Generative AI is Boring
Jere Odell
I didn’t want to start this way. But it’s where I am. Sorry.
Two and half years after ChatGPT launched and I am just now beginning to sort out what LLMs are for. From my reluctant and limited recent use, it appears that Gen AI in writing and image generation is for the bored. The product it produces is not meant to be read. The images are cringeworthy.
The tasks it completes are those that the user was too bored to tackle. Sure, maybe it gets some people off to a start, but even that start is a dull one.
I have been told that I can ask Gen AI tools to jazz it up. Uh huh.
Gen AI reminds me of web searching. (It is web searching of a kind.) Does anyone that is not living in 1993 still find web searching fun?
Yes, I have also used chatbots to help me code. This too was boring – and more parts maddening. The troubleshooting worked, to a point. But only after I bounced and read the documentation myself. I guess if I already knew the ins and outs of calling up an API and retrieving results with Python, this boring task would have been a lot faster.
In the context of work, I could see how AI might enable me to do more boring things in less time.
Yay.
In the context of living, I would rather live.
Give me a year and maybe I’ll change my mind.