"ChatGPT on Poop": How Radical Innovation Unleashed Citizen Doctors
AI is everywhere...and patients are taking it far beyond simple chats
So it has been a minute, as they say…I started this blog in July with the intention of writing every 2 weeks, and it is now September!
While I cannot make up for the past time, the summer pause brought an opportunity to reflect on a new phenomenon: the emergence of citizen doctors.
Vital Signs
Over the past three months or so, I've noticed an interesting pattern. Patients are using generative AI all-purpose chatbots to interpret their test results and then attempt to cobble up an intervention based on the AI inputs.
I called these patients “citizen doctors” because, just like citizen scientists or citizen policemen, they attempt to engage in the work that requires specialized training, experience, and expertise.
Wikipedia provides a nice review of citizen science, including the origin of the term, its current definition, and the evolution of the citizen science role. There is a typology to citizen scientists, with the most recent definition in activities described in 2022 as “co-creation and participatory approaches” and “contributory citizen science.”
Here are a few notable “citizen doctors” examples:
A 73-year-old man with CAD and ischemic cardiomyopathy (received stents and a defibrillator):
“Hi Dr. Druz,
ChatGPT identified several imbalances in my stool, so I would like to follow up when you are ready: absence of Lactobacillus, low Bifidobacterium, Candida overgrowth, low short-chain fatty acids (SCFAs), and reduced pancreatic elastaseâ?? Related to the 28-day course of amoxicillin I completed in May? Lowering beneficial bacteria and opening the door to yeast overgrowth and reduced gut fermentation? I welcome your help and wonder about enzyme/antifungal support. Thanks and warm regards.”
A 49-year-old woman with cryptogenic CVA due to PFO, idiosyncratic reaction to an antibiotic, and/or Eliquis with transient elevations in liver enzymes after the PFO closure procedure:
“After stopping ALL medicines (except metoprolol) and supplements, AST-ALT returned to normal, but in analyzing historic and very recent bloodwork, I see consistently high MCV and MCH, low RBC and RDW, and high ferritin. This, coupled with my history of PCOS and high cholesterol, makes me wonder about NAFLD and the benefit of a high-impact intervention like metformin (also high glycaemic protein)? Did we also test for hemochromatosis? I am trying to connect the dots! Thank you!!”
Bottom Lines
Generative AI is a radical innovation. Radical innovation is often used in a business or technology context to describe technology or a process that is fundamentally different from the current state and offers transformation through disruption.
I found this definition from “Digital Leadership” to be helpful:
“Radical innovation is a term that encapsulates a groundbreaking and transformative form of innovation. At its core, the definition of radical innovation lies in its departure from incremental improvements or minor modifications. Instead, it represents a profound shift in thinking, technology, or business models. This type of innovation introduces entirely new concepts, products, or services that challenge existing norms and redefine industries. The radical innovation meaning is characterized by its disruptive impact, creating a significant departure from the established status quo. In essence, radical innovation is not an incremental enhancement but a revolutionary leap, paving the way for novel solutions, approaches, and opportunities.”
In using generative AI, patients are radically innovating on the traditional process of medical testing and healthcare delivery by taking a shortcut to fulfill their jobs. To fully understand jobs, pains, and gains, and how those shape the emergence of innovative technology or process via products or services that are gain creators or pain relievers, check out the value proposition canvas on Strategyzer.
What often does not get discussed is that radical innovation produces both positive and negative externalities. Externalities are essentially unintended consequences. These terms are most often encountered in economics to explain how activity in one part of a business may produce an unexpected cost or benefit somewhere else outside of a company. Investopedia defines an externality as “a cost or benefit caused by an economic actor that is not suffered or enjoyed by that same actor.”
Generative AI's positive externality is enabling patient agency and access to test results interpretation. But a much more significant negative externality is a breakdown in the medical thinking process and violation of the checks and balances that underlie diagnostic and treatment decisions. This negative externality enables cognitive bias (Dunning-Kruger effect), where “citizen doctors” lean out of co-creation and collaboration, and lean in to shortcuts that will end up taking many down irrelevant rabbit holes.
Fortunately, my patients did lean into collaboration. I believe the significant positive externality of these generative AI experiences is patient engagement with a trusted physician…of course, provided that they have one who is willing to engage comprehensively and collaboratively.
What do you think? What have been your experiences? Please drop a comment and let’s discuss!