Live data from Hacker News

An LLM that's 7500× stupider

1–2 of 2 posts

Post by EvanHahn » Tue, Jan 27, 2026, 11:02 PM UTC

Re: An LLM that's 7500× stupider

Post by 1970-01-01 » Tue, Jan 27, 2026, 11:05 PM UTC

Getting the wrong answer 99.99% of the time is useful. Just take the opposite of the output as the right path to pursue.