i suspect that to fully understand the universe, we need AGI in a similar way to needing a telescope to understand galaxies our unaided faculties are not good enough, and we need better tools--or in this case, smarter tools--to help us make sense of reality
94
79
18
824
also, i think AGI is probably necessary for humanity to survive--our problems seem too big to solve for us to solve without better tools :( it's always hard to think about what tools will enable, but i cannot begin to imagine what we'll be able to achieve with the help of AGI

Jul 27, 2022 · 6:12 PM UTC

36
31
12
369
Load newest
Replying to @sama
An AGI has to maximize impact
1
Replying to @sama
Here's a clue - we're not going to solve our problems. Everything #ElonMusk fears is will happen in some form, the majority of humanity will be wiped out except for the few that get out and all of this AI and robotics BS is for the benefit of the human elites that are left. Any ?
2
Replying to @sama
Is this an unfalsifiable hypothesis?
2
GIF
Replying to @sama
Keep going!
2
Replying to @sama
Perhaps social media platforms - like this - with their users are already AGI. Supraindividual cyborg with silicon white matter and human grey matter.
1
2
Replying to @sama
Yes, but let's make sure we solve AI alignment otherwise it might be the cause of our doom instead of allowing us to realize our cosmic potential.
3
Replying to @sama
we already have better tools, sam — the question is if we're brave enough to use them (☭)
2
Replying to @sama
I apsolutely agree with you dear Sam Altman @sama #SamAltman. Thereupon I am asking you to work hard so we can have AGI asap. #ElonMusk #NASA #JoeRogan 😇🙏
1
Replying to @sama
I apsolutely agree with you dear Sam Altman @sama #SamAltman. Thereupon I am asking you to work hard so we can have AGI asap. #ElonMusk #NASA #JoeRogan 😇🙏
1
Replying to @sama
Part of what AGI will learn will be useful in both advertising and propaganda, related fields in which deception (including exaggeration and oversimplification) has been rationalised and normalised.
Load more