The Humanity in Data
A disclaimer: Below I present an account of a woman that I met while attending a bookclub. The story told here is true, as far as I am concerned, but the woman herself has been stretched and reconfigured, beyond the scope of our brief interaction, to fit the more universal argument I’m attempting to construct. For her sake, and for yours, I wanted to make my unchecked manipulation of her as a character clear upfront.
“Chasing data is not the answer,” said the middle-aged doctor to a room full of mostly like-minded professionals. “With better people, half the problems in medicine will go away.”
She sat perched on the edge of a couch, precariously balanced next to a stack of books about honing discipline, staying true to principles, and practicing reason in the face of adversity. As she spoke, the heads of those listening bobbed in a kind of passive acquiescence, a crowd of warm bodies producing an appropriate response. In our minds, we were ticking off boxes next to the arguments we’d heard before:
The imminent robot takeover…
Hospitals, restaurants, and schools full of screens…
Complex diagnoses made by algorithms with human input disregarded…
Before long, the doctor’s commentary devolved into a rant, a runaway truck whose conversational brakes were lost to her passion and devotion to a profession she nevertheless seems to despise. The discussion ended before anyone could offer a satisfying response, so here I am, weeks later, intent on putting my staircase wit into writing.
The Expert Falls from Grace
Drawn in the proverbial sand are two lines separating me, a 24-year-old product designer with a B.A. in coloring and telling computers what to do, from this late 40s physician with decades of secondary education, advanced training, professional accolades, and an actual list of lives she has saved.
The first line is made of time, the forward march, the generational divide. i.e. I am a Millennial scoff scoff and she hails from GenX — the forgotten species of grungy antiestablishment slackers— the dizzy products of the couples that trudged forward with early pregnancies despite the delayed family planning of many of their contemporaries.
Though her X evokes a smiley stamped pill and mine, an iPhone, we are both versed in the drug that is cultural hegemony. So, I was confused when she proposed a renewed allegiance to expert knowledge re: “Hire better people.” My initial conclusion was that her distrust is made of different stuff than mine, but in more ways than one, I am getting ahead of myself.
See, we had set out that day to discuss the Nobel Prize winning work of psychologists Daniel Kahneman and Amos Tversky via Michael Lewis’ The Undoing Project. Kahneman and the late Tversky are among the intellectual grandfathers of such blended disciplines as behavioral economics, decision science, and HCI. Their findings have influenced the practices of those employed in pretty much any domain that requires decision making about people. The extent of their influence has a great deal to do with why Lewis wrote a book about them.
The simplest but hardest to swallow contribution of K&T is that people are not rational, at least not when it comes to making judgments under uncertainty. Instead, we navigate the world with a toolbox full of choice heuristics which, the vast majority of the time, function for us quite well.
In the direct words of K&T:
People rely on a limited number of heuristic principles which reduce the complex tasks of assessing probabilities and predicting values to simpler judgmental operations. In general, these heuristics are quite useful, but sometimes they lead to severe and systematic errors.
If you are new to the names of Tversky and Kahneman, I will appeal to you, earnestly, to go check out one of the links above; theirs is a brand best served raw.
In the meantime, I’ll return to the doctor.
In no domain are the “severe and systematic errors” in our cognition more pronounced and more dire than in medicine, where all judgments are constrained by the uncertain workings of the human body. Lewis devotes an entire chapter of his book to discussing how K&T’s work made its way into the Emergency Room, offering a case study of a medical practitioner employed by Toronto’s Sunnybrook Hospital specifically to check the thinking of his colleagues. Don Redelmeier, the man in question, worked to turn one of Canada’s largest regional trauma centers into not just “a place to treat the unwell but also [a] machine for coping with uncertainty.”
The main premise of our doctor’s rant, however, was an obsession with the idea that mistakes on the ER floor would be reduced if only the people working were more practiced. Now, K&T take particular care in all of their work, to make clear that cognitive biases don’t just affect college undergraduates or lay people. Biases in judgment under uncertainty are not the result of a lack of experience, they are a product of the particular way the human brain perceives the world.
Though the doctor’s feelings were contradictory to the main premise of The Undoing Project and in turn, of our discussion, she did allude to a point often overlooked by critics of K&T, who complain that their assessment of our cognition is overly negative. Namely, paired with the statement about severe and systematic errors is the statement that “in general, [our] heuristics are quite useful.” It is true that doctors are usually right. It just so happens, that when they are wrong, people die.
While I could never be sure from our brief interaction, I suspect that the doctor spoke mostly from a place of fear. Medical practitioners are not the revered figures they once were. They are harried, rushed, depressed, and most of all—entrenched—stuck in a system that is as frustrating for them as it is for patients. Not to mention, the royal “we” of the technical world keep throwing ‘solutions’ their way that add steps to their process without paying deference to their workflows.
But beyond the clear brokenness of our healthcare system, lurks a much more general fear, one which disproportionately affects those, like doctors, who have sought knowledge and a livelihood through specialization. That fear is this: Technology is waging a war against expertise—and because we remain so so human, it is a war of attrition.
Man versus Machine
The second line separating me and the doctor concerns our dueling perceptions of the current state of affairs and what it really means to “chase data.”
In the two decades we’ve both witnessed, sides have been taken. The battlefield is muddled, bloody, and mostly ruled by Amazon. But from my screen addled, over stimulated, Millennial point of view, there is hope.
My perspective on the future re: the so-called “imminent robot takeover” involves three parts, each of which explores a different human-centered tension.
Check back soon.
Part 1: The Monkey and the Robot Holding Hands
Part 2: Marooned on the Island of Now
Part 3: The Rise and Fall of the Philosopher Queen
With a little help from my friends
My desire to broach a topic as multifaceted as this one (with the awareness that my words may reach very few) mostly stems from my grappling with the paradigm altering work listed below.
Algorithms to Live By: The Computer Science of Human Decisions — Brian Christian and Tom Griffiths
American Gods — Neil Gaiman
How to Create a Mind — Ray Kurzweil
Player Piano — Kurt Vonnegut
The Better Angels of our Nature — Steven Pinker
The Most Human Human — Brian Christian
Thinking Fast and Slow — Daniel Kahneman
Strangers Drowning: Voyages to the Brink of Moral Extremity—Melissa MacFarquhar