alt Hacker NewsThere was a post today or yesterday where Paul Graham and Nate Silver accused the AP of being woke for saying you shouldn’t anthropomorphize LLMs. But I can imagine there to be quite a few unpredictable consequences to encouraging your customers to project emotions onto your product that humans generally project onto each other.
I’m not sure SV is prepared for what happens when you go down that road.
> There was a post today or yesterday where Paul Graham and Nate Silver accused the AP of being woke for saying you shouldn’t anthropomorphize LLMs.
I searched, but couldn't find this. Do you have a link? Sounds like a totally unhinged response from totally unhinged people. There's nothing "woke" about wanting my word calculator to avoid pretending it's a person.
When I integrate LLMs into my apps that respond to user queries, I deliberately have my system instructions prohibit all self-referential usage of first-person pronouns. The system can only refer to itself as "this system", never "I" or "me".
Any chat bot that pretends to be a person just grosses me out, especially in a customer experience context. Gives off very manipulative and patronizing vibes.