Reading Response 2: Rheingold Annotations

In  Howard Rheingold’s chapter 2 of his book he talks about being smart on the internet with the title, ‘Crap Detection 101: How to Find What You Need to Know, and How to Decide If It’s True’. In this text he talks about the steps needed to differentiate whether the information gathered is valid or not. He also lets his readers know why it is important to do advanced search rather than trusting in your initial search. In The Economist, a research took placed at Stanford University to determine whether Artificial Intelligence could determine the sexual preference of an individual based on their faceprint. Throughout this article, they talked about the steps they took to determine if the AI could actually determine if a person was gay. In this study when it was a small controlled group Michal Kosinski and Yilun were able to get statistics closer in range of accuracy with men rather than women. Their explanation was because the “fetuses develop in the womb, they are exposed to various levels of hormones, in particular testosterone” (The Economist, 2017). In the end of the article, the author stated that “the study has limitations” (The Economist, 2017), this is because when the group is controlled they can manipulate how many men in the study are gay to correlate with real-world statistics, but on a larger scale and 1,000 men were chosen at random the system failed to correctly choose the gay men. This can cause problems in search engines and the lives of others. “Spouses might seek to know what sexuality-inferring software says about their partner”, (The Economist, 2017) and based off of the  AI the information can’t always be accurate nor valid.

Here is a picture of Agathe Mougin, she can be mistaken as male and this could trigger the AI to assume she is gay.

Going back to Rheingold’s explanation of crap detection, if The Economist only told the readers the capability of the AI rather than including the limitations it would be considered crap. Although, you should still complete more searches on the AI to see if any other researchers have taken place with similar studies. After reading this I took it upon myself to dig a little deeper, I found that androgynous people share the looks of both male and female. If an androgynous man was apart of this study, he would most likely be considered gay based on the AI. This may not be the case and it is marvelous that Michal and Yilun decided to test this system out on a larger random scale of people before just completing their studies and telling the capabilities of it. I found another compelling article I found while searching the names of the researchers who studied at Stanford, it was written with a different tone compared to The Economist. ‘Why Stanford Researchers Tried to Create a ‘Gaydar’ Machine’ written by Heather Murphy, describes the trials and tribulations faced after the study. Michal was the leader of the study and received many death threats due to this.

In another one of his studies, Ms. Murphy described in her article of the researchers’ links back to Rheingold’s article. “All of our online likes, Dr. Kosinski said, have left us vulnerable to microtargeting by political candidates, companies and others with nefarious intentions” (Murphy, 2017). “Search engines sell sponsored links that appear on the top or side of the page of links displayed in response to a search query” (Rheingold, 2012). Usually, this paid sponsorship is based on the users’ recent searches and likes, which will make it more likely for the user to click on the paid link. “Whenever someone clicks on a sponsored link, a small amount of money goes to the search engine provider. Those clicks add up to billions of dollars each year” (Rheingold, 2012). It is very ironic, but since I’m an avid online shopper I experience this a lot. Something I left in my cart after completing my transaction, will continuously pop up on the side of my computer screen weeks later. The internet is a valuable informative tool, as long as we use it correctly and Rheingold gave us many ways to detect the good from the bad.


Works Cited

“Advances in AI Are Used to Spot Signs of Sexuality.” The Economist, The Economist Newspaper, 9 Sept. 2017,

Murphy, Heather. “Why Stanford Researchers Tried to Create a ‘Gaydar’ Machine.” The New York Times, The New York Times, 9 Oct. 2017,

Rheingold, Howard. “Rheingold-Net-Smart-Ch-2_compatible.Pdf.” Google Drive, Google, 2012,