what’s worse?

>interacting with someone for a while only to discover they’re a bot 

>writing someone off as a bot only to discover that they’re a human

For me, the former has me more impressed by the technology than there being any feeling of being cheated or deceived in some way

The latter makes me feel bad, but I can’t help but think they’re using an LLM to help craft their text; perhaps they’re ESL and just trying to connect. This is me being *not* impressed by the technology and hopeful that it’ll get better and allow for more human connection in the world

Head of Product at @syky /digital-fashion. Love all dogs and especially my own. ex-Ledger

agreed with both points. the confusion i have is this: if you’re using an LLM, why would the language be so awkward, grammar so poor? is that the idea - to come off as human through error? 

i find that the bots aren’t the ones that talk like typical gpt. they’re usually disguised by ESL

I see. Yeah, maybe two steps back currently

agreed. my main issue is it’s a net negative for the folks that are legitimately ESL

I’m not sure how they arrive at the output that is so clearly bad to us

Like I said, I am just hopeful it will get better