NYAS says AI should scale care

The New York Academy of Sciences published a May 21 Academy Blog article arguing that healthcare AI’s clearest near-term use is not discovering new therapies but getting established care to more patients. The piece, written by Brooke Grindlinger and built around comments from former New York City health commissioner Dave A. Chokshi, says many of healthcare’s biggest failures happen before treatment begins — when patients drop out during referrals, intake, scheduling or follow-up. Dave A. Chokshi, a physician and former New York City health commissioner, said in the article that AI’s “greatest promise” may be helping proven care reach “the patients medicine still misses.” The framing shifts attention from high-risk clinical autonomy toward narrower operational tasks that can be measured in booked visits, completed intake forms and successful follow-up. (nyas.org) ### Why is NYAS focusing on access failures before the visit? The May 21 article says healthcare breakdowns often occur in “ordinary” administrative steps rather than in the exam room. NYAS points to referral leakage, intake abandonment, scheduling friction and poor follow-up as examples of where patients can fall out of the system before receiving care. (nyas.org) Brooke Grindlinger’s article presents those failures as targets for automation because they sit inside bounded workflows. That means AI can be used to prompt patients to finish forms, steer them toward open appointment slots, or help staff identify who needs outreach, rather than making open-ended clinical decisions. (nyas.org) ### What exactly did Chokshi say about AI’s “greatest promise”? Dave A. Chokshi made the same argument in an April 13 NYAS-linked essay originally published by Healthbeat. In that piece, he wrote that “at the least, we can make it easier for patients like him to receive the care we already know saves lives,” linking AI’s value to delivery and adherence rather than only discovery. (nyas.org) The May 21 NYAS post extends that view by placing access problems at the center of healthcare AI deployment. In practice, that supports tools that recover incomplete intakes, reduce scheduling delays, clarify next steps after a visit and help providers fill canceled slots from waitlists. That is an inference from the workflow examples NYAS lists, not a separate NYAS ranking. (nyas.org) ### How does this fit into NYAS’s broader AI-in-healthcare agenda? The New York Academy of Sciences held its “New Wave of AI in Healthcare 2026” symposium on May 12 and May 13 in New York City with the Icahn School of Medicine at Mount Sinai. The event description said AI and digital technologies are reshaping diagnosis, treatment and care delivery, and the speaker roster included leaders from Mayo Clinic, Microsoft, UCSF, NIH and Mount Sinai. (nyas.org) The timing matters because the May 21 article appeared days after that conference. NYAS did not describe the article as a formal policy statement, but the blog post and conference materials show the organization is convening discussion not only around frontier models, but also around patient-facing delivery problems and real-world deployment. (nyas.org) ### Which healthcare AI projects fit this argument best? The examples most consistent with the NYAS article are front-office and care-navigation tools. Those include intake completion prompts, scheduling-readiness checks, referral follow-up, reminder and rescheduling systems, and waitlist backfill when appointments open. Those categories follow directly from the access failures named in the article. (nyas.org) NYAS’s article does not rule out AI for discovery or diagnosis. But Chokshi’s argument, as presented by the Academy, puts the immediate case for AI in a narrower place: making sure patients actually receive care that already works. Readers can find the full article on the Academy Blog published May 21, 2026, and related event material on NYAS’s May 12-13 symposium pages. (nyas.org)