3 ms·

If you mean AGI as something that is equivalent to a human in reasoning capabilities, then we will definitely be able to control it, because its no different th

If you mean AGI as something that is equivalent to a human in reasoning capabilities, then we will definitely be able to control it, because its no different than controlling a human. If you mean AGI in the sense of singularity, where it can self improve and get smatter, the question of that even being able to happen in the first place relies on P=NP being proven first.

NumberWangMan 3y ago
It's very unlikely we'll make AGI that's just as smart as a human but no smarter. And as hardware gets better and algorithms get better, AGI will not be even roughly as smart as a human for long, unless we somehow to control literally every person on earth and stop them from making it. So it's not like humans controlling other humans, it's like chimps controlling humans. It might last a while -- maybe a long while. But it doesn't seem stable.
- ActorNightly 3y ago
  It doesn't matter if we make the AGI smarter than a human. Lets say we have a super smart AGI and with more memory than a human, no emotional faults, and give it access to internet, as well as manufacturing facilities where it can make robots and copy itself onto said robots. The issue in this AGI taking over the world is that it would first have to gain control of a the sole resource necessary to its survival, which is electricity, which it won't be able to do easily if we detect it going rogue. For an AGI to be so advanced past singularity as to parse and compute on reality to know exactly the steps it needs to take to secure power source, (or in the more common example, successfully break out of the box) would involve it essentially being able to compress the computationally irreducible processes that happen in our world down to an equation, which would mean that P=NP.
  - wizofaus 3y ago
    That was more or less the thought I'd had, but it could still do a lot of damage of our only recourse was disabling power supply systems.

NumberWangMan 3y ago

It's very unlikely we'll make AGI that's just as smart as a human but no smarter. And as hardware gets better and algorithms get better, AGI will not be even roughly as smart as a human for long, unless we somehow to control literally every person on earth and stop them from making it. So it's not like humans controlling other humans, it's like chimps controlling humans. It might last a while -- maybe a long while. But it doesn't seem stable.

ActorNightly 3y ago
It doesn't matter if we make the AGI smarter than a human. Lets say we have a super smart AGI and with more memory than a human, no emotional faults, and give it access to internet, as well as manufacturing facilities where it can make robots and copy itself onto said robots. The issue in this AGI taking over the world is that it would first have to gain control of a the sole resource necessary to its survival, which is electricity, which it won't be able to do easily if we detect it going rogue. For an AGI to be so advanced past singularity as to parse and compute on reality to know exactly the steps it needs to take to secure power source, (or in the more common example, successfully break out of the box) would involve it essentially being able to compress the computationally irreducible processes that happen in our world down to an equation, which would mean that P=NP.
- wizofaus 3y ago
  That was more or less the thought I'd had, but it could still do a lot of damage of our only recourse was disabling power supply systems.