When Data Science Goes Wrong

In this video, I talk about what happens when data science or machine learning goes wrong. I get many of the examples from the book "Weapons of Math Destruction" by Cathy O'Neil. Book: https://amzn.to/2w1eJVy Even well intentioned data science models can have negative consequences. Take the college ranking system for example. This system was designed to evaluate the quality of schools, and evaluated them across a host of different metrics. One thing that was not considered was the cost of attendance. Universities spent billions of dollars to improve their rankings, and this was reflected in tuition rates. This model lacked an extremely important input and thus caused a major incentive problem. Another example is social media. Companies are doing their best to keep you on their platforms with notifications, likes, etc. This can have an addictive impact as well. The algorithms are doing their job, but we are getting addicted to our phones and devices. A third example comes from the news. News companies want as much interaction with their articles as possible. Unfortunately, the articles that get the most engagement are the ones that are controversial and sensationalist. These machine learning models promote these types of articles because that its what they are trained to do. It is our responsibility as data scientists to make sure our models are used for good and don't have negative unintended consequences. You can do this by getting user feedback, re-training and evaluating, and testing your models continuously on real world data. Dangers of Social Media: https://www.bbc.com/news/technology-44640959 https://journals.sagepub.com/doi/abs/10.1177/0894439316660340 #DataScience #KenJee ⭕ Subscribe: https://www.youtube.com/c/kenjee1?sub_confirmation=1 🎙 Listen to My Podcast: https://www.youtube.com/c/KensNearestNeighborsPodcast 🕸 Check out My Website - https://kennethjee.com/ ✍️Sign up for My Newsletter - https://www.kennethjee.com/newsletter 📚 Books and Pr