[AI Ethics] Is It All Good with No Downsides? A few AI ethics questions that we must all ask ourselves

by Pei-Wen Wang

 

AI can assist humans with countless tedious tasks in life, while it also raises ethical issues. For instance, are all applications beneficial to humans?

AI replaces people’s jobs

AI technology is expected to bring the fourth industrial revolution, where jobs that are repetitive can be gradually taken over by machines, which leads to unemployment.

Prejudice and discrimination involving employment and criminal prevention

One of the most pressing and widely discussed issues in AI ethics is the bias in predictive analytics, such as in employment or crime prevention. There was a time when Amazon used an AI algorithm to screen job candidates based on data, which became an incident of hiring bias. The problem results from the gender bias in candidate selections in the past, which leads to the algorithm also tending to choose male candidates for the job.

In March 2019, the New York City Police Department (NYPD) revealed that they had developed Patternizr, an algorithmic machine-learning software that sifts through police data to find patterns and connect similar crimes. It has been in use since December 2016, but it is not designed to track certain crimes, including rapes and homicides. The software was designed to exclude gender and race from its algorithm, which is a big plus compared to previous algorithms. However, eliminating all kinds of biases based on historical outcomes is not standard protocol.

Misinformation & fake News

Another ethical issue involving AI is forgery, where AI is used to process images, videos, and audio files from the original material to alter the intent of others, often in a malicious way.

Consumer privacy & security

Although the government has formulated many regulations to protect consumer privacy, the potential threat is still substantial.

1. Data collection: Most devices used in consumer electronics use AI to collect personal data in order to provide better services. This is a great feature if the user agrees to the use of such information and when the way the data is collected is transparent. On the contrary, if relevant policies are not consistent or transparent, the user experience can be bad.

2. Identification and tracking: Tracking personal information is not safe for domestic violence survivors who want to hide their tracks.

3. Voice and facial recognition: The voice and facial recognition are convenient features, when it comes to unlocking the phone on your smartphone, for instance.

Conclusion

The progress of science and technology has brought convenience to people’s life, but it is also necessary to pay attention to whether such technology is causing a backlash on human beings due to improper use.

分享到社群

vMaker編輯

The share account of vmaker editors. Send your work to us:contact@vmaker.tw