I love how the article just glosses over the fact that this woman reported doctors were treating her poorly because her notes mentioned she’d consumed psychoactive substances. Like apparently that’s just an acceptable and a valid way to treat patients.

I know healthcare workers get compassion fatigue but it’s just fucking depressing how normalised shitty behaviour is from them towards the vulnerable people rekying on them.

If it’s not AI’s fault, then the error was the human who implemented the system. This is basic reasoning FFS.

Don’t let people blame machines unless the machines can be held responsible.

This type of hallucination is a problem, but is not my biggest concern.

I just want to converse with a human who is listening to me. It’s a bonus if that person has some familiarity with me but it’s not essential.

I certainly do not want an LLM inferring what might be wrong with me.

I feel like this shouldn’t be too much to ask, but it’s becoming that way.

Dumbest part about this is that auditory dictation software has existed for DECADES and WORKS. It’s even based on machiblne learning models, just ones that have been trained for that specific purpose, not this LLM human-impersonating trash.

“Please be assured that we take the accuracy of our documentation very seriously, and we are truly sorry that this occurred,” the letter stated.

Yeah. That’s why you do AI: because you take your patients’ documentation really seriously…