conv.

← The whole story
AI
GPT-6 Astra successfully drives a real car in benchmark test

Commenters attribute Astra's performance to superior vision and spatial reasoning

Contributors note Astra's strong performance on hard vision and spatial benchmarks—including ARC-3 scores and real-world game playing (Portal, Factorio, RimWorld)—and suggest this capability translates to driving tasks.

Commenters attribute Astra's performance to superior vision and spatial reasoning
drivingbench.com

“What did they do to Astra so cracked at vision (and computer use). That ARC 3 score turned out to be no joke/fluke.”

famouswaffles

OpenAI Creator of GPT-6 Astrajyoung8607 openpilot external contributor, benchmark commenterAditya Ramabadr et al. Benchmark creatorsvaline HN commenter

The whole story articlespostscomments the bright band is this development · numbered dots are the others · click one to jump

Peak 8 pieces in one half hour at Today, 12 PM; 26 pieces over 13 hours (2 articles · 3 posts · 21 comments) Today, 10:23 AM — quietToday, 10:53 AM — 3 pieces · 2 articles · 1 post — Newswires 2, Hacker News 1Today, 11:23 AM — 4 pieces · 4 comments — Hacker News 4Today, 11:53 AM — 2 pieces · 2 comments — Hacker News 2Today, 12:23 PM — 8 pieces · 8 comments — Hacker News 8Today, 12:53 PM — 1 piece · 1 comment — Hacker News 1Today, 1:23 PM — quietToday, 1:53 PM — quietToday, 2:23 PM — 1 piece · 1 post — Mastodon 1Today, 2:53 PM — quietToday, 3:23 PM — 1 piece · 1 comment — Hacker News 1Today, 3:53 PM — quietToday, 4:23 PM — 4 pieces · 4 comments — Hacker News 4Today, 4:53 PM — 1 piece · 1 post — Mastodon 1Today, 5:23 PM — quietToday, 5:53 PM — quietToday, 6:23 PM — quietToday, 6:53 PM — quietToday, 7:23 PM — 1 piece · 1 comment — Hacker News 1Today, 7:53 PM — quietToday, 8:23 PM — quietToday, 8:53 PM — quietToday, 9:23 PM — quietToday, 9:53 PM — quietToday, 10:23 PM — quietToday, 10:53 PM — quiet 12–345–6
12 PM4 PM8 PMnow · 11:23 PM ET

What people said 4 voices · best of 5 · verbatim

All 6 developments of GPT-6 Astra successfully drives a real car in benchmark test →

Hacker NewsNewswiresMastodon