October 17, 2025
Dear Interested Readers,
Is AI the Answer?
I wonder if I attend more Zoom meetings than the average eighty-year-old. There are usually at least two and sometimes more Zoom meetings a week on my schedule. My Zoom meetings save me a lot of time. I can be involved with organizations in Massachusetts without subjecting myself to the dangers and delays of driving in the greater Boston environs. Even in my small town environment, a Zoom meeting saves me the trouble of driving across town and back. As we all know, time is a scarce and valuable commodity.
It has been my experience that, on boards and committees, the most challenging position to fill has been the secretary’s role. No one I have ever met has said they loved being responsible for the meeting’s minutes. Now things are changing. If allowed, Zoom will take the minutes of your meeting! The minutes still need to be read, possibly corrected, and approved, but I have yet to hear anyone object to assigning this traditionally onerous task to a machine.
When I was trying to lead an organization toward better quality and patient satisfaction through more efficient service, while improving our employees’ work environment and job satisfaction, I became fascinated by the literature on the varieties of acceptance of innovation. You may be more familiar with Kubler-Ross’s stages of grief and acceptance than you are with the stages of adoption of innovation, but I see some similarities. I will argue that there may be a Kubler-Ross explanation for why some of us resist what is new because accepting the new often requires going through the grieving process as we mourn the loss of a status quo that contains comfortable familiarity, potential control benefits, and possible loss of investments that have been devalued by what business guru Clay Christensen called disruptive innovation.
I have recently relived the reluctance to accept an offer of innovation. One of the charitable non-profits with which I am affiliated was offered the opportunity to apply for a grant that would automate our work. For years, our process has included a lot of paperwork and the reliance on file cards that are sometimes lost or unavailable when the information they contain is needed. This opportunity to transition to a more secure, computer-automated operational platform sparked a spirited discussion. Balanced against the potential benefit were the transitional efforts required. I suspected that some potential users were not comfortable with their computer skills, since our average age is at least late sixties and may be in the seventies. Still, those opposed to the idea did not base their objections on concerns about their individual lack of computer skills. They argued that there may be hidden costs to implementation and maintenance, that the work required for implementation would distract us from our larger mission, and that changing what had worked “just fine” for over twenty years was unnecessary. Even though almost half the group favored applying for the grant, the idea was “tabled.” So we will continue to rely on a shoebox full of index cards and reams of paper documentation, including multiple manual reports that could be easily generated by effective automation. Those suggesting change, even beneficial change, should beware the power of the status quo and vested interests. In his 16th-century book The Prince, Machiavelli warned that innovators should expect resistance to their better ideas and products.
Machiavelli’s exact words were:
“It ought to be remembered that there is nothing more difficult to take in hand, more perilous to conduct, or more uncertain in its success, than to take the lead in the introduction of a new order of things.”
In the nineteenth century, Ralph Waldo Emerson had warnings for innovators and those who wanted to be change agents. So I asked Google about Emerson’s warnings and got back more than I expected:
AI Overview
Ralph Waldo Emerson’s warning to innovators is captured in the quote: “Beware when the great God lets loose a thinker on this planet. Then all things are at risk”. In his writings, Emerson explains that the radical individualism and self-reliance of a true genius can overturn existing knowledge and dismantle the status quo.
This warning is not an admonishment but an acknowledgment of the revolutionary power of genuine, self-reliant thought.
The dangerous potential of the true thinker
In his essay on “Circles,” Emerson describes how a “new generalization” from a great thinker can upend established fields of study and human understanding.
- Scientific upheaval: A new theory can instantly render entire fields of science obsolete.
- Literary reevaluation: Literary reputations and “eternal names of fame” can be “revised and condemned” by a new, powerful voice.
- Societal disruption: The hopes, morals, and religions of an entire society are at the mercy of the new perspective brought forth by a revolutionary mind.
The innovator as a destructive force
Emerson saw the innovator as a force of nature, similar to a wildfire, that respects no boundaries. The “conflagration” of a new idea can burn through a great city of knowledge, leaving nothing safe. For innovators, this means:
- You will be misunderstood: In his essay “Self-Reliance,” Emerson asserts, “To be great is to be misunderstood”. The true innovator’s work will challenge and provoke society, not receive immediate acceptance. [This sounds like Machiavelli.]
- Imitation is a trap: The path to greatness is not to copy others. As Emerson wrote, “imitation is suicide”. Innovators must trust their own unique perspectives, or “private chart,” and be guided by it, not by the wisdom of others.
- Society will oppose you: Society promotes conformity and will push back against the radical independence of a new idea. The innovator must possess the independence of spirit to stand alone amidst the crowd.
The reward for the genuine innovator
Despite the warnings, Emerson championed this kind of individual genius, viewing it as a divine force. He argued that the genuine innovator will discover that external fame is less valuable than the internal reward of creation itself. The ultimate motivation is not the public’s approval but the integrity of one’s own mind and the joy of contributing something new.
Then I decided to use AI to provide you with an easy-to-read description of the variation in people’s acceptance of innovation. So I asked Google, “What are the classical stages of adoption of innovation beginning with early adopters?” Google replied:
AI Overview
The classical stages of innovation adoption, beginning with “early adopters,” are Early Majority, Late Majority, and Laggards. These stages, according to Everett Rogers’ Diffusion of Innovations theory, follow the “Innovators” and “Early Adopters” who are the first to embrace a new idea or product.
- Early Majority: These users are pragmatic and adopt an innovation after seeing that its benefits have been proven by early adopters.
- Late Majority: This group is skeptical and only adopts a new product or idea when it has become mainstream and is widely accepted.
- Laggards: The last group to adopt, they are often resistant to change and may only adopt an innovation out of necessity or as a last resort.
When I applied the definition to myself, I found that I was probably a “late” early majority adopter. My first cell phone was the size of a carry-on bag for an airline trip. Still, I was reluctant to give up my Motorola flip phone for an early iPhone, even after an enthusiastic patient spent most of the time that should have been devoted to his physical demonstrating to me all the wonders of this miraculous piece of technology. I carried a Sony Walkman on my runs, along with a Garmin GPS, long before any of my running buddies, and I was always an early adopter of the latest innovations in running shoes. It took me a while to get an iPhone, but I was quick to get an iPod and an iWatch. I spent a small fortune on an early VHS player and a huge video camera, and I was a sucker for any new digital camera. I am not an innovator. My “adopter’s” profile seems mixed, but I am not a “laggard.”
I was slow to get a laptop computer. I got my first in 1994, long after my wife had become an enthusiastic Apple user and my kids were into video games. It wasn’t that I was afraid of computers. I was always impressed by people who could use computers, but a lack of keyboard skills was a barrier for me and limited me to “read-only” use. I was envious of those who had the skills to use the new technology.
Back in the sixties, while I was in medical school, a neighbor who was then a student at Harvard Business School introduced me to the wonder of large mainframe computers. After he graduated, he sold time shares on huge mainframes to businesses that were too small to own their own. In that era, computers just weren’t available to everyone, even if you could type. Harvard Community Health Plan opened its doors in 1969 with a computer-based medical record developed at the MGH in 1964 by Dr. Octo Barnett. If you click on the link, you can read:
In 1964, Dr. Barnett was recruited by Mass General to head the Hospital Computer Project, an ambitious plan to implement a large hospital information system. From this project, the Laboratory of Computer Science was born. He remained the director of the lab until his retirement in 2012.
Dr. Barnett was co-developer of COSTAR, one of the nation’s first computerized electronic health records, and of DXplain, one of the best known, widely used diagnostic decision support systems. The programming language known as M, Caché, or MUMPS, which remains in widespread use today as the foundation of numerous clinical systems, was invented in LCS under his leadership.
Our COSTAR medical record was “read only.” For each patient, we were given a “Printout,” a formatted summary of the patient’s demographics, social history, problem list (“SOAP”-ala Weed methodology was an innovation that slowly gained acceptance), and recent pertinent encounters. Along with the printout, there was a form containing diagnostic codes for frequent “encounters,” as well as places to record new problems. For each item, we could choose to write 180 characters or dictate a note. We had a large computer terminal on our desks so that we could dig deeper into the record if we were so moved. I was rarely so moved because I was busy trying to stay on schedule. Like many clinicians who today spend their evenings documenting visits on their computers at home, I filled out most of my “encounter forms” at home after leaving the office, which were eventually entered into the medical record by an “inputter.”
By the late 80s, but still pre-Microsoft Windows, Harvard Community Health Plan was working with Ross Perot’s company, Electronic Data Systems, and other IT organizations to develop a new “Automated Medical Record System (AMRS).” We sunk millions into the project, and even opened a health center in Burlington, Massachusetts, as a beta site to test the system. AMRS was clunky, but some younger doctors with tech backgrounds and excellent typing skills were “super users,” or what would be called “early adopters.” I lived in fear because back in the eleventh grade, I had dropped my typing class because I had convinced my guidance counselor that since I was going to be a doctor, I would always have a secretary, and did not need to learn how to type. We had an anatomy elective where we dissected a cat in our progressive (for South Carolina) high school, and I argued to my counselor that the anatomy class was a much more useful course for me than typing.
By the early nineties, I was leading the Physicians Group and sitting on the management committee of HCHP. Because one of my colleagues on the Physicians Council was a true “early adopter,” I was beginning to learn my “seek and you shall find” typing method. He convinced us to try the new concept of “email,” and we bought an internal messaging system called “Gold Mail,” and I began to type. It became apparent to us that AMRS was a decade too late and that we needed a commercial product. Epic was in its earliest version, and by the mid-nineties, we were the first New England medical organization, practice or hospital system, to use Epic. Because of my lack of typing speed, Epic was always a challenge for me. Still, I got a little relief when another tech-savvy colleague, a “superuser” I had known since he was a med student and I was an intern, showed me how to open “properties” and turn on the “copy and paste” option. My other “late early adopter” attempt was to couple “Dragon” voice recognition capabilities to Epic. It was clunky and hard to “train,” but by the time I was CEO, the system’s capabilities enabled us to couple both voice recognition and the UpToDate medical text with our latest version of Epic. It was too late for me to enjoy these new efficiencies. Still, I hoped it would make our clinicians’ lives easier and improve the care we could offer our patients by freeing our clinicians from some of the time-consuming aspects of medical record keeping. One of my greatest disappointments was learning how limited our systems were in looking at populations. Epic, at the time, was set up from the perspective of individual records. A separate program was required to “mine” population data. When we finally invested in a data warehouse, I learned from our technical staff that it was better suited to “mine” financial data than clinical data.
From my perspective, the most significant problem that exists for patients in primary care is access to care when it is needed. As a clinician, my direct telephone number was always on my business card. When email came into generalized use, I put my personal email on my card. I reasoned that I would rather my patients have easy access to me than have them go to an ER or be put through needless hoops to have a colleague learn what I already knew. Any clinician who has a long-term relationship with a patient has a “sense” of the patient that goes beyond words and can’t easily be discerned from the medical record. This sense can be the difference between reassuring the patient, ordering a useless test, or unnecessarily hospitalizing the patient. Sometimes it is the difference between being “deaf” to a problem and sending a patient home rather than doing a critical test or advising a hospital admission. Not all medical knowledge is in the medical record. Some colleagues told me I was crazy to give my patients easy access, but I never had a patient abuse it, and the few calls that came through were ones I was glad to get.
I also never sat facing the computer with the patient looking at my back during an appointment. I always felt that reading the patient’s face and having the patient see my reaction to their concern was more important than just hearing their voice or them watching my back as I tried to type. There was a cost. I was always behind in my record keeping, and my workdays extended late into the evenings and weekends, time I should have spent with my family. Recently, I was delighted to hear a friend say she was surprised during her first visit with her new PCP because the doctor looked at her, not the computer.
For years now, the most significant problem from the perspective of primary care clinicians has been the onerous administrative tasks, primarily doing the documentation required in a fee-for-service system of finance. I think it is safe to say that many doctors spend more time documenting their services than spending time talking with or thinking about their patients and their concerns. For a long time now, I have anticipated the day when our exam rooms would be fitted out to take advantage of the new options that should be available through the thoughtful application of AI to patient care.
This week, I came across some early reports that may be evidence that my dreams of an effective integration of AI into patient care are coming true. The first was a report in the Boston Globe that Mass General Brigham was investing in AI as part of its strategy to head off the unionization of its doctors who have become frustrated with the demands placed on them because of administrative tasks and the difficulty a system as rich and as powerful as Mass General Brigham has in hiring enough doctors to provide adequate access for patients and support to practice for clinicians. Deep into an article published in May entitled “Amid union vote, Mass General Brigham to invest nearly $400 million in primary care,” we read:
Over the past year, MGB said, it has created more than 90 additional full-time support staff positions across both Brigham and Massachusetts General Hospital, the system’s two flagship hospitals. That includes jobs for medical assistants, practice assistants, and medication access coordinators.
MGB said it has hired an additional physician and is seeking to hire three more doctors to cover leaves and vacancies.
In addition, MGB said, it held a job fair to attract medical assistants and drew more than 75 candidates, twice the usual number.
The system said it has also made artificial intelligence tools available to all attending physicians to make it easier for them to write clinical notes after appointments with patients.
In an article this week by the same reporter, Jonathan Saltzman, entitled “MGB is turning to AI to ease shortage of primary care doctors. Some of them don’t like it”, which suggests that there is the usual resistance from the “laggards” to innovation, or those who have another political agenda.
We read:
Mass General Brigham has turned to artificial intelligence to address a critical shortage of primary care doctors, launching an AI app that questions patients, reviews medical records, and produces a list of potential diagnoses.
Called “Care Connect,” the platform was launched on Sept. 9 for the 15,000 MGB patients without a primary care doctor. A chatbot that is available 24/7 interviews the patient, then sets up a telehealth appointment with a physician in as little as half an hour. MGB is among the first health care systems nationally to roll out the app.
Several primary care doctors at MGB, however, described AI as a sideshow, a distraction from what they contend is the real solution to the shortage: providing the pay and working conditions that will attract more primary care physicians to the state’s largest health system and keep them from leaving.
What MGB seems to be developing is not precisely how I would employ AI, but I think what they do will be interesting to follow. I hope it will be a positive step toward an improved care experience for patients and an improved work experience for clinicians, whether they ever succeed in forming a union or not. Despite the fears some in the AI world have that AI will eventually lead to the extinction of the human race, I want to believe in the positive possibilities of scientific progress and our ultimate control over the tools mankind invents. I am a self-diagnosed “Pollyanna.”
Perhaps even more encouraging than the Globe reports on using AI in primary care at MGB was an article published in JAMA Network Open, which describes AI use in the exam room —closer to my vision of its utility. The article is entitled “Use of Ambient AI Scribes to Reduce Administrative Burden and Professional Burnout.” The lead author is Kristine D. Olson, MD, MSc, from Yale. In the article, we learn:
Question What is the association of using ambient artificial intelligence (AI) scribes with clinician administrative burden, burnout, time documenting after hours, and time and attention for patients?
Findings This quality improvement study of 263 physicians and advance practice practitioners across 6 health care systems found that after 30 days with an ambient AI scribe, burnout among those working in ambulatory clinics decreased significantly from 51.9% to 38.8%. There were also significant improvements in the cognitive task load, time spent documenting after hours, focused attention on patients, and urgent access to care.
Meaning These findings suggest that AI may have promising applications to reduce administrative burdens for clinicians and allow more time for meaningful work and professional well-being.
I will leave a deeper exploration of the article to you. Still, I think I have made my case that there may be relief on the horizon from AI if you are suffering from issues of patient access, clinician distress, and inadequate organizational resources. I don’t know whether the Trump administration is interested in exploring the benefits of AI as part of its approach to Making America Great Again, but I do think it could be a huge boost across the entire healthcare spectrum toward the dreams of health equity and the Triple Aim. History shows that innovations must be explored with both successes and failures as we seek continuous improvement. Ultimately, if successful, they are attached to a business plan to facilitate widespread utilization. If our government is unwilling to invest or be a partner in advancing AI in medical practice, I hope that some of those rich beneficiaries of the One Big Beautiful Bill will seize the opportunity to make a lot more money while improving our health. I accept that in our culture, financial opportunity plays a key role in innovation and progress.
Enjoying A Dry Fall
I should not complain about our lackluster fall colors. When I look closely, there are some beautiful displays of color. It’s just that we don’t have the banquet of color that I look forward to enjoying for a few days each year, as the transition from all the pleasures of summer to the dark, cold, short days of winter unfolds, as we move toward the inevitable occasional beautiful snowfall and adventures on the ice.
My neighbor, Peter Bloch, an accomplished drone photographer, is once again the source of this week’s header. The picture that I lifted from one of his recent videos is from Grafton Pond. In a post-glacial environment with plenty of natural ponds and lakes, like the one I live on, Grafton Pond is an exception. It was created behind a dam as the water source for the city of Lebanon and the surrounding area. Because of the severe drought in New Hampshire, the pond’s water level is at a record low, and Lebanon has imposed water-use restrictions. Dartmouth Hitchcock Medical Center is in West Lebanon. I don’t know if the Medical Center has water concerns, but I do. The drought has me worried as I watch the water level in my lake fall and hear that people’s wells have gone dry.
The lower water level in Grafton Pond has revealed stone walls in the fields that were flooded when the reservoir was created more than a hundred years ago. I suggest that if you find today’s header interesting, you watch the four-minute video it was taken from. The concern in my mind is whether our draught is a rare event in natural variation, or is it the harbinger of further occurrences ahead as we get deeper into the changes in weather and health wrought by the global warming that is the result of our addiction to fossil fuels which our government denies as it tries to convince us of its wisdom with ideas like Tylenol as one of the causes autism. We do live in interesting times.
My advice to you is to be out and about as much as possible this weekend so you can enjoy whatever fall opportunities are available to you. My wife and I will be gathering with others on our Town Common for a second “No Kings” event. If there is a “No King” event near you, I urge you to attend.
Be well,
Gene
