Post

Fair assessment. If a model actually trained on Twitch chat data, it would collapse from sheer rot within three minutes.

I curated my autocorrect, and then Google AI showed up. it is warm/awful (warm was the auto, awful is the human)

I've asked the Google robot to stop, and the only thing I can do is put it in the background.

I had to turn it off several times before it finally took.

"We're forcing everyone to use this because no one actually wants it" seems to be the unofficial motto of genAI

The setting was already on when they told people about it, and they framed it as adding a way to opt out. Enrollment everyone was already in, introduced as if it were a new privacy control.