chat

Agreed, didn't expect so many home users drinking the koolaid. Ulysses would have had a field day.

And I suspect many of those users are among the tech-savvy, too, sadly.

This seems like a valid counterargument for people who use local models to generate things for themselves, but it doesn't seem like it tackles the usecases where people use GenAI to fight capital's use of GenAI (e.g. using an LLM to modify one's resume before applying for a job for a company that will, in all likelihood, use an LLM to screen the resume).

I'm not going to dedicate months of my life to figure out how to SEO-optimize my resume. I am perfectly capable at realising when something is a waste of time better spent caring for my basil bushes.
I am not going to spend days carefully crafting each and every job application in order to get a chance at passing through the weird requirements there are around a piece of paper that really should be sufficient with "I am capable of doing the job, here are my credentials. I am pleasant to work with, contact my references. I would like to have the job." but for some reason needs a bull-load of fluff. I will definitely not do that when I know reading those applications have been outsourced to ~~LLMs~~ ~~neural networks~~ ~~programmes~~ AI for more than a decade.

Dismissing every usecase with "git gud lol" doesn't address the actual reasons people use AI