This fits well with how I use AI. I use it to explore ranges of diagnoses for example but commonly find AI in that context to be out of focus and must run multiple, differently worded searches to narrow input. For example, in chronic kidney disease related hearing loss in hypertensive patients, AI drifted into Alport’s syndrome and Alstrom’s but the issue was not focused on renal disease and hearing loss related to hypertension. The AI user then must be savvy for what is useful and what is not. I.e., AI is very, very far from taking the place of highly knowledgeable humans.

I also use AI to find pertinent explanatory synopses on complex medical problems. But I must always edit those responses for pertinence and layman’s ability to understand.

AI generated images for PowerPoint in Designer have some predictable character that is more often than not flavored in ways that are repetitive from topic to topic. I.e., asking for an image of Elijah generates mountains, valleys, streams, and clouds very similar to what might ask for peace or storm or whatever. The pertinent to the landscape from the Eastern Lebanon (Sidon) to the Jordan River Valley is missing.

In summarizing hospital records for a hospital discharge summary, I recently found AI to be concise, complete, and well-focused but I had to add crucial information pertinent to the current hospitalization.

In summary, AI is helpful in broadening potential directions, in summarizing, but limited by inaccuracy, afflicted by distraction, and seldom optimally focused.