AI Talk: Guides, university and malls

Feb. 21, 2020 / By V. “Juggy” Jagannathan, PhD

This week’s AI Talk…

Guides to building AI systems

AI-based solutions and systems are popping up all around us. How does one ensure these systems are effective? Useful? Safe? Ethical? Tech giants Google and Microsoft have released guidelines on how to build such systems. The guidelines underscore the necessity of having a mental model of user’s needs, focusing more on augmenting user’s skills than on automation:

  • Be transparent and explainable.
  • Gain the user’s trust, which is only possible if the system can explain its reasons for its recommendations.
  • Make sure the training data set does not introduce bias.
  • Develop strategies for recognizing errors in operation, particularly when errors can pose safety concerns.
  • Develop a governance model and ensure compliance with such models. 

All good advice! Now, the question is who is applying these principles in the design of systems? I hope the answer is everyone.

University built in five months

Last weekend I was listening to NPR and was intrigued by a story they were reporting. It was about SRM University, a university that was built in five months in Southeast India, not far from where I grew up! Five months does not sound like a big deal when China can build a hospital in a week to combat the novel coronavirus, but what got my attention was not the fact they built it in five months, but their goals and ambitions and what they have already accomplished. SRM is linked to the Minerva, an innovative startup in U.S. which aims to revolutionize education. Students in SRM use a range of high-tech machinery in class, listening with a headset to online lectures delivered in a traditional classroom setting. Students are randomly picked to answer various quiz questions and class participation is part of the grading. The subjects taught include the latest advances in AI and virtual reality to English literature and drama. An example of a topic for drama? Solar powered cars! A perfectly nerdy topic for nerdy engineering students to create a play about! To bring coding parity and encourage girls to join the program, the girls engage in “Python in Pajamas” coding sessions at night. The university has big dreams of being in a league with MIT and Stanford and are gearing up to meet the insatiable appetite for higher education in a country which has 650 million under 25! Apparently, they need to build 1,000 new universities and 50,000 technical colleges this coming decade to keep up with demand! Seems like a tall order.

Who killed the mall?

Saw this article in the New York Times over the weekend. It provides an excellent commentary on what happened to the retail industry. We have all witnessed a sea change in the retail sector, with the constant barrage of store closings from the likes of Sears, Macy’s and others. Most malls are beginning to look like ghost towns. A popular misconception is that the rise of e-commerce and Amazon is the reason for the demise of the mall. But Austan Goolsbee, author of the NYT article, provides a convincing array of other reasons. He notes that e-commerce is still only 11 percent of retail spending. So, what are the factors that contributed to the rapid decline in the popularity of malls? People have changed their shopping habits. More and more now shop in big warehouse discount stores like Costco and Walmart. A primary reason for that is income inequality and the shrinking middle class. The middle class was the main customer base for the standard retail outlet and their dwindling numbers has had a direct impact. The rich save more and spend less on retail and there is also a shift to paying for healthcare services (getting progressively more expensive), education, food and entertainment. The net result is a lot fewer dollars flowing into traditional retail outlets and a lot less traffic in malls. Grist for your grey cells! 


My thanks to Anna Abovyan for forwarding guidelines for building AI systems.

I am always looking for feedback and if you would like me to cover a story, please let me know. “See something, say something”! Leave me a comment below or ask a question on my blogger profile page.

V. “Juggy” Jagannathan, PhD, is Director of Research for 3M M*Modal and is an AI Evangelist with four decades of experience in AI and Computer Science research.