AI Talk: Dark patterns, dueling assistants and a teenage blockchain engineer

June 28, 2019 / By V. “Juggy” Jagannathan, PhD

This week’s AI Talk…

Dark patterns

This week, the New York Times featured an article that talks about “dark patterns.” This term, coined almost a decade ago, is well known in the User Interface design community. Dark patterns specifically refer to underhanded practices used by companies to coax customers into behavior that increases their bottom line. The article calls out practices by an e-commerce website, which informs customers browsing their website that other (non-existent) customers saved lots of money—a psychological ploy to get one to buy their product. Not all of the patterns involve outright deception, but rather make it difficult to unsubscribe, for example, or select an option that means less money to the commercial entity! A Princeton study, released this week, categorized dark patterns into seven categories, from fostering a sense of urgency, to forcing unwanted tangential actions, to being sneaky. Interesting set of tactics! They also note that several third-party entities are cohorts in promoting such patterns. It is highly unlikely that any law will be passed banning any of these practices as it is basically caveat emptor. Better to stay educated on these practices!

Dueling Assistants

The competition is on! The usual cadre of digital assistants was evaluated in a study published in Nature this week. The authors are employees of Klick Health in Toronto, which engaged 46 of their employees in recording a bunch of medication names—brand and generic—and played it back to Google Assistant, Alexa and Siri. Google Assistant, it turns out, is way better than Alexa or Siri at recognizing medication names. Folks working in speech recognition know that the recognizer results are usually good for some laughs, and the results discussed in this paper do not disappoint on that front! Here are some misrecognitions, thanks to Alexa: atorvastatin = “a tour of Staten”; losartan = “low sarten”; meloxicam = “my lock’s a cam.” This particular study focused simply on the recognition of medication names, but the larger intent is to figure out which assistant gives the most relevant information to the consumer.

Teenage Blockchain engineer

I saw this article in Fortune and it links to a YouTube interview of 17-year-old Ananya Chadha who is a blockchain engineer! This bubbly, effervescent, confident teenager is totally fascinating to watch. She seems to have accomplished more in a few short years than a seasoned veteran. One of her main areas of focus is how to use blockchain for social good. I will have to emphasize the “one” in the previous sentence! Check out her website. She is working on brain-computer interface and can control a toy car with her EEG! She created her own blockchain coins, floated her own Initial Coin Offering (ICO) and used the blockchain infrastructure to allow people to submit their genome data using smart contracts! The future she paints is either really scary or totally awesome—you pick. And, she is only 17!

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 Vice President of Research for M*Modal, with four decades of experience in AI and Computer Science research.