The Way of All Flesh

Posted by Ron Decker on 03/25/2019

The topic of artificial intelligence (AI) is a vast and complicated one. It’s alluring to the wacky and visionary, a source of terrifying possibility to the unhinged uncle standing before the half-closed blinds and looking into the grey empty street. The experts say it’s the way of the future and when the entrepreneurial technocrats start making prophesy, it’s best, as the last twenty years have indicated, not to drag your heals too much in opposition. It’s more favorable to look into what it’s all about.

AI is, simply put, a branch of computer science that develops machine learning. This makes it seem like something you can shake hands with, be friendly. Though it’s become a hot-button topic in recent years, it’s not necessarily that new a concept. In the early 1940s Alan Turing predicted that digital computers would ultimately be sophisticated enough to engage in logical reasoning. MIT, circa 1950, dedicated a program of study to it. That was three decades before the first digital wristwatch, twenty-four years before the first pair of Nike shoes swooshed into the market, and forty-four years before Mr. Zuckerberg’s birthday.

In the chaotic world of healthcare, AI offers far-reaching possibilities geared toward addressing problems that have materialized out of our reliance on unsustainable processes and procedures. In “Patient, Heal Thyself,” Roy Smythe puts these clumsy holdovers succinctly, writing, “[w]hile modern medicine has made amazing contributions to acute care, reducing mortality and mass epidemics, the organized structure of it is not effectively serving the general medical needs of the world’s population, and those chiefly responsible for patient care—physicians, nurses, and other caregivers—are being pushed to the point of physical and psychological burnout.”

This is where experts say AI can play a vital role in protecting doctors and other medical professionals from oversaturation in inefficient practices—repetitive tests, monitoring, reliance on limited databases, workflow and administrative tasks, clinical judgments and diagnosis, image analysis—leaving them more opportunity for engagements that require keen human supervision. And of course someone will need to keep an eye on the technologies.

AI also empowers patients to play a more active part in their own care through the utilization of apps, virtual consultations, remote monitoring, and other technologies that prevent avoidable hospital visits and excessive expenses. For remote populations lacking accessible medical care, the technologies will be, in many ways, a salvation.

This all makes AI seem jolly good and warm saline sea breezes and something to anticipate, something to look forward to watching mature, like the neighbor’s five-year-old who seems to be fascinatingly competent with a tennis racket. You look forward to its adulthood. And to a large extent this promise is appropriate.

But, on the other hand, maybe we shouldn’t altogether forget that paranoid uncle tracking a residual unease, surveying the boulevards. A lot of AI’s future rests in cost-saving practices and financial incentives, not altruism. Maybe we should ask questions if the physician shows up for the pre-op consultation and explains a machine will be wielding the scalpel. Maybe we should feel somewhat unsettled when our EHRs are abstracted to some cloud-based data center to live an eternal life alongside electronic reports on other people’s bodies. Maybe we should feel dubious when we can’t get into our GP and, as recourse, a series of codified multiple-choice answers prescribes us antibiotics for our kid’s sore throat.

Subdued paranoia and articulate skepticism have always played important roles in improving revolutions in their infancies. A new form of healthcare is moving inevitably in the direction of AI technologies. The current system is unsustainable—there’s unanimity on this. But that doesn’t mean a dose of slack-jawed skepticism isn’t a boon in our approach to these new, mesmerizing technologies.

Post A Comment
(Will not be published)