23 Feb 2024
I have a long-standing policy of hanging up on robots.
These days, of course, automated calls are 99.9̅% scams. But even before, if it's not important enough to have a human make the call, it's not important enough to have a human answer.
I get a lot of fake-personalised email, the kind that starts "I loved your podcast INSERT NAME HERE, the recent episode really made me think, are you interested in a collaboration", before going on to a generic pitch. It's signed with a human name, but it's very clear that no human even reviewed that message, let alone wrote it.
The modern trend seems to be to have that same robot follow up with increasingly aggravating fake humanity on a fixed schedule: "Re: Collaboration / Hi NAME! I was sad to not hear back from you about PODCAST! Let's circle back, when's a good time to chat?"
I get it. Business is business and outbound sales is a grind. (I used to do it, somewhat under protest - I hand-typed every outgoing message with real thought, meeting with very limited success. Human connection doesn't scale.)
What I object to is the underlying deception. The automated calls I'm getting now don't start with "Tʜɪs ɪs ᴀɴ ɪᴍᴘᴏʀᴛᴀɴᴛ ᴍᴇssᴀɢᴇ ꜰʀᴏᴍ", they start with a human-sounding "Hi how are you [insert three second gap] Good I wanted to talk about..". As LLMs and text-to-speech improves, I'm worried my reaction will shift from "nice try, bot" to "... argh, was I just rude to a person?".
Similarly, I'd respond much more warmly to an email starting with "This is the sales robot at yetanothersaas.example. Based on your metrics we can increase your reach by 27% at a cost of $189/mo. Hit reply to discuss with a human". Why would you want to begin our potential collaboration by lying to me? Yet, everyone does.
The grey area is AI-assisted communication. Imagine a prompt: "Summarise Gwyn's latest email to me". Or "draft a polite reply". Or "send customised birthday wishes to all my facebook friends who are having birthdays this week". For me, the fact that good communication takes time is a way to show value - the fact that it's hard is the point. It's Proof Of Burn. The tools invisibly undermine this, and I'm not sure what will replace it.
Yes, to some extent this is privilege. I don't use grammarly; I am grammarly. But this reflects on values as much as on gifts - I've built these skills through practice, and my writing, my emails, and even my texts are polished because I care about communicating clearly, and I care about the people I'm communicating with.
The insidious nature of AI-assisted communication is that I might not even know, not for sure. But to the extent that I do, then: in a business context, if you didn't type your message, I'm not going to read it. And in a personal context.. please don't.
I want to talk to you, not a machine wearing your skin like a suit.