Healthcare leaders have spent the last two years asking what AI can do. Can it write notes? Analyze research? Predict outcomes? Improve efficiency?
At Becker’s 23rd Annual Spine, Orthopedic and Pain Management-Driven ASC Conference in Chicago, a panel of spine leaders suggested the more important question may be something else entirely. What happens when technology becomes smart enough to sound right?
Because in medicine, sounding right and being right are not always the same thing. That distinction sat at the center of a discussion between Mohamad Bydon, MD, chair of neurological surgery at University of Chicago Medicine, and Xiaofei Zhou, MD, associate program director and neurosurgeon at Cleveland-based University Hospitals, as they explored what AI may, and may not, mean for the future of spine care.
The conversation was not a warning against AI. Nor was it an endorsement of blind adoption. Instead, it reflected a growing realization across healthcare: As AI becomes more capable, human judgment may become more valuable.
The danger isn’t bad technology
One of the most revealing moments of the discussion involved a medical student, a large dataset and a conclusion that appeared statistically convincing. Using AI-assisted analysis, the student identified a relationship suggesting traumatic brain injury causes diabetes.
The data appeared to support it. The conclusion was still wrong.
“There’s no way TBI causes diabetes,” Dr. Zhou said. The example captured one of the most important limitations of AI. It excels at identifying patterns. Medicine requires determining whether those patterns make sense.
As healthcare organizations gain access to larger datasets and more powerful analytical tools, that distinction becomes increasingly important.
“You start to find things that are statistically significant and clinically meaningless,” Dr. Bydon said. For decades, medicine has relied on physician expertise to distinguish meaningful findings from noise. AI does not eliminate that responsibility. It may amplify it.
Why more data doesn’t guarantee better decisions
Healthcare often talks about data as though more is always better. The panelists challenged that assumption.
Modern spine care generates enormous volumes of information. Registries, electronic health records, imaging studies and outcomes databases create opportunities for analysis that previous generations of physicians could scarcely imagine. Yet raw information alone has limited value.
Dr. Bydon pointed to the American Spine Registry, a collaboration between the American Association of Neurological Surgeons and the American Academy of Orthopaedic Surgeons that now contains data from more than 600,000 patients.
The number itself is impressive. What surprised many in the audience was what comes next.
“We spend 75% of our time cleaning the data,” Dr. Bydon said. The lesson is easy to miss amid excitement surrounding AI. Technology is only as reliable as the information feeding it. A sophisticated algorithm analyzing flawed data does not produce better answers. It produces flawed answers faster.
That reality may become one of the defining challenges for healthcare organizations over the next decade. The institutions that benefit most from AI may not be the ones with the most advanced tools. They may be the ones with the most disciplined approach to data.
The physician’s role is changing, not disappearing
The conversation repeatedly returned to a theme that has become increasingly common as AI adoption accelerates. Many people continue asking whether AI will replace physicians. The panelists appeared far more interested in how it will change them.
“I think there’s a capability right now to help with menial tasks,” Dr. Zhou said. “But it’s not at the point where it can really change certain surgical outcomes.”
That distinction matters. Much of AI’s current value lies in reducing administrative burden rather than replacing clinical decision-making: documentation, literature review, workflow management, data analysis. Tasks that consume time without necessarily requiring years of medical training.
If AI succeeds in those areas, physicians may gain something increasingly rare in modern healthcare: Time to think, to teach. And most importantly, time to spend with patients.
The irony is that as technology becomes more efficient, the uniquely human aspects of medicine may become more important rather than less.
What spine training looks like in the AI era
The implications may be most significant for the next generation of physicians. Medical schools and residency programs are already confronting questions that barely existed a few years ago.
How should trainees use AI? When should they trust it? How do they develop independent clinical reasoning in a world where answers are increasingly available on demand?
Dr. Zhou offered a framework that resonated throughout the discussion. “AI is a narrowing tool,” he said. “Whereas the human is an expansion tool.”
The observation highlights a fundamental difference between AI and clinical judgment. AI often works by narrowing possibilities. Physicians create value by recognizing possibilities that others miss. They challenge assumptions. They identify unusual presentations. They recognize when a conclusion that appears statistically convincing fails the test of clinical reality.
Those skills cannot be outsourced. At least not yet.
The experiment that made experts uncomfortable
Dr. Bydon described a recent experiment that illustrates both the promise and limitations of AI. Researchers used the same dataset to generate two scientific manuscripts. One was written by a human. The other was written using AI. Reviewers were not told which was which. The results surprised him.
“It did better than I thought it would, which was a little scary,” Dr. Bydon said. The AI-generated paper was coherent. Organized. Persuasive. It also fabricated references and introduced factual inaccuracies.
The technology was capable. It was not trustworthy on its own. That may be what makes AI fundamentally different from many previous healthcare innovations. Its mistakes are not always obvious. Sometimes they arrive wrapped in confidence.
The organizations that will get AI right
The panelists agreed on one point. Ignoring AI is not a strategy. Organizations that refuse to engage with it risk falling behind. But blind adoption carries its own dangers.
Dr. Bydon described healthcare systems making strategic decisions based on AI-generated analyses that failed to account for referral patterns, market realities and operational complexity.
Technology provided answers. Judgment was still required to determine whether those answers were useful. The organizations most likely to succeed may occupy the middle ground. They will embrace AI. They will also challenge it. They will automate aggressively. They will verify relentlessly.
“People who don’t use it at all will be left behind,” Dr. Bydon said. “People who use it unintelligently will be really up a creek.”
AI will improve. The models will become more sophisticated. The outputs will become more convincing. The capabilities will expand.
The question is whether healthcare’s ability to evaluate those outputs expands as well. Because for all the discussion about algorithms, automation and machine learning, the panel ultimately returned to a much older concept.
Judgment. AI can process more information than any physician. It can search faster. Summarize faster. Analyze faster. But medicine has never been a contest of processing speed.
It is a profession built on deciding which information matters, which conclusions deserve skepticism and which answers should be trusted. The future of spine care may not belong to the organizations that adopt AI first. It may belong to the organizations that remain disciplined enough to question it.
At the Becker’s 32nd Annual Meeting: The Business and Operations of ASCs, taking place October 29-31 in Chicago, ASC leaders, surgeons and healthcare executives will explore strategies to drive growth, enhance operational performance, navigate reimbursement challenges and prepare for the future of ambulatory surgery. Apply for complimentary registration now.
