Hacker News Reader

The LLM critic in me can't help but notice that ChatGPT-3.5 completely whiffed its explanation of integration in the original post, but of course did so in a fluent, confident way that made its response sound like it was correct.

The ease with which LLMs launder misinformation at scale is genuinely dismaying.

View on HN

I'm fairly sure its explanation of integration is correct? Maybe not a good answer for a math major, but it's a good answer for his question.

"when you integrate a function you find the accumulation of its rate of change" is simply wrong. You find the accumulation of the function itself, not its deltas.

GPT is confusing the fundamental theorem of calculus with the meaning of integration.

ETA: what makes LLMs so worrying is that I know calculus quite well, and I think GPT-4 is dumber than a mouse, yet I still felt gaslit by GPT-3.5's wrong answer! I had to make a mental picture to make sure I wasn't contradicting the Mighty Talking Robot. This technology is poison.

We have always had a cognitive bias towards believing glib explanations. We just hold AIs to a higher standard than we do actual human beings. The bias is particularly biased towards attractive individuals we want agree with (so they like us more).

Reply on news.ycombinator.com