niplav @niplav@schelling.pt

vote / bet on values / beliefs

a bit sad :-/

I now remember why I stopped doing as many forecasts: Everything was so confounded by AGI soon

@Aleums good choice

Interpretability research is a thing that is both useful and academia would be very good at it.

Also you probably don't need moleFLOP computers to do it.

YES the AfD for Harberger tax article didn't go through[1]!

[1]: https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Harberger_Tax

@panchromaticity

@ZachWeinersmith this is…good?

be empty of that too

@TetraspaceGrouping two apparently

fiction story about a species where females disguise as males for procreation to steal metabolically difficult to produce eggs from other females

They're STILL releasing new versions of the AmigaOS‽

i heard that you can summon a particular LessWronger by saying “The Categories Were Made For Man, Not Man For The Categories” three times in front of a mirror

@a_lizard@mastodon.social update: eggs acquired, can now parthogenize

AI arms race is more akin to a winners curse than a prisoner's dilemma

Things I regret having spent as much time on:

• Reading classical philosophy (exceptions exist)
• Learning languages

Things I don't regret:

• Having learned coding
• Having done forecasting
• Reading STEM non-fiction

☺️

Does OpenAI still publish interpretability research?