Inboth cases, the benefits are wildly exaggerated

“If they want to try something out, they ask us, and we have a decision-makingframework based on our values and our principles,” Jones says.

Our data would be used by the very people that beat that language out of ourmouths to sell it back to us as a service,” Jones says. “It’s just like taking our land and selling it back tous,” Mahelona adds.

that they’re tested on those without thechoice to opt out;

If we filter out the discourse of marginalized populations, we fail to provide training data thatreclaims slurs and otherwise describes marginalized identities in a positive light,

“The training data hasbeen shown to have problematic characteristics resulting in models that encode stereotypicaland derogatory associations along gender, race, ethnicity, and disability status,”

instead of fighting their way through that and growing from it, they retreat to something thatmakes it a lot easier for them.”

But she’d rather get good grades

“I’m against copy-and-pasting. I’m against cheating and plagiarism. All of that. It’s against the student handbook.” Then shedescribed, step-by-step, how on a recent Friday at 8 a.m., she called up an AI platform to help her writea four-to-five-page essay due two hours later.

“It’s short-circuitingthe learning process, and it’s happening fast.

I can write an essay in two hours that normally takes 12.”