Why AI has fallen well short of outsmarting COVID-19

Understood as a virtual army in the war against COVID-19, AI has vast stockpiles of potential weaponry with which to wage many a battle. That’s the good news.

The not-so-good news is that potential, so far at least, is all the technology has to show for all the chest-beating of its enthusiasts.

Wired writer Gregory Barber breaks down the reasons why in a concise analysis posted May 19.

On the bright side, he notes, a therapeutic drug identified via algorithmic data sifting, baricitinib, is now headed for clinical trials.

In the arena of the purely hopeful, Barber continues, an AI algorithm “could mine lots of patient records and determine who is more at risk of dying and who is more likely to survive, turning anecdotes whispered between doctors into treatment plans.”

Next, sounding a note of sheer realism, he underscores that AI’s success or lack thereof in taking on COVID-19 is “all a matter of data—what data we’ve already gathered and whether we’ve organized it in a way that’s useful for machines.”

“Our healthcare system doesn’t give up information easily to train such systems,” Barber writes. “[P]rivacy regulations and balkanized data silos will stop you even before the antiquated, error-filled health databases do.”

It’s possible the present crisis will shake things up enough for AI to begin whooping some COVID butt, he allows, in so many words. So far, however, signs of an AI cavalry riding in to save the day are hard to come by.

That’s a paraphrase. To read the piece, click here.