@Edweirdian@akoustov Yes very reasonable. There are strong reasons to be wary of LLM and downstream technology, I just don't think those reasons include hallucination and in fact the less they hallucinate (and more they improve in general) perhaps the more wary we should be.
@Edweirdian@akoustov I just think 'hallucination' specifically as an inherent flaw of these systems is less and less of a concern going forward. There are plenty of other reasons one might or perhaps should be wary of their growing involvement in our lives.
@Edweirdian@akoustov I don't have a specific thought on whether an LLM-based system would or should make contribution to your work or field, I think that depends on the field (it seems to be capable of making independent contributions to frontier math, for example).
@Edweirdian@akoustov If a human made a mistake (with the same mind they use to produce 'correct' work) I don't think you'd view them as inherently flawed, you might even look to see how they improved moving forward.
@Edweirdian@akoustov Consequences for allowing anything through with fabricated citations are welcome, but also you should update yourself on the capabilities of current workflows.
@northwoods1980 What is your suggestion for the ED? Is standard of care meaningfully different wrt to workup of closed head injury with poor history/exam in the elderly if they have baseline cognitive impairment?
@aditya_gan3500@DocPriyamMD What do you think the original premise is? I do not think the point being made was that hypertension, acute or chronic, is benign. I interpreted it as hypertension is more likely to be chronic than hypotension and therefore less dangerous across all comers with hypertension.
@aditya_gan3500@DocPriyamMD Just asking whether hypotension or hypertension in general is more dangerous is so underspecified that it’s not a useful statement I agree.
But that’s clearly not the point being made here, which is that chronic hypertension is common.
@aditya_gan3500@DocPriyamMD Those are both specific cases of high and low BP. If someone had triage vitals in the ED and was 60 points above or below 120 systolic I know which I’d be more worried about, which I think is the point being made.
@intensivaev@ThinkingCC Really nice waveform any advice for someone who can get other Doppler waveforms routinely but struggles with an adequate RVOT view let alone pulse wave
@RogerSeheult@operationdanish Regardless it’s an increase in coding for visits that are similar pre and post AI which is the findings that were discussed. Whether you think those patients were under coded before or over coded after.
@shahrukh_bakar@Hragy What else is on the differential for patient with these stress test findings that would make Cath unnecessary or negative? I’m EM so somewhat ignorant. LVH as another reply said?
@RobertHermanMD@Uqayyum123@Bappieidk On the other hand, when the subjective nature of presentation is often focused on for some cases (typical/atypical, etc.) but then seemingly disregarded for others (no cath for “NSTEMI” w persistent sx) that’s also a bummer