glangmead

@peterb@mathstodon.xyz I mean, one can usually just tell, right? I guess I'm ok with that. Meaning, I probably do that too. The hidden portion of all the arguments I've ever seen is the part where someone assesses the authenticity, good faith, and motivations of the interlocutor.

@peterb@mathstodon.xyz Can you suss that from their words? To clarify my own, the booster has to explicitly say "you are signaling" or "look at this bozo who doesn't even know how to build their own LLM."

@peterb@mathstodon.xyz I may have hijacked your reasoned discussion with my radicalized temperament. Let me add this. I agree with the article. I want everyone to think for themselves and be skeptical of grandiose claims. Power to the people.

@peterb@mathstodon.xyz I should have remembered my rule: no discussion without defining terms. A booster, to me, is someone who dismisses all of the top-5 worries as one of a) inauthentic or b) easily defeated by just knowing more technical stuff. In that sense the word "hater" is coined on their side, to capture the inauthenticity they think they see.

@peterb@mathstodon.xyz I'm with you! The crazies are crazy. All I'm saying is, welcome to team Hater.

@peterb@mathstodon.xyz I think this reasonable and measured approach is what a booster would call a hater. And hater is a booster term.

@peterb@mathstodon.xyz Sorry — what I’m saying is the author is a hater. In the sense of taking it upon themselves to decide what is the case about LLMs and what is not.

@peterb@mathstodon.xyz Those are reasonable stances. But are they compatible with boosterism? They want us to believe these are conscious, i.e. anthropomorphize them. To insist they are not conscious is my position and qualifies me as a hater.