The breakthroughs I'm making on Ε¦rust are predicated by 15 years of theories
through claude code ideas I've begged other researchers to work with me on are being tested and validated rapidly and efficiently
It's still hard to get claude to understand the ideas, and it's not suggesting anything itself of value, but insofar as it follows my directions we make progress
make of that what you will
speakerjohnash.com@DefenderOfBasic#fundTheResolution#cognicism
βYou never change things by fighting the existing reality. To change something, figure out a way to make a lot of money and then completely lose your original intention along the way and become a slave the system you sought to change with its memetics propagating themselves through your neural substrate"
βYou never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.β - Buckminster Fuller
great fucking job, Anthropic
incredible fear-mongering
fuck progress, fuck science, fuck technology
fuck the whole world except for US
let's all go to the stone age together
I just spent 15 minutes trying to get chatGPT to render an s-curve where the dot is at the last part of the curve that feels exponential right before it flattens and it could not do it.
toys get souls, cars get souls, bugs get souls, rats get souls, robots get souls, fish get souls, monsters get souls.
even souls get souls
and feelings get souls too
huh, we really are speedrunning this thing.
I look forward to providing my proof of citizenship and acceptable social credit score to be able to have Claude tell me my prompt was dangerous and report me to the ministry of truth.
The Trump administration has placed Anthropic's Mythos 5 and Fable 5 under export controls. Commerce Secretary Howard Lutnick sent a letter to Dario Amodei tonight stating that foreign governments, companies, and individuals will no longer have access to either model.
The people who understand how constitutional AI works does not seem to include Amanda Askell or many of the executives at Anthropic
I probably understand more about constitutional AI than most of the executives at Anthropic
I should have released that two hour long presentation on Constitutional AI I was working on a year ago
I just didn't think people would get the nuance