Alright, so mandatory response to the AI counter-alarmism:
“Runaway AI,” another way to consider the singularity concept, is a sci-fi prediction that computer programs we call artificial intelligence can learn how to make themselves smarter and then smarter still until they achieve godlike competency. As previously stated, this is not a scientific concept and there is zero evidence that it could ever actually occur.
Empirical evidence, maybe, but then again we don’t have empirical evidence the sun will rise tomorrow. I probably wouldn’t call it science, but it’s a theoretical idea that’s not unscientific. It’s up in the air if it will actually work like that; it might not, but it could.
its ability to actually engineer the extinction of humanity is totally vague and underdescribed by the doomsayers.
This one’s a bit more interesting. I actually agree, a wifi connection isn’t as powerful as all the alignment people seem to assume. A couple decades of cybercriminals have forced us to wall off the wide open internet from important things already, and once you have to move through meatspace things get much slower and much more prone to the unexpected. Think about it, how would you go about getting a single piece of hardware built on the internet, without using money? And it would need to set up probably thousands of devices without being detected, in order to become self-sustaining.
They do have good ideas about how an AI could get from a powerful warlord to the only consciousness in the solar system, or from a humming box to a server, but not about that gap between a server and a powerful warlord.
As for the new wave of luddites, this guy has good ideas about how to generally achieve the goals of such a movement, if not very original ones, but I’m still not seeing everyone blaming technology for their woes. I might be totally wrong, but technology seems like something people kind of like today, and he doesn’t spend so much time on why that’s the chosen target instead of minorities or the elite like usual.
Alright, so mandatory response to the AI counter-alarmism:
Empirical evidence, maybe, but then again we don’t have empirical evidence the sun will rise tomorrow. I probably wouldn’t call it science, but it’s a theoretical idea that’s not unscientific. It’s up in the air if it will actually work like that; it might not, but it could.
This one’s a bit more interesting. I actually agree, a wifi connection isn’t as powerful as all the alignment people seem to assume. A couple decades of cybercriminals have forced us to wall off the wide open internet from important things already, and once you have to move through meatspace things get much slower and much more prone to the unexpected. Think about it, how would you go about getting a single piece of hardware built on the internet, without using money? And it would need to set up probably thousands of devices without being detected, in order to become self-sustaining.
They do have good ideas about how an AI could get from a powerful warlord to the only consciousness in the solar system, or from a humming box to a server, but not about that gap between a server and a powerful warlord.
As for the new wave of luddites, this guy has good ideas about how to generally achieve the goals of such a movement, if not very original ones, but I’m still not seeing everyone blaming technology for their woes. I might be totally wrong, but technology seems like something people kind of like today, and he doesn’t spend so much time on why that’s the chosen target instead of minorities or the elite like usual.