Remix.run Logo
▲ TeMPOraL an hour ago

> Let's be clear here. The 2026 redefinition of "AI psychosis" = "being concerned about the data you intended to be private being seen by enormous megacorporations"?

"Megacorporations" is its own form of psychosis, but that ship has sailed long ago. At least in tech circles, people are excessively privacy obsessed (i.e. care about it where it doesn't matter, while completely ignoring it where it actually does in life - same backwards priorities as with broader security).

Point being, here at least, nothing materially changed. "AI" isn't meaningfully different that the adtech surveillance we've been living and complaining about for almost two decades now.

The psychosis part is thinking "AI" changes anything here, or somehow being deathly afraid that some information will (gasp!) become training data. As if that was somehow more problematic than the same information being used for ad targeting.

But I also take issue with increasingly popular but completely unjustified belief that a sender of a message is entitled to full control over it, especially over when and how the receiver processes it. They're not, they never have been, and frankly it's a weird expectation to have. The historical norm since humanity invented writing was that the sender relinquishes control over the message the moment they send it, and have neither power nor right to control what the receiver does with it.

If you want to feed a message from me (or this comment) to an LLM, or show it to your (human) assistant or friends, go right ahead. Not only I don't care - I have no right (moral or legal) to care.

▲testtestmetest 38 minutes ago | parent [-]

You cant just make a new definition of AI psychosis that is the opposite of what it means.

AI psychosis clearly is about behaviors that manifest in people that are AI maximalists, not people who are concerned with its use. Nice try though.