Maybe It’s Time to Pump the Brakes on AI

The healthcare system might stand to undergo a few adjustments before we turn things over to the machines.

There’s a moment early in the 1993 blockbuster film Jurassic Park when the movie’s designated skeptic, Dr. Ian Malcolm (brilliantly portrayed by a seemingly heavily caffeinated Jeff Goldblum), engaged in a spirited debate with the titular theme park’s creator over the ethical considerations of resurrecting dinosaurs for profit, makes something of a profound point regarding the motivation behind the endeavor.

“Your scientists were so preoccupied with whether or not they could,” Malcolm says, “they didn’t stop to think if they should.”

The scene came to mind when digesting a recent RACmonitor article, “AI in Healthcare is a Looming Inevitability.”

“Artificial Intelligence (AI) promises to greatly reduce the cost of healthcare,” the article begins, before delving into the all-too-real-sounding scenario in which AI gradually (and then rapidly) replaces many (if not all) of the 950,000 patient care physicians in the U.S. – a notion seemingly championed not as dystopian or terrifying, but merely fiscally responsible.

“AI has won chess and Go tournaments, and now can pass an MBA exam. It can make complicated decisions. It recently created a complete Al Joe Rogan show, including the voice, indistinguishable from the real personality’s voice. It is better at summarizing information than humans. It writes and responds better than 99.9 percent of humans. In the workforce, AI is replacing humans at a relentless rate,” the article reads. “AI is not unionized; it does not sleep; it works 24 hours per day; it has a capacity to process millions of operations per second. A single giant data center could service the entire United States, and more.”

Well, at least they won’t unionize, right? (But wait – do we even know that for sure?)

Before diving headlong into this, let’s first take a step back and a deep breath and a 30,000-foot view. That could involve a quick glance at where we are, not where we’re going. And where we are is, well, ugly.

To call the American healthcare system’s performance abysmal would even be an understatement. Per federal statistics, we doled out a staggering $4.3 trillion on healthcare in 2021, or nearly $13,000 per person.

What did all that dough get us? Well, The Commonwealth Fund took a look at 11 otherwise comparable nations a couple of years ago. We ranked dead last (by considerable margins) in four out of five key healthcare metrics: access to care, administrative efficiency, equity, and outcomes. In terms of healthcare spending as a percentage of GDP, the 10 other countries studied all clocked in at or below 10 percent. We’re pushing 20 percent (and it was less than 5 percent in 1960).

I’m sure it’s merely coincidental that we’re also the only high-income nation on earth that doesn’t universally guarantee health coverage. Otherwise known as the reason that two-thirds of all personal bankruptcies in the U.S. result from medical bills.

But wait just a second! Isn’t all this an argument in FAVOR of AI, not against it? Is our healthcare system so fatally flawed that burning it all down and starting from scratch could be a viable solution?

You’d think so, right? But no. Doctor salaries add up to about 8 percent of U.S. healthcare costs (not 80, 8). The Kaiser Family Foundation estimated that even a drastic 40-percent cut in such salaries would reduce overall healthcare spending by about 3 percent.

Making up a far bigger piece of the pie is administrative spending, accounting for somewhere between 15 to 25 percent. And “at least half” of such spending “does not contribute to health outcomes in any discernable way,” according to a Health Affairs research brief.

Yes, pointing out the problems is far easier than identifying the solutions, and I’d be the first to admit that there is a long, winding road ahead toward improvement. But I’d humbly suggest that we just need to see some of that improvement before even dreaming of talking about AI doctors becoming the norm. By no means am I against AI as a concept, but artificial intelligence will only be as effective as the human intelligence applying it. And human intelligence got us in the mess we’re in now.

Plus, an AMA survey done last year found a record high of 63 percent of physician respondents reporting symptoms of burnout. Could we cut them some slack before we start suggesting they get the ax? I’m sure if you asked them, they’d say that burnout is preferable to unemployment.

In the meantime, maybe we keep the words of Dr. Malcolm in mind. If I recall correctly, the rest of the movie didn’t do very much to prove him wrong.

And let’s all perhaps be a little more skeptical of anything described as a panacea. Too often such promises turn out to be more of a Pandora’s box.