@theorangetheme @alice

'and then?'

@theorangetheme @alice

Philosophy!

@Numerfolt @alice yeah, we need to switch to offensive mode.

That makes me want to create a nightshade fuse FS.
So when you want to upload the image from your picture folder, it nightshades it on the fly.

@miclgael
Wherever possible, I'm using my duck addresses, because all too often they send a confirmation link they need you to click.

Some places have the domain blacklisted, but not all of them.

@alice

@miclgael

Some shops umm...decline...take-away orders without a name. The POS* computer insists on one. I give them a completely random word or number. Works fine.

@alice

*(POS also stands for, ¨Point Of Sale¨)

@vantiss I have to disagree.

Say I go to Amazon. I use perfect tracking protection and I'm not signed in. I browse for a while, and click every ad they serve me. I've wasted a bunch of different companies marketing money, my click data is worthless, and they don't know what ads to send me. I look like every other AdNauseam user, and they still don't know who I am.

Now say I do the exact same thing, but I sign into Amazon. The exact same thing happens, but they know who I am.

...

And as far as LLMs go, waste their fucking money and resources. Use every free option you can, and take it as an opportunity to poison their feedback. Don't give them any personal info, don't use them for critical questions, just flood them with garbage that pops this bubble even faster.

Even if you don't want LLMs in everything, companies will put them there—unless it burns their wallets. The more we set fire to their AIs, the faster executives will learn it's a bad idea to use them.

@mikro2nd @alice

Bend, fold, mutilate, and spindle!

@agturcz @alice

Please enlighten me... What does that do?

@penguinrebellion that's why I said plausible, but fake.

Generators are good though.

There are, however, reasons to enter something wildly off every so often, like "[email protected]", because it tells companies that field is obviously fake. This both makes the plausible fakes more likely to slip by if they do use your data, but also makes them more likely to discard your data for marketing and analytics purposes in general.