[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 28 KB, 1200x1200, ea-logo-square.png [View same] [iqdb] [saucenao] [google]
14958814 No.14958814 [Reply] [Original]

>> No.14958841

>>14958814
No, that's Christianity because it's beyond human's reason.

>> No.14960110
File: 36 KB, 652x280, whatifalthist effective altruism.png [View same] [iqdb] [saucenao] [google]
14960110

>>14958841
t. Rudyard Lynch

>> No.14960121
File: 2.29 MB, 2834x5102, simmap.jpg [View same] [iqdb] [saucenao] [google]
14960121

>>14958814
If this is a simulation, most of the suffering that supposedly exists in the world might not actually exist. All the people in ward, plagues, famines, etc. might all just be NPCs. I might be the only conscious being that actually exists.

>> No.14960136
File: 197 KB, 800x1200, AI safety vs other cause areas.jpg [View same] [iqdb] [saucenao] [google]
14960136

Scientifically, the only cause area that matters in the long run is preventing AI from subjecting us to eternal torture.

https://en.wikipedia.org/wiki/Suffering_risks
https://www.lesswrong.com/posts/N4AvpwNs7mZdQESzG/the-dilemma-of-worse-than-death-scenarios

>> No.14960167

>>14958814
No. The best use of resources is whoever finds the most value from said resources and is willing to give the best offer for it. What is altruistic? who gets to define what is worthy of research and what isn't? Why can't I just say "fuck off" and use my resources to research how to turn myself into a car so I can finally live out my dream of having someone inside me?

>> No.14960245

no

>> No.14960298

>>14958814
Let's face it; humans are awful in doing effective altruistic deeds.

If you're American, you probably remember how people donated to BLM in 2020.

Did they think about what they were going to use the money for?

How would additional money help their goal of stopping police brutality against Blacks?

Aren't there more leading causes of death than police brutality?

It's all stupid.

>> No.14960326

>>14958814
What the fuck even is effective altruism?

>> No.14960929

>>14960136
My understanding is that the mainline probability is AGI is much more likely to just kill everyone than create suffering.
It's not a convergent instrumental goal.
It's not impossible, but I wouldn't assign it a higher probability than getting corrigible alignment right.
90-95% of the time everyone just dies.
Until EY actually donates MIRI money to oxygenate shrimp out of higher EV, I don't think we're totally fucked.