3 ms·

There's nothing sacred about machine learning, machine vision & feedback loops, assembled into robots. This is what we are.

scotty79 13y ago
Sacred, no, but still worth recognition and respect proportional to complexity.
stiff 13y ago
Machine learning might be a useful model of what happens in the brain but just like the map isn't the territory this doesn't make us robots, what would that even mean? Everything aside, biological systems are completely different than everything we have ever built and so far it seems current machine learning techniques not necessarily have to have much to do with what happens in the brain.
michaelgrafl 13y ago
No, that's a crass abstraction of what we are. To a philistine the Mona Lisa is nothing but a bunch of pigments applied to a surface. That's what makes him a philistine, after all.
- kryten 13y ago
  It's an apt description of what we are. Only we think we are more. Perhaps we're not and the thought is just part of that feedback loop. Art, wine, music, maths. It's all just more inputs...
  - michaelgrafl 13y ago
    It's a marginal description of what we are. It takes a true nerd living in a nerd bubble to interpret a human being as a data processing device.
    - kryten 13y ago
      I think you mean your perception is that my description is marginal, which is my point. We're just maths if you go far enough down. So are machines.
RivieraKid 13y ago
No, machines cannot feel pain.
- leoedin 13y ago
  Pain is a feedback sensor which indicates that something's gone wrong. You could build a machine with pressure sensitive grippers. You could even place a threshold on that, above which the machine reacted by moving in the opposite direction. Biological pain is just a complicated version of that. Pain is just our minds interpretation of damage signals. At what point do you draw the line. "This machine doesn't feel pain, it just responds to damage signals." could apply to a simple robot or a cockroach. At some point, biological computers become complicated enough that we deem pain to have arrived, but that point is fairly arbitrary. Why aren't we just biological machines?
  - RivieraKid 13y ago
    There's a huge difference between feeling pain and reacting to demage. Humans can react to injury in a very complex ways without feeling pain. And buildings too - when you crash with an airplane into a building, the building reacts in a complex way. Does it feel pain? Feeling pain is something fundamentally different than reacting to demage.
    - leoedin 13y ago
      If feel only fits your narrow, human centric definition of it, then of course machines can't feel pain. There isn't much point in having a discussion if the discussion is about your understanding of a word though. If you go by the broader definition of feel, which includes "to have a sensation of something", then I'd argue that the line is far more blurred. If you accept that a rat is capable of feeling pain, then surely you must accept that a functionally identical machine with a neural network functionally the same as a rat's brain can also feel pain? There's no real fundamental physical reason stopping us constructing a computer which functions in the same way as a rat's brain does. Does the computer we constructed suddenly stop being a machine?

scotty79 13y ago

Sacred, no, but still worth recognition and respect proportional to complexity.

stiff 13y ago

Machine learning might be a useful model of what happens in the brain but just like the map isn't the territory this doesn't make us robots, what would that even mean? Everything aside, biological systems are completely different than everything we have ever built and so far it seems current machine learning techniques not necessarily have to have much to do with what happens in the brain.

michaelgrafl 13y ago

No, that's a crass abstraction of what we are. To a philistine the Mona Lisa is nothing but a bunch of pigments applied to a surface. That's what makes him a philistine, after all.

kryten 13y ago
It's an apt description of what we are. Only we think we are more. Perhaps we're not and the thought is just part of that feedback loop. Art, wine, music, maths. It's all just more inputs...
- michaelgrafl 13y ago
  It's a marginal description of what we are. It takes a true nerd living in a nerd bubble to interpret a human being as a data processing device.
  - kryten 13y ago
    I think you mean your perception is that my description is marginal, which is my point. We're just maths if you go far enough down. So are machines.

RivieraKid 13y ago

No, machines cannot feel pain.

leoedin 13y ago
Pain is a feedback sensor which indicates that something's gone wrong. You could build a machine with pressure sensitive grippers. You could even place a threshold on that, above which the machine reacted by moving in the opposite direction. Biological pain is just a complicated version of that. Pain is just our minds interpretation of damage signals. At what point do you draw the line. "This machine doesn't feel pain, it just responds to damage signals." could apply to a simple robot or a cockroach. At some point, biological computers become complicated enough that we deem pain to have arrived, but that point is fairly arbitrary. Why aren't we just biological machines?
- RivieraKid 13y ago
  There's a huge difference between feeling pain and reacting to demage. Humans can react to injury in a very complex ways without feeling pain. And buildings too - when you crash with an airplane into a building, the building reacts in a complex way. Does it feel pain? Feeling pain is something fundamentally different than reacting to demage.
  - leoedin 13y ago
    If feel only fits your narrow, human centric definition of it, then of course machines can't feel pain. There isn't much point in having a discussion if the discussion is about your understanding of a word though. If you go by the broader definition of feel, which includes "to have a sensation of something", then I'd argue that the line is far more blurred. If you accept that a rat is capable of feeling pain, then surely you must accept that a functionally identical machine with a neural network functionally the same as a rat's brain can also feel pain? There's no real fundamental physical reason stopping us constructing a computer which functions in the same way as a rat's brain does. Does the computer we constructed suddenly stop being a machine?