Same even on Android. If a site doesn’t work well with FF, Vivaldi usually works. Vivaldi is also better for PWAs.

Hey man the French revolution involved actual beheadings and it brought much of the prosperity you have today. It’s not this scum ruining your life. It’s untouchable billionaires far away from consequences.

She’s a bystander. This is an insured storefront of a multinational.

CEOs singlehandedly ruining whit people’s PR. Sounds just as dumb.

It happens in every worthwhile protest. If you see it more it’s partly because the media wants you to, partly because things are so bad your country has been letting more people feel this is okay.

Also, everyone thinks the French revolution was so cool, but beheadings weren’t normal back then either, it probably wasn’t the most well-spoken of the protestors doing that violence either.

And abstinence is the least effective birth control, but proper education saves lives. Wished this article took the same approach!

Such a weird and morally loaded article.

It seems to conflate a bunch of choking techniques, only mentioned once with jargon, but keeps talking about breath play only?

Conclusion calls it “not normal” but half of people do it?

Outcomes like Russian roulette? One in six die? Ok?

I’m guessing the history popup still works? Ctrl shift h

The wild thing is, NFTs were creating value before they were called that, and they still are. A git commit hash is an NFT, for all intents and purposes. The only uhh innovation is being able to pay money for it.

“I like it in here.”

Pornhub does this.

Tumblr doesn’t do it, but in stead incentives spreading with the queue feature. People do it out of etiquette/ to not get unfollowed.

Pornhub does this

These days they sand it, and then you’re breathing wood and wax varnish dust.

Oh wow!

Only the cone-magnet-coil-wires part is the same. All the electrical parts are different. There’s no way.

Inflation will take care of that. Memory never really dropped in dollar price.

Lower temperature is not necessarily better quality. At 0 it will get very repetitive, it’s for classification tasks, not for prose.

I suppose the optimal temperature depends on the model and the task, and I don’t think it’s very sensitive. Varying temp might even give better results, who knows.

By being just a little weird, but in a pattern, and not to you. A pattern like: every n tokens raise the temperature (randomness) for one token. Then to dectect AI, you tokenize and calculate how expected each next token is. Then you try fitting the pattern to that.

Need to get stickers of this to put on all these awful posters

Is it webscrapers?