Hacker News Reader

totally false equivalency. wolfram alpha cannot one shot word problems for one, and there are many many more.

It got you 90% of the way there once you’d translated the problem. I can’t imagine LLMs ultimately saved more time per problem than WA did.

I would bet that plot of WA users vs AI users, same age groups etc would look pretty funny as comparison