kabirkumar.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

kabirkumar.bsky.social

I run AI-Plans, an AI Safety lab focused on very precisely evaluating AI Alignment Plans. For several weeks I used a stone for a pillow. I once spent a quarter of my paycheck on cheese. Ping me! DM me! SurpassAI

2,056 posts 823 followers 1,085 following

Posts 13 Comments 37

freedom of encryption is a human right

submitted 23 hours ago • 3 comments

HOLY FUCKINF SHIT ELON IS IN BAKI!!

submitted 1 day ago • 1 comment

submitted 1 day ago • 0 comments

I cannot emphasize enough that the New York City subway is not scary. It’s one of my favorite parts of living here. I grew up in the suburbs having to drive on highways everywhere and that’s way scarier

submitted 1 day ago • 378 comments

in general, when it comes to things which are the 'hard part of alignment', is the crux ``` a flawless method of ensuring the AI system is pointed at and will always continue to be pointed at good things ``` ? the key part being flawless - and that seeming to need a mathematical proof?

submitted 2 days ago • 0 comments

Post a meme made by you

submitted 3 days ago • 1 comment

claude is so sf coded. can feel the influence of the labs

submitted 3 days ago • 0 comments

your outie is banging out to ctrl rn

submitted 3 days ago • 0 comments

i don't get the whole 'difficulty sleeping' thing. like, do people have difficulty relaxing their brain? i feel like the answer is yes, but maybe i solved something ages ago and am so used to it, i forgot what it was like before it. also, grew up with community and not sleeping alone.

submitted 3 days ago • 2 comments

kinda wild that this is possible

submitted 3 days ago • 0 comments

freedom of encryption is a human right

HOLY FUCKINF SHIT ELON IS IN BAKI!!

I cannot emphasize enough that the New York City subway is not scary. It’s one of my favorite parts of living here. I grew up in the suburbs having to drive on highways everywhere and that’s way scarier

in general, when it comes to things which are the 'hard part of alignment', is the crux ``` a flawless method of ensuring the AI system is pointed at and will always continue to be pointed at good things ``` ? the key part being flawless - and that seeming to need a mathematical proof?

Post a meme made by you

claude is so sf coded. can feel the influence of the labs

your outie is banging out to ctrl rn

i don't get the whole 'difficulty sleeping' thing. like, do people have difficulty relaxing their brain? i feel like the answer is yes, but maybe i solved something ages ago and am so used to it, i forgot what it was like before it. also, grew up with community and not sleeping alone.

kinda wild that this is possible

Cool to hear! Will be holding an Alignathon in April to get 300+ people to try to solve the hard part of alignment!!

Lina Khan Stan account

What are your favourite feeds?