Getting Hired by an AI
Going to school and getting a degree is considered the best way to not fall into unemployment or poverty. Although, recently, many do not totally agree with that assumption. But, for the most part, a degree is associated with wealth and happiness to a large degree. That’s why parents thrive to send their children to school and encourage them to earn the highest degree in the field they like. People spend a lot of money, energy, and life in schools because they are certain that they will be reward a lot more and that they won’t need to do as much effort after school to reach what is happiness for them.
In the majority of cases, which keeps decreasing, hard work during school and a good degree leads to being easily hired and getting a higher-than-average salary. There is a smaller, but growing, number of cases where school nor talent makes it through employment, and we often wonder why. The answer is sometimes luck or God’s will for the ones seeking employment. There also is the discriminatory factor (race, sex, religion, country, etc.) that people take into consideration. This brings skepticism in the hard work plus degree equal quickly employed plus good salary equation.
A clearer understanding of the hiring process is explained by Cathy O’Neil, in her book, “Weapons of Math Destruction”. The author talks about the computerized models created to make the process of filtering job applicants faster and easier for hiring managers. When those models are not feed with enough good input and feedback data, we end up in a bad perfect model. For instance, the author gives examples of people who were disqualified by models according to ambiguous and not relevant questions they answer in their job applications. Then, the discriminatory factors are added to the model for filtering certain groups of people. Models only do the work they are trained to do. Users define the variables and outcomes and people and companies loose opportunities in the process.
It is very obvious that sex, race, religion, location, etc. are not factors that can determine personality and job performance. There is no country or religion where everyone has the same talents and one personality. All men are not the same and all women are not the same. It is absolutely wrong to use those things are factors for selecting a new employee. The author mentions a model called Gild that doesn’t take some of those irrelevant discriminatory factors into consideration. I believe that all models should follow the same. Not only in the hiring process, but in all the models where the data is not important.
