How AI Is Helping Surgeons in Seoul Perform Operations With Sub-Millimeter Precision on Patients They Never Meet in Person

In a contemporary operating room in Seoul, there’s a certain silence. It was more mechanical than the gentle, cautious silence of older hospitals. a slight hum. A keyboard tap. The surgeon watches a screen that displays a real-time human heartbeat while seated at a console, sometimes in the same building and other times in a different city. A few inches are moved by their hands. Fractions of a millimeter are moved by a robotic arm inside the patient. Even for those who work there on a daily basis, the scene still seems a little unreal.

More than most nations, South Korea has embraced surgical robotics, and Seoul in particular has turned into an odd kind of testing ground. Hospitals that have partnered with domestic AI companies have started performing procedures without the surgeon of record physically present. A local team prepares the patient. A remote specialist uses a system trained on thousands of hours of previous operations to guide the cutting, suturing, and clipping. The robot remains motionless. At hour six, it doesn’t grow weary. No matter how skilled the surgeon behind the glove is, it moves with a precision that human muscle cannot match.

The learning is what sets this moment apart from, say, the da Vinci era of robotic surgery. The surgeon’s hand is no longer merely extended by these systems. They are observing. Changing. Results from a system called SRT-H, which was trained on surgical videos in the same manner as ChatGPT was trained on text, were released by Johns Hopkins researchers last summer. It completed 17 consecutive tasks, including a full gallbladder removal on a lifelike model, without human assistance. During the process, it even reacted to voice corrections. “Grab the gallbladder head.” It complied.

Such research is applied to remote practice by the Seoul ecosystem. Without taking a plane, a Gangnam surgeon can perform surgery on a patient in Busan or possibly Manila. The consequences are truly enormous for war zones, rural hospitals, and nations lacking sufficient specialists. And, to be honest, a bit unnerving.

Something about the old surgical social contract—the notion that you can look someone in the eye when they cut you open—is subtly rewritten here. Patients might not mind. When a grandmother in a small town receives top-notch retinal treatment from a Munich-developed system that can precisely measure 15 micrometers on a retina that is only 200 micrometers thick, she might never question who was holding the syringe. She will simply check again. That could be sufficient.

The questions continue to pile up, though. When a procedure’s autonomous step goes wrong, who is responsible? What happens to the gradual, supervised training that young surgeons have always relied on—learning by doing, observing, and making minor, fixable errors under supervision? What will a surgical residency look like in 2035 if robots are able to perform better than humans?

It’s difficult to ignore how quickly this is moving. Ten years ago, it was unthinkable that a robot could carry out a task completely by itself. Nowadays, the only meaningful discussion is not technical but rather regulatory. AI-guided procedures are being used in Japanese hospitals. Live-animal trials are being prepared in German labs. Within a generation, remote, AI-assisted surgery will be commonplace, according to Seoul’s infrastructure.

There’s a weird double feeling as you watch this happen. relief, as those who would have perished or lost their vision will now be able to see. Beneath it, there’s a quieter feeling that the human hand is gradually being displaced. not thrown away. recently relocated.