Algorithms of Character- Upstart

Upstart is a company that allows its customers to invest in people in return for a percentage of their future profits. In order to choose people who would be good investments, the company uses and algorithm for predicting financial success and willingness to payback their debts and investors. Upstart applicants submit data such as test scores, the college or university they attended, their gpa, salary histories, etc. and then use this to predict their work ethic and willingness to succeed and payback their investors.

The benefits of this algorithm are that it enables people, such as college students, who have not had the time to build credit, get funding for their own loans. Furthermore, the algorithm could be used in the future for other companies in the loan business. An issue that arises with this is that it could also close a door for those who did not do as well in these areas. As the algorithm is supposed to asses willingness to payback, it could close doors for those who did not go to college and obtain a college degree. This factor in question leads to an issue with the criteria the algorithm is based on.

Though this analysis is data-driven, and the algorithm is meant to be objective, it is written by a human. “Even if the facts aren’t biased, design can be,” said Mr. Gu who wrote the algorithm, but who also would not have qualified for an upstart loan using his own algorithm (New York Times). This example shows the flaws and social bias in the algorithm. Furthermore, it conveys that there are multiple different definitions of success, and successful people can come from many different backgrounds such as Mr. Gu who had perfect SAT scores, but dropped out of Yale. One of the biggest names in the technology industry, Steve Jobs, did not finish college and had trouble in education institutions even at an elementary level. He however, was extremely successful and became someone who would have been able to payback an Upstart loan several times over.

Though the algorithm has since been changed, there is a lot of uncertainty over whether or not similar problems will arise. Furthermore, since this is relatively new, in theory it is suppose to be successful, but a great return on this has not been seen as it will take a few years for investors to see a return. The Upstart program allows investors to mentor their beneficiary, and is to be repaid some of the beneficiary’s profit over the course of some years.

The algorithm itself provides an opportunity for those like myself, who are weighed down with college debt, but who have an incentive and will to succeed. However, the program does not fairly provide this opportunity to people who have not done well in standard educational institutions.

Source: http://bits.blogs.nytimes.com/2015/07/26/using-algorithms-to-determine-character/