Healthcare is “so complicated,” even for artificial intelligence

The team at


AI’s potential in healthcare is curbed by a maze of policy, regulation, and importantly, concerns for patient privacy. “The current healthcare delivery system unfortunately remains structurally ill-suited to absorb and deploy rapid advances,” write researchers from the Stanford One Hundred Study on Artificial Intelligence.

This week, ImpactAlpha is extracting nuggets from Stanford’s century-long effort to understand AI’s long-term possibilities and dangers. Even with a flood of digital data from electronic records, mobile health and wearable devices, identifying patterns between symptoms and diseases usually falls to physicians. “With automated assistance, the physician could instead supervise this process, applying her or his experience and intuition to guide the input process and to evaluate the output of the machine intelligence,” the report suggests.

Elder care is a prime opportunity for AI-enabled healthcare. The U.S.’s 65-plus population is expected to double between 2012 and 2050. As American seniors embrace technology to prolong their independence quality of life, smart devices will become part of daily life, from robots for dressing and hygiene to devices for improving hearing, sight and mobility.

Regulatory and structural barriers have prevented more rapid clinical and diagnostic adoption of AI. That may not be entirely bad. The report warns that AI could introduce potentially life-threatening risks unless systems are “well-targeted.”

This post originally appeared in ImpactAlpha’s daily newsletter. Get The Brief.

Photo credit: ThinkStock