A Better Way to Think About AI

"Sometimes it seems the most direct route is to automate wherever possible, and to keep iterating until we get it right. Here's why that would be a mistake: imperfect automation is not a first step toward perfect automation, anymore than jumping halfway across a canyon is a first step toward jumping the full distance. Recognizing that the rim is out of reach, we may find better alternatives to leaping-for example, building a bridge, hiking the trail, or driving around the perimeter."

"Rather than asking AI to hurl itself over the abyss while hoping for the best, we should instead use AI's extraordinary and improving capabilities to build bridges. What this means in practical terms: We should insist on AI that can collaborate with, say, doctors-as well as teachers, lawyers, building contractors, and many others-instead of AI that aims to automate them out of a job."

"Radiology provides an illustrative example of automation overreach. In a widely discussed study published in April 2024, researchers at MIT found that when radiologists used an AI diagnostic tool called CheXpert, the accuracy of their diagnoses declined. "Even though the AI tool in our experiment performs better than two-thirds of radiologists," the researchers wrote, "we find that giving radiologists access to AI predictions does not, on average, lead to higher performance." Why did this good tool produce bad results?"

Automation will increase, but imperfect automation is not a first step toward perfect automation; attempting half-measures can produce harmful outcomes. Safer and more effective approaches involve building bridges: designing AI to collaborate with human professionals rather than replace them. AI should assist doctors, teachers, lawyers, contractors, and other workers so human judgment remains central. A radiology experiment with CheXpert showed that providing AI predictions reduced diagnostic accuracy because clinicians lacked clear guidance on when to defer and often overrode confident AI outputs. AI development should prioritize augmentative tools and protocols for reliable human-AI collaboration.

Read at The Atlantic