←back to thread

LLM Inevitabilism

(tomrenner.com)
1616 points SwoopsFromAbove | 1 comments | 15 Jul 25 04:35 UTC | HN request time: 0s | source
Show context
cdrini ◴[15 Jul 25 07:21 UTC] No.44568772[source]▶
How do you differentiate between an effective debater using inevitabilism as a technique to win a debate, and an effective thinker making a convincing argument that something is likely to be inevitable?

How do you differentiate between an effective debater "controlling the framing of a conversation" and an effective thinker providing a new perspective on a shared experience?

How do you differentiate between a good argument and a good idea?

I don't think you can really?

You could say intent plays a part -- that someone with an intent to manipulate can use debating tools as tricks. But still, even if someone with bad intentions makes a good argument, isn't it still a good argument?

replies(4): >>44568810 #>>44568918 #>>44569240 #>>44571980 #
throwawayoldie ◴[15 Jul 25 15:11 UTC] No.44571980[source]▶
One good way is if they back up their assertions and explain their reasoning, which generally people arguing for LLM inevitability don't--they make assertions like "LLMs are only going to get better".

Contrast this with someone arguing the other side of the question like Ed Zitron or David Gerard. You may or may not agree with their arguments, but they explain, in great detail, with numbers and citations, how they arrived at their conclusions. If you don't agree with those conclusions, you should be able to point at a certain part of their argument and say "this is where you went wrong."

replies(1): >>44573659 #
1. cdrini ◴[15 Jul 25 17:27 UTC] No.44573659{3}[source]▶
+1, I think the mark of a good argument is to back up assertions and explain reasoning. Although I disagree, I'm seeing well-reasoned arguments in this thread both for and against the claim "AI is inevitable".