Although technology workers are important stakeholders of artificial intelligence development and deployment，But few people care about their understanding of ethical issues in the development of artificial intelligence。In view of this，Jude Brown, the person in charge of the Gender Research Center of Cambridge University in the United Kingdom, uses empirical research，Data from science and technology workers from a large artificial intelligence company，To better understand the people’s views on artificial intelligence ethics issues。The difference between similar research in the past is，This study clearly uses feminism methods to analyze problems，Pay particular attention to how to make the sound and narrative of technology workers leading the research direction。Recent，This research was published in "Big Data & Society) Journal.

　　bet365 live casino games It is important

The current economic environment changes rapidly，Brown and others think，To ensure that the development of artificial intelligence is in line with ethical specifications，Listening to the needs and concerns of science and technology workers are important。But，As shown in the institute，When the enterprise formulates artificial intelligence ethics strategy，Rarely soliciting opinions from employees，This is mainly based on the assumptions of experts such as ethical scholars and other experts with epistemology。This exclusion phenomenon lacks STEM (science、Technology、Engineering、Mathematics) The background of science and technology workers is particularly obvious on the body，For example, a lawyer、Marketing staff or human resources professionals are often deeply affected by new technologies or new products，But their opinions and feedback are rarely collected or Bet365 lotto review included in the product development and design process。

Brown and others conducted in -depth interviews with more than 60 employees of a large artificial intelligence company，These employees are distributed in artificial intelligence、Data science、Legal services and human resources and other fields。Researchers follow the methodology of feminist theorist Sara Ahmed (Sara Ahmed)，Allows the respondents to tell the story of their own artificial intelligence ethics，and use the research of feminist political theory and feminist science and technology to explain。The use of feminism methods and the attention of personal stories，Enterprise as a hypothesis complicated by a single entity that has a single understanding of artificial intelligence ethics。That is to say，This study emphasizes different ethical viewpoints of employees，and the importance of understanding what artificial intelligence can do and what to do。The point of view of technology workers with different forms of knowledge，This research shows in a company with well -known global brands and highly developed internal corporate culture，The significance and importance of artificial intelligence ethics will undergo major changes between individuals and groups。

Enterprise、Government、International Organization、academic institutions、Civil organizations, etc. have regarded the concept of artificial intelligence ethics as an important way to guide artificial intelligence development and deployment。Many artificial intelligence ethics frameworks have been formulated at the international level，For example, the EU's "Algorithm accountability system and transparency control framework"。Although different artificial intelligence ethics projects include their own unique political agenda，and related to different stakeholders，But there are still some common concern points，For example, justice and fairness。But Brown and others think，Listening to the story and point of view of technology workers，You can see a more complicated picture，This proves how some common bet365 live casino games ethical terms are not enough to achieve effective change。

　Tell the story of artificial intelligence

When the technology worker tells the story of personal intelligence about artificial intelligence，Researchers focus on three topics。First，"Bias" (bias) What does it mean for technology workers，How do they believe that this concept is compatible with their work。2，Science and technology workers for the development and diversity of artificial intelligence、Equality and tolerance issues and the relationship between corporate diversity initiative。third，Fusion of the new artificial intelligence system with the existing "reserved system"，and the process of maintaining the artificial intelligence system，Challenges brought to the development of ethical artificial intelligence。

First，Artificial intelligence ethics policy and discourse highlight the prejudice problem in the artificial intelligence system，and frequently choose to prefer to make a strategy of reducing prejudice。But，This study indicates，The definition of prejudice of science and technology workers is not consistent，And the views they may have prejudice on which systems and data they may have prejudice。Science and technology workers can learn more about structural inequality how to affect the data they use，You can benefit from it。Inspired by feminism theory，Researchers encourage artificial intelligence ethicalists and technicians to think about the concept of artificial intelligence "partiality"，That is the local perspective adopted by each machine learning model。Artificial intelligence "favoritism" originated from the situation knowledge in feminist theory，This theory claims that knowledge is specific to the situation，instead of common。So，Data should be understood as an unbalanced experience in the background of inequality，and can only provide any point of view of any given phenomenon。By this way，Brown et al.，The limitations of the artificial intelligence system should be a featured feature defined，This may Bet365 lotto review be better than advocating "eliminating prejudice" and other attempts to restore the system to a neutral state，More conducive to the formation of artificial intelligence ethics。

Next，Just as the number of artificial intelligence ethics frameworks shown，The diversity or lack of diversity of the product team will affect the design of artificial intelligence、Development and deployment。But，The population structure of the artificial intelligence industry is still serious at homogeneity，So that some scholars proposed，Artificial intelligence is suffering from the "diversity crisis"。such as，Only 26%of the artificial intelligence practitioners in the United States are women。In order to correct the lack of women's number of women in the artificial intelligence industry，Artificial intelligence ethics initiative emphasizes the need to increase the diversity and representativeness of the design team。But most respondents did not see any relationship between the development of labor and artificial intelligence ethics。Because in practice，If the diversity is considered a choice item rather than a continuous process，It is difficult for technology workers to understand why diversity is so important，and its core role in creating ethical artificial intelligence。Brown et al.，Diversity、Equality and tolerance plan and artificial intelligence ethics target shall establish a direct contact，Diversity issues should surpass the problem of insufficient representative individuals，It should also be noted that the occlusion nature of different forms of situation knowledge。Government and other institutions can surpass the narrative of the "diversity crisis" of artificial intelligence talents，and start to better develop the development of artificial intelligence、Function and social impact is placed in the specific background。such as，Some companies are not only committed to improving the diversity of design teams，And advocates fully realizing the vision of "design justice"，The voice of those who are usually marginalized during the bet365 live casino games technical design and development process。

Last，Research indicates，As a large computer system that has been used for a long time and has many years of software development and data accumulation，"Reserve System" is one of the main obstacles to develop artificial intelligence ethics。The ethical supervision of retaining the system may need to study potential ethical issues that are decentralized or reconstructed by data。Some data scientists are very alienated with the background and source of the data they are processing，Therefore, it is called "data concentration stranger"。The distance between data scientists and data collection and maintenance，exacerbated the risk of producing artificial intelligence products that are not representative。These concerns highlight a fact，The implementation of ethical artificial intelligence，Must require long -term and complicated efforts，Can't simply achieve it through the one -time "prejudice inspection" in the development process。"Maintenance" is a neglected field in artificial intelligence research，Research on Feminist Political Economics has long proved the importance of maintenance，How this job is underestimated。Brown and others think，The same situation is also applicable to the technical system。Some interviewees think，Debugging and other maintenance work is boring and unpopular。Despite this，They emphasize，How to monitor and calibrate the artificial intelligence system，to ensure its long -term fairness and consistent performance，The long -term practice of artificial intelligence ethics is important。

　　Dialectical view of the situation

The core of the feminist method used by this study is the concept of contextual knowledge，This concept emphasizes that knowledge is scenario，It is shaped by the specific conditions generated by it。So，Although some artificial intelligence ethicsists emphasize moral reasoning，and try to transplant these principles into the industry background，But Brown and others think，Bet365 lotto review Exploring the contextual point of view of those who are engaged in the artificial intelligence field are important，The specific method is to explore the situation knowledge of science and technology workers through open dialogue。Researchers do not rely on more limited survey categories (for example, determine the answers of the respondent in advance according to the question set) or more stringent interview and answer structure based on questionnaires )，Because some of the viewers' views will be caught by the structured method because of irrelevant or too trivial methods。Open dialogue can encourage science and technology workers to explore their own production、Test、Feeling and expectations of using and sales technology，To better understand the conditions of people's decisions。

Of course，Researchers do not treat the story told by science and technology workers without criticism，There is no assumption that they contain the only "truth" about artificial intelligence，and the challenges facing artificial intelligence at the ethical level of science and technology workers，Instead of looking at the story told by employees dialectically，At the same time, realize that their experience may be blurred or covered with the real harm that their company's products may produce。By using feminist methods，Researchers created a more detailed picture，Drawing the views on ethical issues when developing and maintaining the artificial intelligence system。Of course，There is also insufficient research on this research。such as，Researchers pay attention to a large technology company headquartered in developed countries，This obviously cannot represent the entire technology industry。