this post was submitted on 06 Dec 2024
275 points (99.3% liked)

United States | News & Politics

1980 readers
2019 users here now

Welcome to [email protected], where you can share and converse about the different things happening all over/about the United States.

If you’re interested in participating, please subscribe.

Rules

Be respectful and civil. No racism/bigotry/hateful speech.

Post anything related to the United States.

founded 2 years ago
MODERATORS
[email protected]
[email protected]
275
UnitedHealthcare CEO murder suspect no longer believed to be in NYC, NYPD says (6abc.com)
submitted 5 days ago by cm0002 to c/[email protected]
83 comments fedilink hide all child comments
 

cross-posted from: https://lemmy.crimedad.work/post/159369

A law enforcement source told our sister station, ABC7 Eyewitness News in New York, that the suspect's unmasked smile came as he flirted with the front desk clerk who checked him into the hostel, and encouraged him to drop the mask so she could see his smile. The suspect obliged, pulling his mask down long enough for the surveillance camera to capture his face.

you are viewing a single comment's thread
view the rest of the comments
[–] NotMyOldRedditName 57 points 5 days ago* (last edited 5 days ago) (1 children)

I don't know if the number was real but someone posted a picture that their denial rate was 32% whereas the lowest provider was 9%

[–] [email protected] 79 points 5 days ago (1 children)

Their denial rate was 8% when Brian Thompson became CEO and jumped to 22.7% after one year of his leadership. 32% is believable after 3 years, especially if it's true that the decision to train an LLM for the purpose of auto rejecting claims was his directive

[–] [email protected] 30 points 5 days ago (1 children)

Why a LLM when an autoreply of "Denied" would work?

[–] [email protected] 18 points 4 days ago

Presumably because the LLM can give a justification as well, rather than just a straight "no". Doesn't mean it's a good justification but I'm sure they have metrics that say when it looks somewhat plausible that more people don't challenge it.