About whether AI can attempt arbitrary problems from a plain-language spec.
Has this happened?
Yes 127 (88%)Not sure 12 (8%)No 5 (3%)
Votes cast 1–2 October 2026: 100,590 votes from 9,694 people.
Hacker News has set AI a lot of challenges over the years. Which ones has it met?
About whether AI can attempt arbitrary problems from a plain-language spec.
Votes cast 1–2 October 2026: 100,590 votes from 9,694 people.
You know, if you're at the point where you can give a human-readable spec of the problem and the AI can make a passable attempt at it, that's basically the Turing Test -- hence why I think it deserves its status as holy grail. Something that passes would really give the impression of "there's a ghost inside here".