A lot of people at my work, especially managerment, are very pro AI. I’ve haven’t openly shared my opinion of AI/the fact that I don’t use it, because the hype around AI seems almost cult like at my work. It was months before anyone brought up hallunications.
Part of me wants to share my reasons against AI at work. Some possible reasons I’m thinking of sharing are cooking the planet, you don’t know when it is hallucinating so how do you trust it, critical thinking rot.
Any advice on discussing the negatives of AI at work? Or should I just keep my head down and let sloppers slop?
Ed Zitron’s newsletter has some incredibly useful resources to reference. (I would link it, but I am sleepy at the moment, I apologise).
My opinion is that unless you feel like it’s putting your career at risk, it’s worthwhile to discuss the negatives of AI. When people who aren’t very techy are surrounded by people bigging up AI and there’s no-one dissenting, it’s quite common that they will assume that this is just something that’s too complex for them to understand, so some people are sort of conditioned to believe that AI must be magic, even if their own personal experience or intuition tells them that it’s bullshit.
I also believe that anti-AI takes will age like wine; even though there are some legitimately awesome uses of this tech, overwhelmingly, the majority of the shit that’s being pushed is dumb as hell. AI can be a useful tool in certain cases, but not in the way it’s being rolled out. Instead of empowering humans, it’s just making more slop and hallucinations to wade through. I’m reminded now of a good post from Cory Doctorow that discusses AI and “reverse centaurs”. I can’t explain it better than him, but you could probably find it easy by searching, if you’re interested.
Cory Doctorow describes a “reverse centaur” as a worker that’s tasked with babysitting (in this case) the output of an LLM to make sure that it isn’t wrong, as it often is. A plain old “centaur” would be a worker that uses technologies that make the work that they do easier and faster, like a centaur who uses their horse legs to run faster than a human. The “reverse centaur” would be a human with the head of a horse, where the human body would listen to the head of the horse (or LLM). That situation doesn’t benefit the human nor the horse part.
Exactly, thank you for adding this.
The fact that AI is making reverse centaurs out of us really sticks out to me whenever I watch Star Trek. The way they interact with the computer in Trek makes me wistful for what AI could be, if it were being developed in a more human centred way.