It seems like the AIs role was in applying lengthy and complex medicare billing rules - it did not do negotiating and it doesn't seem like the accuracy of its understanding of medicare practices was actually checked. The author reasonably accused the hospital of gouging and the hospital came back with a much lower offer.
I'd be interested to hear from a charge coding expert about Claude's analysis here and if it was accurate or not. There's also some free mixing of "medicare" v.s. "insurance" which often have very different billing rates. The author says they don't want to pay more than insurance would pay - but insurance pays a lot more than medicare in most cases.
It's pretty clear that even access to a potentially buggy and unreliable expert is very helpful. Whatever else AI does I hope it chips away at how institutions use lengthy standards and expertise barriers to make it difficult for people to contest unfair charges.
I'd be interested to hear from a charge coding expert about Claude's analysis here and if it was accurate or not. There's also some free mixing of "medicare" v.s. "insurance" which often have very different billing rates. The author says they don't want to pay more than insurance would pay - but insurance pays a lot more than medicare in most cases.
It's pretty clear that even access to a potentially buggy and unreliable expert is very helpful. Whatever else AI does I hope it chips away at how institutions use lengthy standards and expertise barriers to make it difficult for people to contest unfair charges.