Content
@
https://warpcast.com/~/channel/doctorsphere
0 reply
0 recast
0 reaction
Omar
@dromar.eth
Went to a doctors gathering last night. Only one person out of around 30 I asked knew the difference between a completion model vs reasoning model. I'm not sure if it is expected of them, but found it stunning.
2 replies
1 recast
4 reactions
keccers
@keccers.eth
TBF to them that feels like a fairly deep cut!
1 reply
0 recast
1 reaction
Omar
@dromar.eth
Yeah its a notch too far down the domain to expect them to know it. I was interested because using GPT-4 vs the o1/3 versions makes such a big difference in the results you see. GPT-4 still hallucinates medical info quite a bit, whereas o1/3 can be handle more diverse pt data without having blind spots.
0 reply
0 recast
1 reaction
