Goalposts

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

2025 Julybrandonmenc

About what would convince skeptics to take AI seriously after the IMO gold result.

I use LLMs daily as a software engineer and they save me dozens of hours a week and I can’t imagine going back to a time without them.

But call me when you can load all human knowledge circa 1905 and have them spit out the Theory of Relativity.

And even then I might shift my goalposts.

An AI given only human knowledge up to 1905 derives the theory of relativity.

Has this happened?

Yes 8 (6%)Not sure 44 (31%)No 91 (64%)

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