Americans are often quite reasonably frustrated by the endless growth in health-care expenses. It's a problem we sometimes would like to believe is all our own, due to the unusual way that health insurance is handled in this country. But in no small part, what frustrates Americans also frustrates people in other countries -- it just manifests itself differently elsewhere.

■ The root cause is that everyone has an incentive to pursue the maximum available coverage for their own health, and the resources to deliver health care are limited by a variety of real constraints. Take, for instance, the NHS in the United Kingdom: It's widely admired and, generally, free to the user. But demand has exceeded supply quite a lot, to the extent that ambulance response times have stretched to shocking levels, well in excess of the NHS's targets (not to mention patient expectations).

■ And nurses and ambulance staff are going on strike, saying they aren't paid enough. Fundamentally, it is the same problem as anywhere else: Health care is an area of almost uniquely limitless demand, and somewhere along the line, that demand runs up against some kind of limitation in supply.

■ High technology often doesn't help the supply side as much as intuition suggests it should, but we may be on the cusp of something new -- if and only if the regulatory and professional environments are prepared to accommodate the possibility.

■ Consider this possibility: For as much as medical professionals criticize patient reliance on "Dr. Google", it seems inevitable that self-service diagnostic tools built on AI platforms and delivered by credible providers will be offered as alternatives to urgent care and walk-in clinics, perhaps at kiosks in drugstores and grocery stores. Lots of tests require facilities (like the ability to draw blood) that most people don't have at home, but which could be provided in fairly frictionless ways at establishments already connected to health care.

■ A human professional would probably still need to review and sign off on the diagnosis, but most of the heavy lifting would be automated. Just as AI is already showing up in radiology, so too could it have a lot to contribute elsewhere in medicine. But a human's guidance still makes enormous sense, just as it remains logical to keep human pilots in the cockpit even though autopilot can fly a plane by wire. Humans working in concert with computers are better than either working alone.

■ The trends involved are so powerful that the helping hand of automation seems impossible to ignore. Between known workforce shortages in the medical field, the quest for cost containment, and the rising health demands of an aging population, all of the pressures causing troubles now are set to make things worse unless there is some kind of fundamental change to the systems of provision. More burnout is unfortunately likely to beget even more burnout.

■ We're already growing familiar with telemedicine. Like a lot of other things, it was forced on a lot of people due to the onset of the Covid-19 pandemic, but now it seems far less unusual or far-fetched than it once did. Medical care driven mainly by machines -- call it automedicine -- is almost certainly coming fast behind. It's up to lawmakers to start thinking right now about what "automedicine" will need to look like. If the people in state legislatures and the halls of Congress don't appear to be up to the task, then perhaps patients need to consider the consequences.

Dr. Google, as rendered by Dall-E

The real Dr Google is coming to deliver your…