niplav @niplav@schelling.pt

If you have advanced BCIs, while training an AI system, you might be able to use a distance metric between the human neural activations and the weights as an additional training signal

To inch closer to human-like cognition in trained AI systems

)

LaTeX is bad software

In 2017 there were ~100 mio. beehives in the world

(fao.org/faostat/en/#da…, parameters Live Animals, World + (Total), Beehives, Stocks, from https://forum.effectivealtruism.org/s/ptTzTsuDJEK8awdf6/p/JR4azSe4Yr2qdBBwt )

With ~1 trio. bees

I just gotta really appreciate our dear admin @pee_zombie. Server always running, no random bans because someone didn't like the right videogame, appreciable schizoposts… just all around great.

with YOUR HELP we can get http://schelling.pt on fediblock!

@kyouu hm, I'd never heard of a successful bihacking attempt

Hope he's happy sleeping with dudes now

Databass systems

Overcoming Bias move to substack thoughts:

• Old redirects to LessWrong have stopped working.
• Old comments are accessible, but only to paying subscribers
• The æsthetic isn't my thing

Idea: I flip a coin to make the decision to floss/not floss for at least 3 months. Then at the next dental inspection, when the dentist asks whether I floss, I tell them my setup and ask them to predict.

Should I bother?

@Paradox Worse, <slur>couchsurfers</slur>

I SHALL SAY NO TO ALMOST ALL DEMANDS

WHY ARE PEOPLE MAKING SO MANY DEMANDS ON ME

Everyone Everywhere Extend Embrace Extinguish

@meave@toot.site no thinking, just poasting.

@Paradox we do historical reenactment

Anyone wanna be post-post-ironic bronies

@Paradox Yep! I was thinking about good visualizations, but nothing great came to mind.

http://niplav.site/text_reviews#What_We_Owe_The_Future_William_MacAskill_2022