Friday, November 1, 2019

Artificial Intelligence Not All Rosy

In the Peace Weekly's Current Events Diagnosis column a college professor gives us her ideas on the world of artificial intelligence. Cars without a driver, robots, waiting at home to help with the cooking and cleaning. All we need to do is tell our speakers to play our favorite movie or music and it will be done. The washing machine will check the amount of detergent and when it runs out, automatically order a replacement.

AI will not only make our lives easier but according to TrendWatch com, a leading trend research firm: health will be a big area of concern. With knowledge of the consumer's eating habits, DNA, blood test information, they will be able to communicate medical information using apps and other devices. 

An example would be checking your eating menu and your blood pressure, recommending a two-hour walk after dinner and vitamin drinks. TrendWatch com expects AI will manage human health with numerical indicators in the manner engineers solve their problems.

A future in which AI will manage the efficiency of production, the convenience of life and our human health is not all rosy. In science fiction films, many concerns about the ethics and fairness of AI are ongoing. 

Evidence suggests that AI's current algorithms are not fair or ethical, even if the AI doesn't become a monster many problems are expected. Facial recognition system technology, for example, is said to reflect prejudice against race. White men are trusted more than the black and hiring algorithms are known in certain cases to be unfavorable towards women.

According to a survey of 27,000 consumers by a US company, 97% of consumers are concerned about the ethical use of technology, and 94% believe that the government should intervene. Fortunately, several technologies have been developed and applied at the private level that first check the fairness and ethics of AI algorithms.

Facebook uses this technology to check fairness on its platform to select companies that discriminate by age and gender in job advertisements. To increase the transparency of AI services, IBM provides program tools that identify which individuals or groups will be victims of prejudice, and the reason for unfair results, and provides developers with information to correct problems.

Competence and morality of AI are equally important. A thorough review of the moral issues that arise when technology is applied should be undertaken. Humans are more than just quantified objects that AI recognizes. If the artificial intelligence that manages human life and health is mistaken, its ramifications are huge. The human error we can forgive AI errors are not forgiven.

The big problem with AI will always be the human element. Mistakes that go in will have a ripple effect when they come out. Artificial Intelligence system will only be as good as the data that goes in.