Goalposts

Hacker News has set AI a lot of challenges over the years. Which ones has it met?

2023 OctoberSakos

About GPT-4V reading an image whose "attack while its tail is up" advice is wrong.

I'll be impressed when ChatGPT recognizes that the "Attack while its tail up" message is erroneous and the opposite is true. GPT and everything else like GPT is useless until it can differentiate fact and fiction and understand the difference.

ChatGPT recognizes that an image's "Attack while its tail up" message is erroneous and that the opposite is true.

Has this happened?

Yes 25 (23%)Not sure 78 (72%)No 6 (6%)

Votes cast 1–2 October 2026: 100,590 votes from 9,694 people.