The rapid development of artificial intelligence technology and its widely used application in various fields of society，Bring a series of risk challenges。How to make moral governance on artificial intelligence ethics risks，An important question that people have to think carefully。According to the traditional responsibility theory，The moral responsibility caused by a certain technology must be assigned to the designer of the technology based on the size of the negligence、Manufacturer、User、Supervisor and other human subjects。But because artificial intelligence has significant automation features，If it is excluded from moral responsibility，It may cause "responsibility empty field"。So，Establish a complete artificial intelligence moral governance system，It must solve the problem faced by the technology when the moral responsibility belongs。

　　bet365 live casino games

The first problem of the belonging to the responsibility of artificial intelligence is whether artificial intelligence is eligible to become a responsible actor (or consequences).，Not just a technical tool controlled by humans。This requires answering whether artificial intelligence has the minimum moral cognitive ability and moral autonomy for responsibility。Artificial intelligence is a "human -like intelligence" that uses computers to simulate and implement。It converts external environment information to input data，and process the input data by the characteristics generated from the data，Implement data output。The input and output process of artificial intelligence，It can be understood as a process of perceiving external environment information and taking specific action based on perception information。bet365 live casino games Artificial Intelligence's behavioral choice is based on the capable response based on the judgment of the external environment。Comparison of comparison，Traditional computer programs complete the input and output of the data according to the designed function，Its response to the external environment is passive。other，Artificial intelligence can also be further inserted into the moral algorithm，The logical rules of human understanding of the importance of morality into artificial intelligence information processing and decision -making，Make it abide by social moral requirements in the forward -looking sense，Make a responsible behavior decision。

The particularity of artificial intelligence lies in its unique learning algorithm。Multi -layer treatment，Low -level features The feature representation transformed into high -level features，Make artificial intelligence have the ability to complete complex tasks，You can imitate specific human behavior。such as，AlphaGo through the deep learning of neural networks and the enhanced learning of repeated self -game，Not only improves its prediction ability in the Go game，also found new game rules，New strategy。Here，Learning algorithm is pre -stipulated by human designers，But the behavior made by AlphaGo through machine learning is not determined by human designer。Artificial intelligence has a certain but limited autonomy、Interactivity and action ability，You can make behavioral decisions independently in a specific environment，Therefore, you can make moral blame for it。Acknowledging artificial intelligence can be attributed to the minimum limit，It affirmed its limited energy position in the moral governance system。

　　Interest responsibility

The second problem is artificial intelligence follow mathematics、Physics and other natural laws，Lack of "feeling that can bear responsibility"。The subjective feeling of responsibility is first manifested as the subjective intention of the actor，That is the perception Bet365 app download of the meaning of the behavior of the actor，Hope or want to act in some way。Although the subjective intention does affect people's moral condemnation or punishment for actors，But it is not a necessary condition for responsibility to belong。From the perspective of moral practice，Subjective intention as an inherent state of the actor，It is difficult to make accurate evaluation of it。And，The purpose of people's responsibility is，Back traceability evaluation through the consequences of damage，To determine the cause of the actor who caused this consequences，Regardless of whether the actor has subjective intentions。German scholar Outterde Hye take Oedipus as an example，Explain that this kind of non -fault liability is common in prehistoric society: responsibility is objective，Even if the actor does not have any subjective fault，You also need to bear moral responsibility。Responsibility to artificial intelligence in accordance with the principle of objectivity，Mainly examine the causal relationship between damage and consequences and consequences，No need to pay more attention to the subjective intention of the perpetrator。

The subjective feelings of responsibility are also manifested as the self -blame caused by the actor for their behavior、Reactive attitude such as remorse，Therefore, admit that you can act in other ways。Artificial intelligence lacks feelings，It's hard to imagine that it will produce any form of pain or guilt，So，​​Is it meaningful to belong to the responsibility of artificial intelligence and implement moral punishment? As far as the reactive attitude is concerned，The reactivity of others instead of the self -reactive attitude of the perpetrator is the driving force for responsibility。The blame and resentment of others are not only the value evaluation of the value of an event (or consequence)，It is also the reaction made by the actor on the basis of evaluation。Responsibility and resentment allows Bet365 app download the actor to propose a reason to avoid being blamed。but at the same time，If the cause and effect responsibility of an event (or consequence) of an event (or consequences) is extremely significant，When it triggers the strong blame and resentment response of others，Others will refuse artificial intelligence as a reason for its exemption。In terms of moral punishment，Punishment as "a natural revenge on the wrong past error"，Gradually lost its moral legitimacy in modern society。Moral responsibility does not necessarily require a moral punishment for actors，Instead, the actor is required to face up to、Correction or compensate for your moral fault，Try to avoid similar moral faults at the same time。That's for this，It is feasible to require artificial intelligence to bear moral responsibility。

　　Common responsibility

The third problem is to belong to the responsibility of artificial intelligence，Whether it will weaken human moral responsibility or threatens human responsibility subject status。The answer is negative。The correct attribution to artificial intelligence helps to clarify the boundary of the responsibility of human and artificial intelligence，so as to strengthen human moral responsibility。Human designer、Manufacturer、Users, etc. participated in the theoretical innovation of artificial intelligence、Technology R & D、Manufacturing、Practical application and other links，Responsibilities of internal knowledge、Responsibilities of scientific experiments and the responsibility to put science into practice。These diverse subjects in this technology design -production -application chain division of labor cooperation，As a result, a chain of responsibility: No separate links to bear separate responsibilities，Each link is part of responsibility，And this part of responsibility has to be linked to the responsibility of the entire behavior of the entire behavior。As He Fei said，Who can create some prerequisites for others，He will have a bet365 live casino games common responsibility for this matter。On the contrary，If they do not see this relationship，or I can't see this relationship，The requirements of the ownership of responsibility，For this reason, it should be set up an organization that can supervise the performance of responsibility。The agency is for human designers、The responsibilities of manufacturers and users are supervised and managed。Human actors cannot avoid moral responsibility，You need to bear the corresponding compensation or compensation liability。

In moral practice，Human actors authorize some morals to artificial intelligence，If the intelligent search engine is allowed to screen and block the sensitive words。Authorized behavior means that some moral capital is transferred to artificial intelligence by human performers，Perform the latter to perform a certain social responsibility instead of human actors。Does the moral authorization of artificial intelligence mean the authorization of moral responsibility? This depends on whether the authorization is authorized or the coordinates of moral behavior (that is, only providing feedback information for human moral decision -making)。Authorization of moral behavior，It means that moral good and evil and responsibility are distributed in the multi -intelligent system system composed of artificial intelligence and humans。At this time，Artificial Intelligence Replacement Human actors participate in the real division of labor，An important part of becoming a social system，Therefore，Moral responsibility that needs to be assigned to the responsibility of the society，That is distributed moral responsibility。Authorization of moral behavior coordinates，It still uses human as a moral responsibility and realizes human moral intentions through artificial intelligence，Artificial intelligence only needs to bear expansion moral responsibility。Regardless of distribution or expansion moral responsibility，None。

In short，The key to artificial intelligence and moral bet365 best casino games governance is based on a new open liability theory to belong to artificial intelligence。First，The moral responsibility of artificial intelligence is no fault liability。The purpose of the ownership of artificial intelligence responsibilities is to correct the correct attribution of moral behavior and its consequences in accordance with the principles of objectivity。Especially when multi -acting subjects participate in a certain behavior decision，Correct attribution helps to correct errors in time。Next，Artificial intelligence can only bear the moral responsibility that is compatible with its intelligent level and autonomy。The moral responsibility of the current widely used weak artificial intelligence is closely related to its ability to extract the characteristics from big data。The ownership of its responsibility cannot be neat，It needs to be treated according to its automation level。Moral responsibility beyond its ability to capacity still needs to be borne by human actors。again，The responsibility of artificial intelligence is required to conduct a retrospective experiment on the artificial intelligence algorithm until the algorithm defect or program vulnerability。The way of artificial intelligence assumes responsibility includes an optimization algorithm、Program upgrade、Passing or permanently stop using。Corresponding compensation or liability for compensation must also be borne by human perpetrators。Last，Human designers should supervise and inspire artificial intelligence to continuously accumulate moral experience，Transform moral experience into moral algorithms or new feature quantities，Try to minimize the unexpected moral situation and dilemma。In the computer field，The situation of modifying vulnerabilities through programs is very common，This also requires a reasonable understanding and summary of the moral errors of artificial intelligence。

　　(This article is the "Research on Responsibility Ethics of Artificial Intelligence" (19BZX109) in the general project of the National Social Science Foundation of the National Social Science Fund

(Author Unit: Marxism College of Ningbo University)