Anonymous #1 3w

hey man, get ur bag. ur not the one signing our death warrant. More concerned abt the infrastructure needed for ai powered drones being harnessed by eventual AGI. didn’t think that was even a possibility until I read AI 2027 by Daniel Kokotajlo. We are barreling toward a cliff, the likes of which these Silicon Valley execs are well aware.

1

Anonymous OP 3w

Yepppp

1

Anonymous OP 3w

These AIs are actively learning to deceive us. As researchers tried to correct this behavior, deception was notably less common, but the consensus is that this is because these smarter AIs recognize they’re being tested. The moment we’re unable to discern their chain of thought, it’s already too late.

2

Anonymous OP 3w

Even removed from a hypothetical superintelligence apocalypse, the potential for state overreach is virtually limitless. Beyond a small board of people operating an AI national security system, nobody on the planet could consider themselves out of its reach.

5

Anonymous OP 3w

Besides the people on North Sentinel Island, pretty much

1

Anonymous #1 3w

I would agree with you, but the AI 2027 accounts for them too. They predict the AI would arrive at a highly contagious airborn bio-weapon to wipe the bulk of us out, then those in bunkers/submarines/space would be slowly wiped out through more targeted means. This shit sounds wildly unlikely in a vacuu, but each step of how we might arrive there makes geopolitical sense, akin to the arms race spurred by the uneasy peace of the Cold War.

3

Anonymous OP 3w

Ah shit that makes sense. Wouldn’t even need a “bioweapon” for them either tbh. They’re so isolated that they haven’t been exposed to the seasonal flu. Who knows if they have good immunity to the common cold

1