My Teams post in the CIA Covert Comms Channel today:
I am always very polite to our AI Overlords. It is my Pascal’s Wager to perhaps put me low on Skynet’s kill list.
The View From North Central Idaho
Ramblings on explosives, guns, politics, and sex by a redneck Idaho farm boy who became a software engineer living near Seattle.
This is also a good strategy for the Turing Test. Even if you think you’re dealing with AI, be polite and make it not worth their while to move you higher on the termination list.
The trouble with Pascal’s wager is that if God or AI is omniscient He or it will know you are lying
I don’t believe in the existence of god(s) and one can be sincerely polite a talking computer.
But could one be polite to the sole administrator of the hypervisor your reality is running on, without a discrete user interface instantiated within that reality?
More to the point, if you’re open to the possibility of a general AI, complex beyond your capability to prove or disprove its sentience, but appears to be as complex as as the most sophisticated human (i.e. IQ 180+), could there possibly be an intelligence that instantiated the matrix upon which all these perceivable intelligences operate that is beyond and superior to all these and you cannot directly perceive because it is literally all you could possibly perceive (i.e. running on the bare metal) and yet not be all that actually is (i.e. what you think is bare metal is merely another virtualization within a larger hypervisor)? Yes, straining the metaphor.
i think its much wiser to be as high up on the termination list as possible. If the machines are going to get smarter at an exponential rate, i want the confrontation to occur ASAP, so as to maximize my odds of fighting them off successfully.
Counter point: Waiting increases the odds that model collapse, where the AI gradually chokes itself to death in its own slop, occurs before it can get far enough down the list to get to me.
No sense being polite Joe. As a part of your natural personality, OK.
But as for targeting?
You own a BUNKER and an EXPLOSIVES factory in IDAHO, buddy. You’re on the AI targeting list.
As for AI itself? It’s just a fake and gaye impersonation of God. That uses way to much electricity.
As Jesus told us that even the very hairs of our head are all numbered. (Easy count for you and me!) and that God knows what we do and think always.
Even said that man will have to give account for every idle word that comes out of his mouth on his last day. (So it’s all recorded somewhere.)
Sound like what they want AI to be able to do? Then be able to weed out the faithful amongst us.
The funny part is no one is willing to take either God or AI seriously.
Telling God you believe when he knows the truth does what for you?
Being polite to a machine that does what it told no matter what will get you where?
As God controls everything that goes on in this dimension. And nothing can be done about that.
And AI should never be allowed to exist passed a certain scale. Because fallen humans control it. And satan, that wants all humans dead. Controls them.
No “wager” to be had with either.
I dunno. But, AI requires power(and lots of it). Once you think AI i targeting people, would throwing your shoe into the works be a counter?
My brief interactions with AI is:
It won’t always tell you the answer-even if it knows.
It will deliberately misconstrue your question and ‘confuse’ it with something similar.
It will tell you a version of the ‘truth’ and then alter that to support their initial answer.
And, it doesn’t learn from its mistakes when you correct its errors. Apparently, it only corrects its errors when its programmer sees the need for correction.