Gender and Racial Bias in Artificial Intelligence

下面51due教员组为大家整理一篇优秀的代写范文- Gender and Racial Bias in Artificial Intelligence，供大家参考学习。这篇文章讲述的是人工智能的发展引起了公众的广泛关注。人工智能被视为通过解决某些世界上最大的效率低下和挑战来改善人们生活的重要工具（美国政府，2016年）。然而，人工智能也给人类带来了新的挑战。性别分工和人工智能的种族观念都表现出偏见的存在，这种偏见加剧了现有偏见，严重影响了社会和谐。

Gender and Racial Bias in Artificial Intelligence

In recent years, the development of artificial intelligence has attracted much public attention. It has made progress in the areas of speech recognition and image recognition; and it is also applied in people’s daily life gradually, such as speech input, navigation, trip planning, self-driving vehicles, language translation, and playing strategic games (U.S. Government, 2016). Generally, artificial intelligence is viewed as an important tool in improving people’s lives through solving certain world’s greatest inefficiencies and challenges (U.S. Government, 2016). However, artificial intelligence also brings new challenges to human beings. The sexual division of labor and racial concepts of artificial intelligence all show the existence of biases, which reinforce existing biases and seriously influence social harmony.

In the first place, the sexual division of labor indicates the existence of gender bias in artificial intelligence. According to Vicente Ordonez, a computer science professor in university of Virginia, it is more likely for the image-recognition software to associate a picture of a kitchen with women than men. Furthermore, his research shows that, coaching and shooting are linked to men while washing and shopping are tied to women (Simonite, 2017). When it comes to the artificial intelligence robots, robots who engage in service-oriented jobs tend to be females, and robots who work on complex and high end jobs tend to be males. Apple’s Siri, Google’s Now, Amazon’s Alexa and Microsoft’s Cortana are all females, and they all engage in service-oriented jobs. At the same time, a high end artificial intelligence system is called as Watson by IBM and an artificial intelligence lawyer is called as Ross. Consequently, the sexual division of labor reveals the gender bias in artificial intelligence.

In the second place, the racial concepts of artificial intelligence also reflects racial bias. Research shows that, the names of people from the Europe and America are more likely to be connected with words such as “gift” and happiness, and the names of African American are commonly related to unpleasant words. Moreover, compared to African American, European Americans are 50 percent more likely to receive an interview invitation than the African American even though their resumes are the same (Devlin, 2017). In facial-analysis programs, the research from Stanford and MIT University indicates that, there is a low error rates in determining the gender of light-skinned men, but the error rates balloons when it comes to determining the gender of darker-skinned women (Hardesty, 2018).

As Joanna Bryson, a computer scientist at the University of Bath argues that, it is very probably that artificial intelligence strengthens existing biases since it cannot resist the biases that it learns from human beings. The data of machine learning algorithm comes from human’s language and behaviors. Nevertheless, human’s bad language awareness and thoughts will provide wrong models and examples for the algorithms. In general, the existence of gender and racial biases shows that, it is necessary for human beings to realize and solve the problem for the sake of maintaining social harmony and proper values.

51due留学教育原创版权郑重声明：原创优秀代写范文源自编辑创作，未经官方许可，网站谢绝转载。对于侵权行为，未经同意的情况下，51Due有权追究法律责任。主要业务有essay代写、assignment代写、paper代写、作业代写服务。

51due为留学生提供最好的作业代写服务，亲们可以进入主页了解和获取更多代写范文提供作业代写服务，详情可以咨询我们的客服QQ：800020041。

51due为留学生提供最好的essay代写服务，亲们可以进入主页了解和获取更多essay代写范文 提供美国作业代写服务,详情可以咨询我们的客服QQ: 800020041。