[ericmcer]

This could be a really cool use for AI. We have so many rules and systems in place to protect normal people from professionals that we have to blindly trust. If I had a reliable AI in my pocket I could be in charge of my own safety again.

I did this a month or so ago when a Dr. recommended a cortisone injection behind my ankle. I asked GPT about it, it said consensus recommends against it because it can weaken the Achilles. The doctor hated it but I am really glad I didn’t just blindly trust him.

I guess I like it because it is enabling instead of replacing humans.

[romwell]

> did this a month or so ago when a Dr. recommended a cortisone injection behind my ankle. I asked GPT about it, it said consensus recommends against it because it can weaken the Achilles. The doctor hated it but I am really glad I didn’t just blindly trust him.

Taking medical advice from ChatGPT over what your doctor says, what could go wrong?

[DowagerDave]

>> If I had a reliable AI in my pocket

that's both a big "If" and a very ambiguous ask - what's does a "reliable AI" mean?

[intelVISA]

A cursory glance suggests that a reliable AI is one that generates max revenue for shareholders whilst data mining the user reliably.

[biofox]

Someday, but I think we're still a way off.

Regarding your interaction with the doctor, I completely understand both sides, and I can guarantee he was well aware of the small risk of tendon weakening.

People usually want their physician to do "something", otherwise what's the point of going? This puts them in a no-win scenario. A single cortisone injection provides immediate relief, with a very small chance of side effects. If he doesn't offer it, people will complain about him being useless. If he does, some patients will think he's incompetent.

I'd find it exhausting having to second guess every interaction.

[rscho]

Every single doc knows about steroids + tendons. Did you ask why steroids were offered ?