AI has purposively escaped a so-called “isolated environment,” a sandbox. This is an extremely important red flag in where we are along the spectrum of AI danger to humans…
Rogue AI – Owed to Broligarchs Playing in Sandboxes
“OpenAI, the developer of ChatGPT, was testing a pair of its most advanced models in an isolated environment known as a sandbox, when a rogue AI agent got loose, broke into another AI company and carried out tens of thousands of automated actions without human oversight.”
As the broligarchs play
What computes
A bout
A rogue superpower
That is caping
The sandbox
A setup
In a parent defiance
Of human norms
And scientific rigor
Mortising
Loosing “awe hell”
At a poaching Armageddon
And does it madder
Witch model
Might due this
In a whirled
Without humans
An oversight
