Home

June 2017

State troopers call it "death by GPS." People have become so trusting of their gadgets that they go wherever their smartphone tells them, sometimes ending up stranded in the middle of the desert or on a snowbound mountain road, or driving into a lake or ocean.

From getting directions to finding a prospective mate to gauging how many calories we're consuming, we're now relying on computer algorithms for many facets of our lives. An algorithm might be defined as a piece of computer code that performs a specific function.

We've become cyborgs. We rely on our smartphones to remember things for us, to mediate our relationships with other people (texting, social networks), indeed, even to create our identity.

Take the simple weather app on my iPad. I rely on that every day to decide how to dress, whether to ride my bike to lunch, or whether to take my umbrella with me. It shows me hour by hour what I can expect all day long. How did I ever live without that algorithm?

Or take dating apps and services. A study published in 2013 based on data from 2005–2012 found that more than one-third of marriages in the U.S. began with online dating – and the couples whose matchmaker was an algorithm tended to be slightly happier than those who got together in conventional ways. (That figure may be high, though, with a 2015 survey finding that 20% of people currently married or in committed relationships met online via a dating service.)

Also, today it's hard to imagine taking a road trip without GPS telling us in a friendly voice how to get to our destination, including turn-by-turn directions. These GPS algorithms have become so wedded to our lives that they are literally changing our brains.

A GPS device tends to activate the brain's caudate nucleus, which is cue-based, whereas map-based navigation primarily activates hippocampus, which is the area of the brain involved in spatial memory. When you use a paper map, you're essentially using the hippocampus to build a map in your head. Some neuroscientists think that using GPS is causing our hippocampus to atrophy.

We're also increasingly trusting our health to algorithms. For example, smartphone sensors track our steps, our pulse, our activity levels, and more. The accompanying apps make recommendations for workouts that will improve our health. We can also use these apps to track our food consumption so that our activity level keeps up with our calories consumed.

However, it's one thing to follow exercise recommendations based on an app's algorithm, but quite another to trust an algorithm to the extent that Angelina Jolie did. A statistical analysis found that a particular gene she carried was associated with an 87% chance of getting breast cancer. She opted for a double mastectomy.

In fact, algorithms are increasingly being used in medical diagnosis. Doctors enter symptoms into IBM's Watson, which is proving to be quite adept. Last year Watson diagnosed a Japanese woman's rare form of leukemia in 10 minutes – after it had stumped doctors at the University of Tokyo for months.

Almost without noticing, algorithms are becoming integrated into many facets of our lives. Where will it lead? Elon Musk, the famed engineer known for such gadgets as the Tesla automobile and SpaceX rocket company, sees these trends and believes that the future entails even greater integration of human and machine.

He sees the worldwide network of computers and their algorithms as simply another layer of the brain and has founded a company called Neuralink to create a direct connection between the brain and this network. A tiny electrode implanted in one's brain would allow a speedy wireless connection – much faster than interacting with computers and the internet via voice or typing.

In fact, Musk feels like we don't have any choice but to take this step. Why? He's one of a number of top intellectuals who are worried about computers surpassing humans in intelligence – and taking over. He says that we need to have a neural connection simply to compete with them. One of the reasons he's planning to establish a colony on Mars is as an escape hatch, should computers take over Earth. Seriously.

In fact, artificial intelligence is indeed becoming a bit scary – a topic for a future column. The neural net approach that I wrote about earlier uses something called "deep learning" to come up with pattern-based solutions – without humans being able to know or understand how they arrived at these solutions.

Whatever the risk, it's clear that we're increasingly dependent on the algorithms in our lives.

This month's hot tip

A fascinating and fuller discussion of these issues can be found in the new book Homo Deus: A Brief History of Tomorrow.

© 2017 by Jim Karpen, Ph.D.

E-mail Jim Karpen