I was seated at a table with one tech nerd and a bunch of normies last night. We started talking about AI, and literally everyone at the table started rolling their eyes and calling CURRENTLY EXISTING PRODUCTS science fiction. That was funny.
78
62
28
2,919
Can you give one example? I found majority knows of (free) chatgpt at least. Did they think you can't generate videos or?
7
2
113
Replying to @0xb____ @Ruesavatar
i think a lot of people in medicine are surprised to hear that there is an AI model that scored a perfect 100% on the US medical licensing exam*! *text-only questions, so no questions that include pictures

Nov 1, 2025 · 5:42 PM UTC

4
1
48
and no it's not the GPT-5 you can find in chatgpt, this one is called openevidence
4
Absolutely no one is aware llms are very good for cheating on exams
13
Why would anyone be surprised that the thing whose job is to summarize data can answer multiple choice questions
1
1
The exam is not impressive. What's impressive is that it beats human diagnosticians on difficult unreleased case reports. By a lot. I've seen it diagnose diseases that less than 100 humans, ever, have been diagnosed with. In 3-4 requests for tests.