priority treatment order，and will affect or even determine the success of treatment，Corresponding ethics and responsibilities issues will occur。How much should a doctor rely on artificial intelligence technology? The diagnosis error caused by this、Who should be responsible for medical accidents? Is the subject of supervision、R & D staff or clinician?

Traditional transportation driving fields have a set of rules about the identification and division of responsibility; in the field of traditional Bet365 app download medical care，The main relationship is the relationship between doctors and patients，Its responsibility is relatively easy to define。But it can be seen from the above two cases，Intervention due to artificial intelligence factors，The definition of the responsible subject and how should the responsible subject be responsible to become more complicated。

　Who will be responsible

What needs to be clear is，Artificial intelligence cannot be used as a responsible subject at the current level of development。Although there is still a controversy about whether artificial intelligence has the main position，But now artificial intelligence is used as technical artificial material，In essence, it is still a tool made by humans to make a certain goal。Artificial Intelligence based on data and algorithms to make a judgment and make a choice，It has automatic but not autonomy，With the ability of action but not awareness、Thinking ability，Therefore, it cannot be the subject of responsibility。

Therefore，When discussing the responsibility of artificial intelligence，In essence, the human responsibility of mankind。In the "Cyber ​​Security Standard Practice Guide -Guidelines for Artificial Intelligence Ethical Security Risk Prevention" released by the Secretariat of the National Information Security Standardization Technical Committee in January 2021，Master four responsible subjects: Research Developers、Design Makers、Deployment Applicants and users。In the "New Generation of Artificial Intelligence Governance Principles -Artificial Intelligence of Developing Responsible"，The principle of sharing responsibility was proposed，Involved "R & D、Users and users "several subjects。Although the responsible subject defined in different documents is not exactly the same，But what can be clear is，The responsible subject of artificial intelligence is not a single group，but the subject of diverse behavior。

Due to too many behaviors, the subject participated in the development of artificial intelligence technology、Manufacturing and use，It may cause specific responsibilities to be difficult to determine and push the subject of responsibility。As Scholar bet365 live casino games Katinka Waelbers, a scholar at Mastricht University in the Netherlands，"Natural scientists often claim that they are just developing knowledge，The actual application is not within the scope of their responsibility。Engineers and engineering companies believe that their equipment is no problem，with this to shirk responsibility。In their opinion，The use of technology (or abuse) is the root cause of the problem。And users tend to claim that they lack the ability: how can ordinary consumers know the impact of technology? Even if you have knowledge，What else can I do? They often point out that the government should bear responsibility。But，Actions of Western government agencies often lag behind technology development "。This dilemma existed on the issue of artificial intelligence liability，The second question will be answered by the responsible artificial intelligence: how to be responsible?

　How to be responsible

How to solve the question of how to be responsible，Each responsible subject needs to be clear and assume corresponding responsibility。Comprehensive consideration of the definition of the responsible subject in each document，The subject of responsibility can be divided into regulators、R & D、Manufacturer and user。

Supervisor。This type of subject mainly refers to government departments，It also includes non -governmental agencies participating in artificial intelligence supervision、Organization and group。From the perspective of the current world，Government agencies of major countries have released the outline of the artificial intelligence development report，Formulated the corresponding national policy and development strategy，Ethical supervision of artificial intelligence development。In our country，The State Council issued the "New Generation Artificial Intelligence Development Plan" in July 2017，National Standardization Management Committee、Central Network Information Office、National Development Reform Commission、Department of Technology、Five departments of the Ministry of Industry and Information Technology jointly issued the "Guidelines bet365 Play online games for the Construction of the National New Generation Artificial Intelligence Standard System"，For the key development direction of artificial intelligence、The construction of the technical research standard system for top -level design，Provide a guidance scheme。Artificial intelligence is developing rapidly，Government and related organization groups need to accurately grasp the general trend of artificial intelligence development，Formulate a more forward -looking standard system for the development of artificial intelligence、Management specifications，Continue to follow up the latest dynamics in the field，and R & D、Manufacturing、Use and other links to monitor and manage。

R & D person。This type of subject refers。The ethical principles to be followed by artificial intelligence、Value and standards need to be put in the research and development link。Due to the behavior and function of artificial intelligence, it is determined by the algorithm，R & D subjects need to fully consider the risks that their R & D results may cause，Clear human values ​​and ethical rules that artificial intelligence needs to follow，If fair、Security、transparent and diversity, etc.。The principle of transparency is the basic requirements for the development of "responsible" artificial intelligence。Because，In the case of "Black Box"，It is difficult to take responsibility for the subject。Another，Artificial intelligence technology is not independent of society，R & D subjects need to consider the social and cultural environment of artificial intelligence development，Consider the priority level of value settings，If my country requires the attention of collective interests in the development of artificial intelligence、Social Stability, etc.。

Manufacturer。This type of subject refers to the organization or individual engaged in the manufacturing of artificial intelligence products。The "Artificial Intelligence Standardized White Paper (2018 Edition)" released by the National Standardization Management Committee defines artificial intelligence products as "integrated technological achievements in the field of artificial intelligence、Smart terminal，and divide artificial intelligence products into smart robots、Smart Carrier Tool、Smart terminal、Natural language processing、Computer vision Bet365 lotto review and other types。In the production of these artificial intelligence products，Production subjects should strictly abide by the regulations and industry standards in the field of artificial intelligence，Ensure that the manufactured products meet the interests of society and humanity; fully evaluate the possible risks in the use of the product，Set the corresponding emergency mechanism; ensure the explanatoryness and explanability of the product，and provide users with product functions、Clear information in usage and risk response。

User。This type of subject refers to the organization or individual of artificial intelligence technology products in various cases。Use the subject should have the basic knowledge and ability required by artificial intelligence，Can be correct、Specific use of technical products，and cannot use artificial intelligence to implement any behaviors that harm others and society。To use the subject to better assume its main responsibility，It is necessary to popularize artificial intelligence basic knowledge education。This is not only the basic requirements of the responsibility of using the subject，Also in line with other ethical principles (such as fairness、Anti -discrimination, etc.) requirements。

The above analysis of the responsibility of the responsible subject and the subjects should be assumed。In specific practice，Under the guidance of the overall framework and basic principles of artificial intelligence development，According to the development stage of artificial intelligence in different fields、Application scenario、Use features and other categories to perform more detailed discussions and research。For example，In the field of autonomous driving，Related research pointed out that automakers should be the main responsible party; in the medical field，Except for the subjects mentioned above，It also involves professional evaluation organizations in medical institutions。

Artificial intelligence of responsible for development，Need to build the principles of the corresponding ethical governance。But we can't just stay in the principle、Active discussion level，It also needs the cooperation and participation of each responsible subject。Only parties form a joint force，Responsible artificial intelligence may be realized。

　(This article is the National Social Science Fund major project "Research on Artificial Intelligence Ethics Risk Prevention" (20&ZD041) Staged results)

(Author Unit: College of Marxism University in Tongji University)