←back to thread

Auto-grading decade-old Hacker News discussions with hindsight

(karpathy.bearblog.dev)
615 points __rito__ | 1 comments | 10 Dec 25 17:23 UTC | HN request time: 0.218s | source

Related from yesterday: Show HN: Gemini Pro 3 imagines the HN front page 10 years from now - https://news.ycombinator.com/item?id=46205632
Show context
tptacek ◴[10 Dec 25 20:16 UTC] No.46223130[source]▶
'pcwalton, I'm coming for you. You're going down.

Kidding aside, the comments it picks out for us are a little random. For instance, this was an A+ predictive thread (it appears to be rating threads and not individual comments):

https://news.ycombinator.com/item?id=10703512

But there's just 11 comments, only 1 for me, and it's like a 1-sentence comment.

I do love that my unaccredited-access-to-startup-shares take is on that leaderboard, though.

replies(3): >>46227906 #>>46228813 #>>46231858 #
1. kbenson ◴[11 Dec 25 05:12 UTC] No.46227906[source]▶
I noticed from reviewing my own entry (which honestly I'm surprised exists) that the idea of what it thinks constitutes a "prediction" is fairly open to interpretation, or at least that adding some nuance to a small aspect in a thread to someone else prediction counts quite heavily. I don't really view how I've participated here over the years in any way as making predictions. I actually thought I had done a fairly good job at not making predictions, by design.