In this week’s roundup, learn how artificial intelligence is being used to predict whether someone will die from COVID and how it is helping curb traffic accidents in urban environments. Explore the four ways to democratize data science within your organizations and why so many data science projects fail to deliver. And finally, consider who is making sure that AI algorithms aren’t built with racial bias.
By Tamaar Beeri, contributor for JPost.com
A study conducted by the University of Copenhagen Faculty of Science found that AI can predict with up to 90% accuracy if someone is going to die from COVID-19 before they are even infected. Key risk factors like age, BMI, and hypertension were used to formulate the algorithm. This technology could help hospitals take preventative measures and prioritize patients to better manage the pandemic.
By Universitat Oberta de Catalunya Metz, contributing writer for TechXplore.com
A research project at the Universitat Oberta de Catalunya (UOC) is harnessing AI to make decisions that will make cities safer for pedestrians and drivers. The researchers looked into the correlation between the complexity of urban areas and the likelihood of an accident happening there. The data gathered can be used to train neural networks to detect probable hazards and work out patterns associated with this high-risk potential, aiding traffic authorities in reducing accidents.
by Thomas C. Redman and Thomas H. Davenport, contributing writers for HBR.org
Organizations often leave data to a team of data scientists and focus efforts where there is lots of data. However, using data more strategically and broadly across the organization provides a higher chance for a successful data science transformation. There should be an effort to get everyone involved in data science and tap into it to inform big swing decisions. Here are four ways to make data science more strategic and democratic in your organization.
By Mayur P. Joshi, Ning Su, Robert D. Austin, and Anand K. Sundaram, contributing writers for SloanReview.MIT.edu
While more companies are embracing data science as a function, many have not been able to consistently derive value from their big data, AI, and ML investments. And there is a big cost to this. Evidence suggests that the gap is widening between organizations successfully gaining value from data science and those struggling to do so. Here are five mistakes that companies make when implementing data science projects and how to avoid them.
By Cade Metz, contributing writer for NYTimes.com
Timnit Gebru was one of two AI experts recently pushed out of Google. Gebru has been dedicated to promoting ‘ethical AI’ throughout her career, specifically addressing racial bias that is built into algorithms since the teams creating them are typically not diverse. This article dives into the reasons why this work is critical, citing tangible examples in which AI got it very wrong exhibiting the need for ethical AI. And with two key players forced out of the conversation, it begs the question: who is making sure AI technology isn’t racist?
Did you see an interesting article in the last week? Share it with us! Send it to astuttle [at] lityx.com.