It would be really nice if there were a super-detailed mode that did not use quotes so heavily. Ideally, the level of detail refers more the the inclusivity (i.e. all issues, concerns, responses to questions, etc. are incorporated) as opposed to the style of writing. Simply quoting the patient verbatim makes from somewhat awkward notes and it is more "literal" or "concrete" than it is "detailed" as minor concerns.

Heidi selectively excluding information becomes problematic as it is hard to remember /notice afterward when reviewing Heidi's note what is as opposed to what is not missing. The only way to do so is to review the transcript - which is extremely cumbersome (particularly given the way it is formatted) - or to take notes myself during the encounter which defeats much of the purpose of using the AI scribe in the first place.

Often I will ask a questions specifically like "what is the most intense physical activity you do on a regular basis" which the patient will answer - but because the patient never complained about inactivity or exertional intolerance Heidi considers it irrelevant.

Ideally, in a super detailed mode the patient's answers to every question I ask would be incorporated somewhere (even if the response was negative) because the fact that I asked the question indicates that the answer was a piece of information I considered relevant to the current presentation.