Which healthcare jobs will AI eliminate? And which will it leave intact—or create from scratch?

AI won’t displace many patient-facing clinical workers over the foreseeable future. However, it’s certain to cull the healthcare workforce of roles heavy in documentation duties, administrative tasks or diagnostic dot-connecting.

What’s more, the next 10 years will see the technology’s job-replacement rate largely determined by non-technological factors. These will include still-nascent puzzle pieces like regulatory frameworks, liability laws and the general public’s trust in autonomous medical AI. 

On the other hand, new opportunities will arise for those who attain proficiency in AI implementation, customization and ethics. 

So predicts medical AI expert Jesse Pines, MD, MBA, of George Washington University, Drexel University and US Acute Care Solutions, at which he serves as chief of clinical innovation. 

AI is highly unlikely to swipe jobs “demanding high-stakes human connection, fine motor skills and complex decision-making such as nursing or surgery,” Pines writes in commentary published by Forbes July 3. 

That said, he adds, most of today’s healthcare professionals would do well to “cultivate AI literacy to understand its capabilities and limitations.”

Is your job hanging low or hovering above? 

The healthcare job titles facing the most risk for replacement by AI are six in number, Pines believes. 

These are human scribe, medical coder, appointment scheduler, front-desk receptionist, insurance verification specialist and pharmacy technician. 

Meanwhile, eight titles make Pines’s list of roles safest from the AI scythe. Among these are registered nurse, paramedic/EMT, mental health therapist, midwife, home health aide/certified nursing assistant and dental hygienist. 

Pines also calls out three—and only three—physician-level jobs as highly secure: surgeon, emergency medicine specialist and dentist. 

It’s not clear from this if he means to imply that other kinds of doctors should watch their backs for AI’s encroachment on their turf.

New jobs now taking shape 

As for work opportunities brought on by the advancement of AI in healthcare, Pines discusses three.

1. Clinical AI implementation specialist. 

These workers are translators between technology teams and clinical stakeholders, Pines explains, adding that they oversee deployment, adoption and ongoing evaluation of AI tools. 

“The role typically requires a clinical background: nursing, pharmacy, respiratory therapy or allied health combined with training in health informatics and change management,” he writes. 

• The average salary for a clinical AI implementation specialist, he reports, citing indeed.com data, is $70,000 to $100,000 per year. 

2. Healthcare AI ethics and governance analyst. 

“As AI systems take on consequential clinical and administrative roles, health systems will need dedicated professionals to evaluate systems for bias, fairness, safety and regulatory compliance,” Pines states. 

AI ethics and governance analysts in healthcare review algorithm performance across patient subpopulations, maintain documentation for regulatory audits, design clinical validation frameworks and advise leadership on risk, he notes. 

• The average annual salary for an AI governance analyst is $141,139, Pines says, naming ZipRecruiter as his source. 

3. Health AI data scientist/clinical data engineer. 

Because health systems generate massive stores of clinical data, they need specialized experts to curate, label and transform raw inputs into training datasets and validated AI models. 

“Clinical data scientists and data engineers with healthcare knowledge are among the most sought-after technical professionals,” Pines writes. These roles require proficiency in Python or R, SQL, machine learning frameworks such as TensorFlow or PyTorch, and a working understanding of clinical terminologies including SNOMED, LOINC and HL7 FHIR, he points out. 

• Going by ZipRecuiter, the average annual salary for a healthcare AI data scientist is $122,738, Pines shows. 

At present, the worst stance healthcare workers can take toward workplace AI is “passive avoidance, such as waiting until the tools are mandated and then scrambling to adapt,” Pines writes. 

Better strategy: 

“Seek out opportunities now to interact with AI systems in your clinical area, whether that means participating in a pilot program, attending a CME course in clinical informatics or simply reading peer-reviewed literature on AI performance in your specialty.”

Get the rest.