Book Review: The Algorithm Will See You Now

“The Algorithm Will See You Now” is the first novel written by JL Lycette. It is a remarkably polished and engaging thriller that delves into the promise and pitfalls of artificial intelligence, corporate medical greed, and the ever-increasing tension between the science and the humanity of medicine.

The author’s day job persona is hematologist-oncologist Jennifer Lycette, MD (UCSF undergrad, UW medical school), so the book’s medical and hospital details are accurate.

I admit that I bought this book purely because it sounded kind of interesting and the Kindle version was on sale for just $0.99. Cheap Kindle books are usually amateurishly self-published junk, but not this one — it’s a gem.

The story is set in 2035. Surgery resident Hope Kestrel, MD works for Seattle-based health system Prognostic Intelligent Medical Algorithms, where she has earned the official title of high resident, which PRIMA uses instead of chief resident, in its OASIS unit (Oncologic and Surgical Intervention Success.)

Medical technology has come a long way by 2035, as diagnosis and treatment is performed at PRIMA entirely with AI and a voice-powered chart assistant called Osler. The PRIMA health system’s AI determines upfront whether a patron (its word for “patient”) will benefit from a particular treatment or is instead likely to be a “non-responder.” The goal is to “optimize.” Or as Hope describes, “The AI frees both patients and doctors from the fallacy of choice. The algorithms are more trustworthy than people.”

An anonymous podcaster is critical of PRIMA, however. The algorithm is a mysterious black box. Patrons don’t pay deductibles or co-pays at PRIMA, but aren’t eligible for services for which AI predicts they will be non-responders, and rumor is that PRIMA licenses the technology to insurers who use it to deny coverage. The podcaster summarizes, “The most dangerous lies are the ones that use the truth to sell themselves.”

Hope is fully behind the AI system, declaring that any doctor who won’t embrace algorithm-directed care is as dangerous as surgeons in the 1800s who mocked sterile technique because they didn’t believe in germ theory.

The story continues with each key player, including the most vocal AI proponents, struggling with the health issues of loved ones and themselves, for which AI seems heartless and even wrong based on accidentally or intentionally flawed data training. They don’t face the patrons who are non-responders, but they can’t avoid seeing how the system works for those they love.

There’s also the business angle, as power-hungry PRIMA executives are working on the acquisition of Seattle’s biggest medical group, with plans to go regional, national, and then to be positioned to run a privatized Medicare system.

PRIMA’s medical residents are constantly scored, reassigned, and threatened for various transgressions, one of which is having non-responder patients for which blame must be assigned since nobody will admit that the algorithm could be wrong.

A recurring theme is the accuracy of the AI. Is it wrong one time out of 10,000, as PRIMA touts, or is only 95% accurate as behind-the-scenes data seems to suggest, leaving one patient in 20 to go untreated or to receive treatment unnecessarily that presents side effects and risks? And what are the clinical implications of overfitting, where the computer thinks it sees a pattern that doesn’t exist in the real world?

An uncomfortable moment ensues in a medical staff meeting of the practice that PRIMA hopes to acquire. A skeptical doctor expresses doubt about the AI’s accuracy, then when pressed to state his own accuracy rate, responds instead with a quote from Sir William Osler: “Medicine is a science of uncertainty and an art of probability.” It’s a valid point — doctors really don’t know their own accuracy, so to criticize that of the machine is a tenuous position.

The book is a compelling read that ricochets off the dangers of big data, corporate ambition, and what doctors are supposed to do when told that AI will be taking over their decision-making. It is really written like a movie screenplay, mostly “show don’t tell” scenes rather than exposition, playing for me like an episode of “Black Mirror.” I bail out early on most novels, including those of big-name authors, when I run out of patience for irritating writing style, sloppy editing, and inconsistent character behavior, but this one was remarkably well crafted and hard to put down.

The author wrote most of this book in the pre-2020 dark ages before COVID-19, ChatGPT, and health systems that are fueled by acquisitions racing each other to become national providers to compete with corporate giants such as Amazon and CVS Health. It is not preachy or prescriptive about any of the touchy healthcare topics that its characters are living through 10+ years in the future, which requires the reader to decide whether the tale it tells is entertaining or cautionary.

The author says the idea for the novel came when she saw IBM Watson touted for oncology before it fizzled. She practices as a rural community oncologist, and while she hopes to see the day when precision medicine’s abilities expand enough to cover more than a few patients, she worries about under-resourced systems and disparities in care. Her second book, due in November 2023, is a prequel to “The Algorithm Will See You Now.”