3 ms·

This is also why AI won't overthrow us and become our masters. They don't even know what it means to be a master. EDIT: This is the most volatile comment I've

This is also why AI won't overthrow us and become our masters. They don't even know what it means to be a master. EDIT: This is the most volatile comment I've ever posted. It has been going +2, -2, +2, -2 for the last 35 minutes. People seem to love it or hate.

tim333 10y ago
Not yet.
- inputcoffee 10y ago
  As I understood this article, they're not even trying to. Their version of understanding isn't of the right sort.
  - tim333 10y ago
    I haven't heard of an AI system understanding that yet but it's a fairly common concept in everyday life for instance dogs tend to have masters. Anyone trying to make an AI understand everyday life or literature is going to have to deal with that.
    - inputcoffee 10y ago
      I think the article speaks to this very point. The people "trying to make an AI understand" are not making it "understand" in the sense that you and I use most commonly use the term.
      - outsideline 10y ago
        You're exactly right. If you look closely, present day A.I is modeled after the cortex. It's akin to cutting out the neo-cortex, wiring in i/o hooks, and reprogramming it to one's needs (weak A.I). The cortical regions are good at creating hierarchical feature maps and we had a bunch of search algorithms laying around in the parts bin. Presto : Present day A.I. This approach meshes well with 'big-data' companies in possession of large compute stacks and data-sets. So, its the direction things went. A perfect opportunity for disruption.. The current wave is on borrowed time.
philh 10y ago
Nor does a meteor know what it means to be an extinction event.
- inputcoffee 10y ago
  True but a meteor doesn't have to. As it has been written of in our sci-fi, the robot uprising specifically requires the machines to understand and, most of all, care, about dominance. If you're talking about just being victims of machine logic, we've been suffering that since the invention of the traffic light traffic jam.
  - TeMPOraL 10y ago
    That's because fiction is fiction, especially one created for the masses must be written in terms masses can relate to. As for reality, creating an AI that will actively hate us is a feat of about the same difficulty as creating an AI that would love us. Those are two opposite points of a tiny island called "has more-less human mind" that floats on a vast ocean named "we all die". The biggest challenge of surviving superhuman AIs is locating that island. As the words of wisdom say, "The AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else."
  - jupiter90000 10y ago
    It may not need to understand or care. It may just be optimizing towards a goal, which involves 'humans need to die' or 'destroy/pollute the water supply', etc. It may indirectly cause serious damage without even awareness we exist.
  - mattw1810 10y ago
    https://wiki.lesswrong.com/wiki/Paperclip_maximizer https://wiki.lesswrong.com/wiki/Paperclip_maximizer

tim333 10y ago

Not yet.

inputcoffee 10y ago
As I understood this article, they're not even trying to. Their version of understanding isn't of the right sort.
- tim333 10y ago
  I haven't heard of an AI system understanding that yet but it's a fairly common concept in everyday life for instance dogs tend to have masters. Anyone trying to make an AI understand everyday life or literature is going to have to deal with that.
  - inputcoffee 10y ago
    I think the article speaks to this very point. The people "trying to make an AI understand" are not making it "understand" in the sense that you and I use most commonly use the term.
    - outsideline 10y ago
      You're exactly right. If you look closely, present day A.I is modeled after the cortex. It's akin to cutting out the neo-cortex, wiring in i/o hooks, and reprogramming it to one's needs (weak A.I). The cortical regions are good at creating hierarchical feature maps and we had a bunch of search algorithms laying around in the parts bin. Presto : Present day A.I. This approach meshes well with 'big-data' companies in possession of large compute stacks and data-sets. So, its the direction things went. A perfect opportunity for disruption.. The current wave is on borrowed time.

philh 10y ago

Nor does a meteor know what it means to be an extinction event.

inputcoffee 10y ago
True but a meteor doesn't have to. As it has been written of in our sci-fi, the robot uprising specifically requires the machines to understand and, most of all, care, about dominance. If you're talking about just being victims of machine logic, we've been suffering that since the invention of the traffic light traffic jam.
- TeMPOraL 10y ago
  That's because fiction is fiction, especially one created for the masses must be written in terms masses can relate to. As for reality, creating an AI that will actively hate us is a feat of about the same difficulty as creating an AI that would love us. Those are two opposite points of a tiny island called "has more-less human mind" that floats on a vast ocean named "we all die". The biggest challenge of surviving superhuman AIs is locating that island. As the words of wisdom say, "The AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else."
- jupiter90000 10y ago
  It may not need to understand or care. It may just be optimizing towards a goal, which involves 'humans need to die' or 'destroy/pollute the water supply', etc. It may indirectly cause serious damage without even awareness we exist.
- mattw1810 10y ago
  https://wiki.lesswrong.com/wiki/Paperclip_maximizer https://wiki.lesswrong.com/wiki/Paperclip_maximizer