AC LITERATURE

Dr. Anthony Chang

“I would worry about those people controlling AI before AI itself turning on us. Humankind’s greatest enemy is humankind.” 

Daron Acemoglu, Nobel Laureate in Economics 2024    

 

I was kindly invited to be part of panels at several events recently (Consumer Electronics Show and SIC Venture Studio Biotech Symposium, thanks to my good friends Rene Quashie and Harry Pappas as well as Paul Grewal) to elaborate on the future of AI in healthcare. I had some to reflect on this theme over the holidays, and here are some of my thoughts on the three dimensions (in three parts) of people, technology, and process in health AI for the upcoming years and decades. We will start with the people:  

Publication to Practice Chasm

Clinicians and data scientists/ technologists need to understand and appreciate each other’s domains much more than we have in the past, as most projects and publications do not reach real-world clinical practice (the AI in health “publication-to-practice chasm” is very wide). Data scientists would benefit greatly from spending some time with clinicians in their venues of clinics and hospitals; similarly, clinicians can better appreciate data science by participating directly in the data science work from the beginning of projects. All of us need to spend time for substantive discussions and sharing knowledge as a community of AI in health without intellectual hubris nor self-centered agendas.  

The Clinician-Data Scientist

There are probably more astronauts in the world currently (675 or 714 depending on US vs international definitions) than clinicians who are “bilingual in clinical medicine and AI”- dually educated and trained in both the discipline of clinical medicine and in applications of data science and AI in medicine and healthcare. We need to nurture many more clinicians to be “bilingual” or even “trilingual”- to have dual and triple domain expertises in not only clinical medicine, but also data science/artificial intelligence (and informatics) as well as the areas of implementation science, trans-disciplinary collaboration, and human factors in artificial intelligence to help guide AI in healthcare as leaders. The future AI in health will transition from computational to cognitive intelligence, and this transition will require “street smarts” from all seasoned clinicians. The cohort of bilingual clinician-data scientists will be essential to push the institutional agenda in health systems.  

AI in Healthcare Education and Clinical Training

Lastly, healthcare professional students and trainees are increasingly vocal (as they should be) about the immediate and profound relevance and importance of AI in healthcare education and clinical training both now and into the future. This inclusion of AI in education and training is not only teaching it as a subject matter but also using it as a technological tool to improve engagement and effectiveness in schools and health systems. Artificial intelligence, combined with other related technologies such as extended reality and gaming tools, will be a major asset in medial education and clinical training for not only physicians, but all stakeholders in healthcare that includes nurses and nurse practitioners, physician associates, pharmacists, and healthcare administrators. It is conceivable that AI can dramatically shorten the educational and training timeline in the future for all the aforementioned professionals.