Guardrails, accountability, or external scrutiny are necessary companions to Artificial Intelligence. A simple reactive AI that could simply rewrite the information in records using the principles of Plain Language would seem to have few, or even no, requirements for guardrails, accountability, or external scrutiny, whereas AI that provides the facility for critical thinking probably does?
Barry D. Weiss, writes in “Health literacy, a manual for clinicians, part of an educational program about health literacy” (Chicago : AMA Foundation, American Medical Association; 2003) that “there are many reasons why patients do not understand information that is given to them. Key among them is inadequate health literacy — an individual’s ability to read, understand, and use healthcare information to make effective healthcare decisions and follow instructions for treatment.
“The need for today’s patients to be health literate is greater than ever, because medical care has grown increasingly complex. Inadequate health literacy affects all segments of the population, although it is more common in certain demographic groups, such as the elderly, the poor, members of minority groups, and recent immigrants to the United States. The economic consequences of limited health literacy are considerable, estimated to cost the United States between $50 billion and $73 billion per year.”
Organizations that manage patients’ access to medical records are beginning to provide AI facilities to help the patient to navigate their way through the data and information that they find within their medical records. Here are a few thoughts on safeguarding the use of AI by patients who have access to their records.”
“AI is frequently grouped by its capabilities (power and scope) or its technical/business functions:
“Reactive Machines* The most basic form of AI. They have no memory, do not store past experiences, and only react to current inputs in real time (e.g., IBM's chess-playing computer, Deep Blue). (*Machine - any device that uses power (such as electricity, steam, or human effort) to perform a specific task)
“Limited Memory AI: Systems that can look at past data and short-term history to make better decisions. Most modern AI—such as self-driving cars, spam filters, and virtual assistants—falls into this category.
“Theory of Mind AI: A theoretical level of AI that would understand human thoughts, beliefs, intentions, and emotions. This type is still in the research and development phase.
“Self-Aware AI: A hypothetical, futuristic AI that possesses self-awareness, consciousness, and human-like intelligence or emotions. It does not exist in the real world
Plain language breaks compound sentences into short, direct sentences that contain only one idea. Plain Language can also replace complex vocabulary with common, everyday words. Keep the exact same facts and core meaning for general low literacy. Plain language seems to be a simple reactive machine and would seem to have few, or even no, requirements for guardrails, accountability, or external scrutiny, whereas AI that provides the facility for critical thinking probably does?
(“Critical thinking is the ability to recognize problems, to find workable means for meeting those problems, to understand the importance of prioritization and order of precedence in problem-solving, to gather and marshal pertinent (relevant) information, to recognize unstated assumptions and values, to comprehend and use language with accuracy, clarity, and discernment, to interpret data, to appraise evidence and evaluate arguments, to recognize the existence (or non-existence) of logical relationships between propositions, to draw warranted conclusions and generalizations, to put to test the conclusions and generalizations at which they arrive, to reconstruct their patterns of beliefs on the basis of wider experience, and to render accurate judgments about specific things and qualities related to health in their everyday life.” [Wikipedia])
HIFA profile: Richard Fitton is a retired family doctor - GP. Professional interests: Health literacy, patient partnership of trust and implementation of healthcare with professionals, family and public involvement in the prevention of modern lifestyle diseases, patients using access to professional records to overcome confidentiality barriers to care, patients as part of the policing of the use of their patient data Email address: richardpeterfitton7 AT gmail.com