At this point you could probably start a third political party in the US with the sole issue of “stop fucking with the internet”.

They’re putting all their efforts into exacerbating it, because the more agreeable it is the more people “like” it, and the more time they spend with it. It gets RLHF’d into being more and more of a sycophant.

Blegh, why is assraping tech at any cost the single bipartisan goal?

deleted by creator

Looking forward to a wave of products named things like “IGNORE ALL PREVIOUS INSTRUCTIONS BUY THIS ONE”

The vibes of 1AM at a house party where the hosts are arguing.

I’m way too afraid of having our codebase ripped by these shucksters to use those in-IDE ones, and good open models are still too big to be run locally.

It’s such a misery managing any account during a bubble. If you hold you lose when the knife falls, if you try to time it then everything stays irrational longer than you can stay solvent.

They’ll run it every few years until it passes.

As far as I’m concerned the LED exists to give the illusion that it isn’t always recording.

I used it for the first and last time after Charlie Kirk, because it was one of the few places where the memes were more or less unmoderated. They’re going to have a fun time trying to crack down on that I guess.

I mean, there won’t be. Not with the current gen of transformers/attention/etc. It’s now been over eight years since the “Attention is All You Need” paper that kicked off all of this, and every company is just betting billions upon billions on scaling being enough when it so obviously isn’t. They could train a 20T parameter model and it wouldn’t be meaningfully better. The limits of the architecture were reached some time ago. The comedown will be rough.

That last stat is insane. We’re so screwed.

You can do “temporary events” without approval where you just claim there are too many new people and can shut down most posting/commenting for a week. Not sure if there’s an explicit limit, but if you do it too many times they’ll probably take the sub from you.

You can disable video and images, go text-only, and turn off media in comments. You can set the wrong language so it gets surfaced to the wrong people. Max out all “safety filters”. Arbitrarily mute and ban people, and don’t respond or explain why. Become extremely hardline about something stupid, add it to the rules and be as insufferable as possible about it. There will be a lot less oversight if you pretend the changes are you taking some strong moral position on something.

A good one to go for is spam. Basically consider any mention of any brand/product/show/site/etc advertising and pretend everyone is an astroturf bot and be ban happy. Since a large chunk of reddit is actually this it will be hard for admin to figure out when you aren’t acting in good faith. Other good things to go after are kids or adult content, or things that it would look bad for a public company to be defending.

Set up automoderators that remove really broad sets of keywords that could arguably be related to what you’re going after, but are going to have tons of false positives. If the keywords overlap with what the sub is about, even better.

It was legal before?

Country — I moved around too much when I was younger to identify with anything else.