I saw a patient recently with a new brain tumor. She came with an internet search that suggested she had five years to live. I cannot beat that predictive confidence. The best I could do was to poke holes in the assumptions that servers and algorithms had made, to question the arc of the future, and finally to tell her that the tumor was small, likely benign, with minimal effect on her life expectancy and we should just follow it to see how it behaves over time. She came to me expecting brain surgery and a get-your-affairs-in-order recommendation. She left for a celebratory lunch with her husband. I’d done something good.

Most of us consult the internet for a health problem, at first to learn more. Nearly inevitably, the search suggests that the symptoms could be something serious. Anxiety ensues, leading to a search for more information and then more anxiety, how life can even result in death – so true! – and finally ending in a rapid, degenerate spiral of existential dread. A neurosurgery office visit often proceeds from there.

I get it. The “now” paradigm of finding information is extremely attractive. Why wouldn’t any one of us go to our fingertips to learn more? I do. If a patient is taking a new medicine I haven’t heard of, I can look it up super fast. A consultant near where a near where a patient lives? Hello, Google. The Pubmed search for that rare condition or a new grading system for everyday pathology? Yup. We are learning creatures, and the internet serves our nature. However, I often spend a significant amount of visit time talking people off the precipice from their confidently preformed machine diagnosis, and then as they brace for the risky treatment that is definitely going to go badly, like all of the revenge blogs they’ve read about surgery that didn’t go well. PSA: The people who do well aren’t blogging about surgery. They’re out living their lives.

At every level of education, there were those couple of people who sat in the front of the room, wrote down everything the professor said, used multicolored highlighters, and asked questions about obscure minutiae in order to demonstrate their mastery of the knowledge in question, while sucking up valuable class time. Gunners could answer any question presented in multiple-choice form, produce detailed knowledge, properly formatted, on-demand, and elbow their way to the top of the class. But as junior medical students, out of the classroom, a lot of these people struggled to take care of actual patients. They couldn’t put together a diagnosis, couldn’t engage with patients and staff, couldn’t use their knowledge to solve open-ended problems, couldn’t get comfortable with the uncertainty. Google is a gunner. It has a vast depth of information but will never touch a patient. It has parameters, but no context. No internet search hears a patient’s vocal tone as they express their disabling symptoms, nor touches them to assess their strength, gait, cranial nerves or reflexes, listens to their heart and lungs. There will be no review of images and no full-throated assessment, no assurance that everything is going to be fine, and here are the next steps.

Please do not misunderstand me. I am not a Luddite. Technology is amazing. It has made me a better surgeon. It has helped patients with their own health and will affect medical practice in ways we haven’t even imagined yet. One example is neuronavigation, now a standard part of neurosurgery. It allows me to point to a spot on a patient, and a computer tells me where that is on a CT or an MRI within a couple of millimeters. It’s a super helpful enabling technology that has changed the way I work, but surgeons don’t just press the button and walk away to get coffee. Technology is not yet an autonomous operator. The first part of my job a robot will learn to do by itself is to drill a burr hole. That’s about my easiest task as a neurosurgeon. Nearly everything else about neurosurgery is profoundly human — that hard conversation about prognosis, the happy occurrence of relieved suffering, even how to know that I’ve achieved surgical objectives safely and it’s time to close. Robots need to learn to drive, fold laundry, and do manual disimpaction first, before attempting brain surgery.

Surgical craft is still an analog activity. It’s my hands and tools interacting with patients’ spines and brains, a shared experience, though from very different perspectives. For patients, it’s pain and relief, restored function, and at times setbacks from complications, all within a milieu of expectation. Willingly subjecting oneself to pain and risk for the prospect of a better life afterward is uniquely human. I am not aware of any other species that does this. As a surgeon, I make inferences and decisions based on what patients tell me and on observed digital and sensory information: visual, auditory, tactile, sometimes olfactory. Usually not tasting though.

My elder daughter nudged me to play the cello for a while, so I finally rented one. The cello has been around for about 500 years and requires no intermediary technology. My hands, ears, brain, bow and instrument are all that are required to make a sound. I bought a book, watched YouTube exercises, studied diagrams from image searches for months. Learning was really slow. It accelerated considerably once I found a teacher. She shows me how to hold the bow, gives me exercises to train my ear, points out quirks of the instrument, and I’m learning to read music in bass clef. Without her, the unfiltered deluge of information about playing the cello was an impediment to my learning, opaque and overwhelming. With her as my guide, tech is a jet boost to my playing. I have apps that measure the effectiveness of practice, check my intonation against an objective standard, and help me with intervals and theory. My left hand is greatly aided by technology, far beyond where it would be without. On the other hand, literally, there is no bowing app to guide my right hand. The low C string is sonorous but unruly when I bow it. I’m uncertain how I play it will affect its sound. Bowing a string is a real world action, something out of reach for an app.

Surgery is just as analog as the cello is, except that it’s an instrument I know how to play. I am the teacher who’s been doing it for 25 years, the one with the accumulated knowledge. There is a lot of information about it online, free and easily accessible, but of variable quality. For surgery – or nearly any health problem — patients need a live guide to help put it all together, because humans are still the effectors of health care, the bow to the string. I’ve come to embrace the vast repository of the internet and all sorts of enabling technologies as side-by-side learning and guidance tools for my patients and for me. I can point them to trusted sites with diagrams and videos so they can better understand what surgery is, what it can and cannot do, outside of our limited time together face to face. For me, tech is an indispensable aid to help me do my work better – before, during, and after surgery.

Patients sometimes apologize for searching online, but it’s completely natural. The compulsion to search is a hunt for some kind of truth, a certainty that can comfort us – that everything is going to be fine. But information is like laundry – a lot more useful when it’s clean and sorted. Machines help, but it’s still humans doing the folding, and no combination of ones and zeroes will result in a black-and-white answer. Raw information has limited utility, possibly harmful, but it seems more objective and easier to trust than one fallible person like me, who could be tired or having a bad day, could be giving bad or biased information, who can’t even play the cello very well. Surgery is tactile and capricious, effective but imperfect, and technology-enhanced yet human and manual. I cannot answer with complete certainty how it will go or what the outcome will be, but I want my patients to do as fabulous as they want to do, and I’m going to do my best to make it so. Most any other clinician I know wants the same for their patients too. That’s for sure.

Patrick Connolly is a neurosurgeon.