Hacker News new | ask | show | jobs
by wizzwizz4 246 days ago
It only "passes the bar exam" when AI, or some other flawed process, is the examiner. See e.g. https://doi.org/10.1007/s10506-024-09396-9 for a debunk.
1 comments
hdseggbj 246 days ago
That's not a debunk. "Calls into question" does not equal "in truth, it failed the exam. "
link
hluska 246 days ago
No, it’s a debunk. ChatGPT-4 scored in the 48th percentile (15th percentile in essays) amongst individuals that passed the bar exam. That’s very poor performance.
link
hdseggbj 246 days ago
Thus it scored higher than almost half the humans who passed the test. In other words it too passed the bar.
link