I think about 15% of the essays submitted to First Opinion these days boil down to: "AI'll fix it."
Unsurprisingly, most of these pieces are written by executives at companies selling health care AI. Inevitably, I pass on them.
I'm not against publishing a piece written by a CEO — I think that it's often valuable to hear their perspectives and get them on the record.
But I am against essays that spend 80% of their words spelling out a (very real!) problem, and just 10% on the "solutions" part, handwaving away potential problems. It's ultimately unsatisfying. (You didn't ask, but the other 10% of the standard version of this essay? Usually it's in the conclusion and repeats the thesis statement with some starry-eyed big-picture language.)
It's not just that "AI will fix it" essays end up seeming too self-promotional — they also just aren't terribly interesting, absent some surprising argument.
I'm far more compelled by pieces that make a sharp point about AI's limitations. For instance, First Opinion published a really thought-provoking essay by Hamid Tizhoosh on the "AI pseudo-revolution" in health care. "Despite grand predictions that AI would revolutionize medicine — perhaps even replace doctors — its impact has instead stalled, hanging in the balance of inaccuracy, bias, and overhyped potential," he writes. Tizhoosh has been working in AI in medicine for 30 years and is disappointed to see health systems racing to adopt systems before they are truly ready. I highly recommend its clear-eyed look at AI.
It's part of STAT+, so it's behind the paywall. I wish we didn't need a paywall, but good journalism costs money to produce. The good news is that until tomorrow, you can get 50% off a STAT+ annual subscription. Go here and use code AUGUST50.
Recommendation of the week: I'm running dry here! What books, movies, TV shows, and podcasts have you enjoyed lately?
No comments