On the "Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine"

Comments from “Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine” (New England Journal of Medicine; March 30, 2023) include:

  • “The uses of artificial intelligence (AI) in medicine have been growing in many areas, including in the analysis of medical images,the detection of drug interactions, the identification of high- risk patients, and the coding of medical notes.

  • “GPT-4 is a work in progress … it can, for example, write computer programs for processing and visualizing data, translate foreign languages, decipher explanation-of-benefits notices and laboratory tests for readers unfamiliar with the language used in each, and, perhaps controversially, write emotionally supportive notes to patients.

  • “We have found GPT-4 to be extremely powerful, it also has important limitations.

  • “The system can make mistakes but also catch mistakes — mistakes made by both AI and humans.

  • "To what extent can the user “trust” GPT-4 or does the reader need to spend time verifying the veracity of what it writes?

  • “Chatbots will be used by medical professionals, as well as by patients, with increasing frequency.

  • “We speculate that GPT-4 will soon be followed by even more powerful and capable AI systems — a series of increasingly powerful and generally intelligent machines. These machines are tools, and like all tools, they can be used for good but have the potential to cause harm. If used carefully and with an appropriate degree of caution, these evolving tools have the potential to help health care providers give the best care possible.” 

 

OUR TAKE

GPT-4 and similar AI systems are trained with data from freely available Internet sources (e.g. medical texts, research papers, health system websites, podcasts and videos.) Systems trained with restricted data (from electronic health record system, medical schools or other sources) will likely provide increased accuracy and value to users. 

Data from neural sensors, collected from the brain, will expand the opportunity in this evolving medical technology market.

Concerns about tracking/privacy and hacking/security will likely lead to new types of medical laws and legislation.


Paul Dravis