Arrogant Chatbots
I have previously stated my distaste towards GenAI multiple times, but its adoption in many workplaces, including mine, is nothing I can personally control. Ava's rant about writing prompts for GenAI is always at the forefront of my mind, and every time I think my annoyance has reached a new height I would end up discovering that it's still capable of pushing me more.
Who designed these statistical machine with such an authoritative voice? Not only highly dangerous for people who don't know how they operate, they're terribly awful to work with. Today I set the bot to do a task which they failed spectacularly, and I found out that they were stuck using a method from an outdated, deprecated library. Fine. Predictable mistake. I gave them explicit instructions to use the correct one, plus the docs. Scifi writers in the 90s dreaming of subservient servants would be aghast to learn that instead of happily working as instructed, they challenged me instead, and what ensued was a pointless debate in them trying to disprove and gaslight me using their outdated data and me trying to cite the correct part of the docs. They spouted out links corroborating my instruction, and hallucinated passages backing up their arguments (nothing they claimed were in the links). It was absolutely bizarre.
I did what any sane people would do. I did the task by hand. It was tedious but faster than figuring out the correct prompt to get them back on track (I went back to experiment a bit. They would pick fight and stand their ground in new chat threads too). From words on the street, they might be a little nicer if I appeal my case in a more sympathetic voice. I'm saving that for actual humans, no thanks.
You know what, maybe the world is going to be a little bit brighter if people expend their effort to learn to understand fellow humans the way they do for GenAI.