Ten years ago, I attended an ACLS (Advanced Cardiac Life Support) course where each doctor took turns practicing cardiac resuscitation on a manikin. A computer inside the manikin changed heart rhythms every minute, while a doctor tried to identify and treat them. I watched one doctor in particular. Whenever he erred, the computer’s ECG tracing “flat lined,” which meant he had “killed” the manikin. The first time this happened the doctor expressed frustration. But as he killed the manikin over and over again, he got over being sensitive, and killing this way or that got to be almost mechanical. I thought he would tire of the amusement, but instead his industry grew maniacal, a sort of duel between human and machine—only I began to wonder which was which. The manikin had pretty blinking eyes and a pleasant grin, while the doctor, his facial expression blank, his arms and legs in alive positions but never moving, looked as if he had been turned to metal.
Until recently, many professionals, including many doctors, thought they needed to become more machine-like to keep their jobs. Such worry contributed to the specialization trend of the last century. The more circumscribed their task, the more professionals maximized their knowledge, efficiency, reliability, and predictability when performing it; by emulating the virtues of the machine, professionals hoped to keep both their mechanical competitors and human competitors at bay. But this strategy has been compromised. AI has opened up a large gap between machine performance and human performance, one that people will likely never close. Worse, by over-specializing, professionals have unwittingly shrunk their job duties to fit inside a machine’s scope of practice, thereby making themselves more vulnerable to AI replacement. An AI cardiac resuscitator is only a matter of time.
Ultimately, their mistake was definitional. They equated “expertise” with the “best.” An imaginary continuum existed in their minds: at one end, the “worst,” including stupid, lazy, foolish, and unreliable people; at the other end, the “best,” meaning the machine, whose talents a human being could only approximate. The “best” human workers were those who purposely limited themselves to a small corner of life, and mastered it, allowing them to approximate the skill set of a machine.
With this strategy’s eclipse, a new imaginary continuum is needed, along with a new model of excellence to strive for. Here is one:
Take a continuum that has an amoeba at one end and a mathematician at the other. The amoeba has the most interest in its immediate environment. It reacts strongly and quickly to the smallest stimulus. The mathematician has the least interest in his immediate environment. He is detached from temporal considerations and lives in an insular world of numbers. Somewhere between the amoeba and the mathematician sits the competitive worker of the future: the worker who reacts to a stimulus as an amoeba does, although with more thought, but not with so much depth of thought that he or she ignores the stimulus altogether.
Poker offers a vivid expression of this worker in action, along with the continuum on which that worker sits. Both constructs came to mind during interviews I recently conducted with several poker champions.
I began by asking them if the “best” poker players win the most money. No, they chuckled; poker players who win the most money are good at game selection; they purposely play high-stakes home games against terrible opponents. There are three types of poker players, they continued: recreational players who want to have fun, business-minded players who want to make the most money, and elite players who want to push the edge of the game.
At one end of the continuum lies the recreational player. Recreational players don’t think about complex play. They don’t even think about winning. They just want to enjoy the immediate pleasure of being dealt cards, throwing chips into the pot, and occupying the center of attention. These are people who say, “It’s more fun to play badly than to play well.” They bet when they should fold because it’s more fun to bet, and they certainly didn’t drive fifty miles to a casino just to sit at a poker table and do nothing, one champion said. Recreational players are slapdash and impulsive. Like the amoeba, they are all feeling.
At the other end of the continuum lies the elite player. Such players are all reason. They adopt a mathematical strategy called “game theory optimal” (GTO) that builds on the work of mathematician John Nash, and which guides their play independent of whatever cards their opponent holds. GTO is best calculated using a computer, but elite players do pretty well on their own. Math tells the player when to fold and when to bet; no discerning eye, imagination, or intuition on the player’s part is required. On the contrary, to pursue a strategy where one tries to guess what an opponent has in his hand—the old way of playing poker—is to open oneself to counter-exploitation. Sentiments once heard in poker play, such as “I felt for sure he was bluffing” or “I just had the feeling he had an ace,” mean nothing in elite poker, I was told. No poker professional would utter such phrases today; they are the sign of the amateur. The elite poker player, like the mathematician, is completely rational.
In between the recreational player and the elite player lies the business-minded player (whom I call the “wise” player). Unlike recreational players, wise players are disciplined and logical. Unlike elite players, they rely on intuition to scrutinize their opponents and predict their behavior. Reality absorbs them. They study an opponent’s eyes, hands, and posture; they look for twitches and fidgets. They combine feeling and reason to gauge their opponents, to judge them, to goad them, and to trap them. Caring little for abstract mathematical theories, their conversations about poker often include anecdotes, judgments about people, gossip, and practical facts.
Elite players, I learned, could not make the most money even if they wanted to, as GTO lowers the chance of large winnings. GTO play is a defensive strategy and not the maximally profitable one. Although GTO play wins over the long run, it wins less. It is the wise player who can spot a fool a mile away who wins the most money in poker.
All this offers some useful insights into the AI and robot economy of the future. Many workers must give up trying to emulate the machine, or using their machine-like qualities as their major selling point. In those jobs that pay people for approximating a machine without actually being a machine—for instance, professional athletes, dancers, chess grandmasters, and competitors on Jeopardy—aspiring to the level of the machine will still make sense, as being almost as good as a machine without being one enjoys cachet. But most jobs lack this quality; more precisely, cachet doesn’t add to the bottom line, and employers will likely replace a worker with a machine if the machine is cheaper, more productive, and more reliable. Even actors are vulnerable in this regard, as movie producers test the waters to see if audiences will accept a machine-generated actor. True, such an actor lacks the cachet of a human actor, but if audiences no longer care, the human actor will likely lose out to the machine actor, as the latter stays at the perfect weight, never forgets his or her lines, and never causes scandal or complains.
Rather than aspire to the extreme end of the old continuum—to be the “best,” or the most machine-like—future workers should aspire to the middle of the new continuum. Better to become the wise person whose comparative advantage lies in being able to grasp reality in great depth. Such a person may not care about money in the way a wise poker player does. Nevertheless, the wise person possesses unique cognitive skills that will likely be rewarded with better pay.
Poker illustrates this point. When a wise player plays against an average player, the wise player wins more money than the machine does in the same situation, as the wise player is better than the machine at using the old exploitative strategy. The wise player knows how to get the average player angry or flustered, or how to pick up on a “tell.” When the opponent is life more than it is numbers, the wise player wins more money than the machine does because the wise player knows more and understands more—an important cognitive skill.
What will distinguish wise doctors from AI? They may not know off the top of their heads every possible diagnosis associated with a particular symptom, but they will know which patients are more likely to be careless and forgetful; rather than prescribe them a week’s worth of penicillin pills for strep throat, for instance, they will wisely give them a single injection of penicillin to last them an entire week. What will distinguish wise lawyers from AI? They may not remember every precedent in the law books, but they will have a vital feel for when clients seeking a tax shelter “must do” something rather than “may do” something. What will distinguish a wise interior decorator from AI? They may not know every possible design option for a kitchen, but they will grasp the power dynamic between the two homeowners and know whose preference matters more.
How does one become wise? Through experience and instinct—and by studying the liberal arts. Not the political activism or ideological constructs that often masquerade as the liberal arts, but the liberal arts as traditionally taught. The liberal arts enhance psychological acuteness and let people see into the heart of things.
Science and technology have made life better. Excited by the promise of more advances to come, and the high remuneration that comes with employment in these fields, people these days tend to give the liberal arts short shrift. Unlike STEM fields, the liberal arts do not seem to “add any value,” which is why student interest in the liberal arts has plummeted. But this is wrong thinking. The liberal arts do add value. They teach a person to open up vistas and plumb depths, using all the different facets of his or her intelligence and personality. At the very least, they offer Shakespearean advice about the future economy: “Neither an amoeba nor a mathematician be.”