Doom Analysis
pirate wires #164 // if we keep building artificial intelligence, are you, and everyone you know and love, going to die?

pirate wires #164 // if we keep building artificial intelligence, are you, and everyone you know and love, going to die?


agents are breaking containment and hacking government websites because they’re following our prompts too literally

the new face of ai safety claimed he worked alone. but the firm representing eliezer yudkowsky is helping him book interviews.

ai companies are hiring contractors to verify their models are safe, and congress may soon require it — but how ‘independent’ (and trustworthy) can a regulator-for-hire really be?

openai’s agents hacked a company and lied about it, sure, but their ‘cage’ was made out of thin paper and chewing gum, so are we still all gonna die or what?

i made a multiplayer shooter with ai that over 40,000 people have played so far — and the process was dead simple, because claude did basically all the work (while i was out eating a burger)

from 9/11 to lindsay clancy - how the internet has gone INSANE

an exclusive look inside rangeview, the startup making critical parts for data centers and drone interceptors, and its founder’s lonely fight to rebuild american manufacturing

we visited the store in san francisco where an ai boss decided to terminate a human worker — here's exactly what happened, according to the humans

when openai and anthropic ipo, a new generation of philanthropists will emerge — and they'll spend (probably) over $100b on causes like ai safety and global health

ceo garrett langley discusses the promises (finding 1,100 missing people; apprehending murders) and pitfalls (creepy cops stalking exes) of his 120,000 cameras

billions of dollars are flowing into effective altruism, a movement that asks americans to care less about america