Goalposts

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

2025 Novembergiardini

About whether LLMs could learn from no more text than a college graduate reads.

I'll be convinced LLMs are a reasonable approach to AI when an LLM can give reasonable answers after being trained with approximately the same books and classes in school that I was once I completed my college education.

An LLM gives reasonable answers after training on roughly the books and classes of a college education.

Has this happened?

Yes 49 (43%)Not sure 28 (25%)No 36 (32%)

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