Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

the only reasons i can imagine that a customer would want to use an AI chatbot for support instead of chatting with a person is either because they don't currently have the option to chat with a person 24/7 at all (AI is better than no chat support), or their experience with human chat support has been terrible (long wait times, slow responses, unhelpful agents, annoying language barriers, responses so unnatural and overly scripted that they might as well be bots, etc).

There's nothing AI brings to the table that a competent human wouldn't, with the added benefit that you don't have to worry about AI making things up or not understanding you.

Or maybe they just want to try and convince the AI to give them things you wouldn't (https://arstechnica.com/tech-policy/2024/02/air-canada-must-...)



I agree, although "competent human" is not really the bar when we're talking about actually existing phone support. A while back I was trying to solve an issue with Verizon and calling three separate times I got three completely different approaches to solving my problem, all three of which were totally incorrect. (And the correct solution was literally just "go to this URL and fill out a form".) At least one of those people gave me advice that would have put me in legal hot water. It was rough.


I dunno. Sarah Connor explained it pretty nicely in the 90s

https://m.youtube.com/watch?v=tksN5Jaan9E

Kinda on the nose




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: