I'm an air traffic controller. There has been a bit of a to-do lately among the political class because some air traffic facilities still have controllers using paper flight progress strips. But then, the political class doesn't really understand what air traffic controllers do.
My primary job is twofold, prevent bad things from happening (they kind of understand this) AND, arguably more importantly, when bad things happen, facilitate a "recovery" such that people do not get hurt (& they also like it when we keep the flights on time ;-).
No matter how good your technology might be, it will occasionally fail. Computer screens (with those lovely digital flight strips) can go blank. Paper strips can get coffee stains (try spilling your coffee on the computer & see how well it works) or they fall on the floor. You pick them up & you continue working the traffic.
Critical systems rely on resilience & recovery. The more complex the systems, the more fragile they become & the more dependent on them we can become. Have any of you seen the movie WALL-E?
1. I never did care for Picasso, but then, de gustibus non est disputandum
2. I think AI will be used extensively in medicine and is positively dangerous to the patient. IN part, this is this fault of what we laughingly call 'intelligence' in AI; in part this is due to the design of a medical system which purpose is to maximize profits, not health. When all other cost-cutting measures have been used, reducing the number of people who must be paid is the American way. AI's promise, from the corporate perspective, is just that: get rid of the expensive doctor in the equation Failing that, limit the doctor as much as possible so that they can see more patients in less time.
Hmm... you pose an either/or question which I find difficult to answer. Do I think that a system will be implemented which makes extensive use of AI? YES. Do I think that it will work? NO.
I see where you're coming from. That's actually a very good point: people vary considerably. Often, we'd look at the activity of a disease, for example if they're inflamed, we might find their C-reactive protein is high, and so on.
I'm an air traffic controller. There has been a bit of a to-do lately among the political class because some air traffic facilities still have controllers using paper flight progress strips. But then, the political class doesn't really understand what air traffic controllers do.
My primary job is twofold, prevent bad things from happening (they kind of understand this) AND, arguably more importantly, when bad things happen, facilitate a "recovery" such that people do not get hurt (& they also like it when we keep the flights on time ;-).
No matter how good your technology might be, it will occasionally fail. Computer screens (with those lovely digital flight strips) can go blank. Paper strips can get coffee stains (try spilling your coffee on the computer & see how well it works) or they fall on the floor. You pick them up & you continue working the traffic.
Critical systems rely on resilience & recovery. The more complex the systems, the more fragile they become & the more dependent on them we can become. Have any of you seen the movie WALL-E?
Re: the two surveys:
1. I never did care for Picasso, but then, de gustibus non est disputandum
2. I think AI will be used extensively in medicine and is positively dangerous to the patient. IN part, this is this fault of what we laughingly call 'intelligence' in AI; in part this is due to the design of a medical system which purpose is to maximize profits, not health. When all other cost-cutting measures have been used, reducing the number of people who must be paid is the American way. AI's promise, from the corporate perspective, is just that: get rid of the expensive doctor in the equation Failing that, limit the doctor as much as possible so that they can see more patients in less time.
Hmm... you pose an either/or question which I find difficult to answer. Do I think that a system will be implemented which makes extensive use of AI? YES. Do I think that it will work? NO.
Then I guess the answer is "it won't play a substantial role in *fixing* the EHR" :)
But it's your vote.
Very interesting - possibly you need TWO new inputs
The "how sure am I"
And "how well did (the fix) work"
Hmm. If the 'fix' didn't work, it's not much of a fix, is it?
But THAT information is useful - with these conditions on that individual it did not work - or only worked partially
I am just an engineer but I think that we will end up finding that some things - some "fixes" work well with some people and not at all with others
I see where you're coming from. That's actually a very good point: people vary considerably. Often, we'd look at the activity of a disease, for example if they're inflamed, we might find their C-reactive protein is high, and so on.