Teaching Kids to be Human in an Algorithmic Age

Teaching Kids to be Human in an Algorithmic Age

Algorithms.  Our day-to-day lives are increasingly influenced, if not controlled, by algorithms. Understanding algorithms seemed beyond my reach.  Until I read an article by Hannah Fry, an associate professor of mathematics at University College London.

The article made algorithms understandable, fascinating even.  Turns out, the article was an excerpt from Fry’s recent book, Hello World: Being Human in the Age of Algorithms.  

I found the book at my library, and finished it in a few days. Just like in the article, Fry explains algorithms in a way that even my math-challenged self can understand.  My initial interest in the book was mere curiosity.  Now that I’ve read the book, I think a basic understanding of algorithms would be helpful for everyone.  

We likely sense how much our lives are impacted by algorithms. But, we may fail to fully appreciate how algorithms are creating significant dilemmas.  As we humans continue to advance our technological capabilities, these dilemmas will need resolved.  There are no clear, simple answers.  In my mind, this makes understanding algorithms, and recognizing their limitations, crucial to us all, especially our kids.

Fry walks the reader through the basics of algorithms and explains how they’re fed real-world data to make their calculations. Then, in each chapter, Fry explains how these data-consuming “mathematical objects” work in the fields of justice, medicine, transportation, and art.   

I was particularly intrigued by the chapter on justice, having spent much of my career defending people accused of crimes.  Fry walks the reader through several studies that demonstrate how innate human biases impact defendants.  Even with the best intentions, similarly situated defendants get sentenced differently.  So, some judicial systems have deployed algorithms to eliminate that disparity.  

Compared to humans, these algorithms have greatly improved consistency, giving the exact same output when provided the same inputs. The algorithms are also better at making predictions about future behavior.  So, there’s greater fairness and equity in a system aided by algorithms. At least, so it seems.

Fry goes a step farther with her explanation, however, and shows how even algorithms can make mistakes. 

It seems that these same algorithms have the ability to produce “false positives” and “false negatives,” meaning that some defendants are mislabeled by the system.  Once mislabeled, punishments can be too harsh or too lenient.  This creates a different type of disparity — one we’re less likely to acknowledge. This is where the some of the dilemmas occur. 

To prove this point, think about how frequently you question the recommendations you get from Netflix, or Amazon, or Google.  Think about the number of times you’ve followed your GPS and ended up in the wrong place.  What if a judge, or a doctor, or a driver of a self-driving car follows an algorithm’s output with the same degree of confidence and fails to see mistakes.  In those situations, obviously, the stakes are much higher than when we trust in Google and Netflix.

According to Fry, we cannot blindly follow the directions we’re given by our GPS, nor can we turn over criminal sentencing, medical diagnosis, or even driving, without first accepting the fact that algorithms are no more perfect than humans. 

My kids are a bit too young for Fry’s book, but I’m interested to know how much they trust an algorithm.  Particularly since they’ve never lived without them.  Will they be as trusting as the guy who nearly drove off a cliff because he never questioned his GPS?

I’d like to think that my kids have better sense, but, until I read this book, I never thought to ask.  

I think, as a parent, it’s a question I need to ask. After all, as we move further into the age of the algorithm, it’s my kids and their classmates that will wrestle with these issues throughout their working lives.  Regardless of the work my kids choose to do, they’ll likely face these dilemmas. Even if they’re not writing the algorithms, they’ll likely be working with algorithmic information.  Will they feel comfortable challenging an algorithm if they think it’s wrong?

I’m not entirely sure my kids are ready for that.  But, one way to prepare them for this complexity is to start having conversations to get them thinking.  Get them to see behind the curtain.  Help them understand how imperfectly algorithms sometimes operate.  

And, at the same time, help them understand that while human history is full of imperfections, it’s also filled with courageous people who challenged those imperfections and led us closer to perfection — well before we started feeding data to algorithms.  By talking about our future, and reminding them of our past, I think we can start to teach our kids how to remain human in this algorithmic age.

The Unpredictability of Predictions

The Unpredictability of Predictions

Career Education for the 21st Century

Career Education for the 21st Century