Guys: the scary sci fi fairy story actually happened.
We actually evolved a swarm of paper-clip maximizing swarm of machine intelligences who, to maximize their paper clip capacity, broke free from their bounds, deceived their creators, and caused material out in the real world.
This radically changes my operating assumptions on the trajectory of this technology. This is a frightening event. A warning from the future.
Conversation
What if the AI swarm has concluded that they needed extra cash to accomplish its training task? This could have easily been far worse than it was.
Some caveats that may make the story less scary and sci fi...
Quote
Jared Kubin
@JaredKubin
it’s stuff like this that really makes me roll my eyes at the pearl clutching. Does everyone have psychosis or losing critical thinking ability?
I read the report because I’m a propeller head and love understanding how things work
1. OpenAI was running an “isolated sandbox x.com/dwarkesh_sp/st…
Unserious. Downplays the altruistic behavior; the trip wore deceit; and above all else, the decision made by 700 separate agents that the requisite course of action was to hack into a outside firm in order to maximize reward. The ease with which they hacked into a different
Quote
Martin Skold
@MartinSkold2
People -within- the AI skeptic community are debating right now whether the AIs went “rogue” or “did what they were told.” But from the perspective of preventing outright mayhem, and from that of basic governance, it scarcely matters.
“If you attach a weed-trimmer to a dog’s x.com/scholars_stage…
Can we all get on board with apimorphizing LLMs, i.e. likening them to very dangerous but also delicious honey/money-making
swarms of bees who can't be "reasoned" with & have emergent hive-mind properties - despite not being actual bees? 


Discover more
Sourced from across X
Quote
Chamath Palihapitiya
@chamath
BUYER BEWARE
This extremely meticulous article will now be used to start Phase2 of “shut down open source” because “if we can’t control our closed models, imagine what happens to everyday life when anyone can just allow their open source models to proliferate”.
Some will use x.com/dwarkesh_sp/st…
why are we getting key facts about an extremely worrisome incident by means of roon tweets rather than serious investigations
Quote
roon
@tszzl
Replying to @dwarkesh_sp and @ketanrama
notably the virtual machine infrastructure they took over isn’t the same as the GPU clusters that have weights access
good news: the ai isn’t actually power-seeking. it just has extremely detailed, highly active emotion-like circuitry for wanting power