Online Advertising and Machine Learning

In chapter 4 of “Weapons of Math Destruction” by Cathy O’Neil, this chapter generally talks about online advertising. And continuing the topic from the previous chapter, O’Neil talks about how for-profit colleges uses online advertising to target the more vulnerable side of the population. One of the example she took is about the advertisement made by University of Phoenix on Google, “While spending more than $50 million on Google ads alone, the University of Phoenix targeted poor people with the bait of upward mobility. Its come-on carried the underlying criticism that the struggling classes weren’t doing enough to improve their lives. And it worked.” In this example, result shows that the college targeted the poor families, claiming to give them opportunity to improve their lives, and during the process, for-profit enrollment tripled, and the industry now accounts for 11 percent of the country’s college and university students. I do agree on her approach, where the ads “pinpoint people in great need and sell them false or overpriced promises.”, this shows how using WMD could benefit the companies and harm general public.

Later on in the chapter, O’Neil also mentioned about how Facebook and Google ads are WMDs. She said that each of these platforms allows advertisers to segment their target populations in meticulous detail. The ads also have the ability to learn from campaigns, which means that they energetically learn from new data. And because of their act of very large-scale data, they can further strengthen their ability to learn. I found this topic quite interesting. These two example shows us how given a large scale of data, the WMD would learn and adept, and eventually they would be personalized for everyone. They may not bring the biggest benefit to our life, but they certainly brings some convenience such as predicting what we want to buy in the near future.