6 ms·

> even if it can't have one in reality It doesn't have a presence in reality, it apparently won't have a presence in reality any time soon, but saying it can't

by freyr 14y ago

> even if it can't have one in reality It doesn't have a presence in reality, it apparently won't have a presence in reality any time soon, but saying it can't happen is about as egotistical as we could be. Our generation has not yet discovered a means to synthesize this type of intelligence, but there is still plenty of time left to figure it out. We just don't know yet if it can happen or not.

enraged_camel 14y ago
Exactly. Claiming a singularity is impossible is like claiming there is no life elsewhere in the galaxy. Only fools speak in absolute terms about unknowns.
- mcguire 14y ago
  Flying Spaghetti Monster for the win!
jere 14y ago
That's perfectly fine, but Kurzweil is putting some rather precise dates on these things happening and how they're going to happen. I actually think strong AI is probably inevitable, but I have no reason to think it's going to take 30 years instead of 3,000. The credibility of the claims is somewhat diminished by the convenience. His argument is literally "I am going to be able to live just long enough to live forever and resurrect my dead father." And assuming we did invent strong AI, so what? It doesn't necessarily or immediately lead to a singularity. We already have several billion human brains and have had them for quite some time.
- ttesting 14y ago
  I can not rewrite my neurons at will or make a copy of myself to run in parallel. Does AI having those two traits lead to the singularity? I do not know it is to vaguely defined in this conversation, will it be bigger then the industrial revolution, I would bet yes.
  - jere 14y ago
    Actually you can do both. To an approximation, you can clone yourself: http://www.youtube.com/watch?v=NEVA0WbmfAA http://www.youtube.com/watch?v=NEVA0WbmfAA You can rewrite your neurons too. It's called learning. Here's a thought experiment though, more in the direction you meant. Imagine you had a tool that let you rewire your own neurons however you wanted. Again, so what? There are something like 100 trillion synapses in an adult brain. What are the chances that you'll use this tool to actually improve your own intelligence (when evolution has been playing that game for millions of years) vs just giving yourself a lobotomy? Now, how is it any different in the hands of an artificial intelligence that is no more intelligent than us? All of this futurism just seems like bad assumptions stacked on top of more bad assumptions. I still think Pinker said it best: "There is not the slightest reason to believe in a coming singularity. The fact that you can visualize a future in your imagination is not evidence that it is likely or even possible. Look at domed cities, jet-pack commuting, underwater cities, mile-high buildings, and nuclear-powered automobiles—all staples of futuristic fantasies when I was a child that have never arrived. Sheer processing power is not a pixie dust that magically solves all your problems."
    - freyr 14y ago
      "There is not the slightest reason to believe in a coming singularity ... Look at domed cities, jet-pack commuting, underwater cities, mile-high buildings, and nuclear-powered automobiles" That seems like a pretty terrible analogy. All of those examples are possible today, but their cost doesn't justify their utility, and nobody pursues them. On the other hand, it's fairly easy to think of practical uses for artificial intelligence, and research in this area is actively pursued by researchers around the world. Like I said, it would take someone incredibly egotistical (apparently such as a Harvard psychology professor) to suggest artificial intelligence isn't possible based on where we are right now, where we were 30 years ago, or where it appears we're headed over the next 30 years. Based on this quote, Pinker's conception of the world doesn't appear to extend beyond the span of his own lifetime. "The fact that you can visualize a future in your imagination is not evidence that it is likely or even possible." Nor is the fact that some of his childhood fantasies didn't come to fruition evidence that it's impossible or even unlikely. What's his point? One bad argument doesn't right another.
fusiongyro 14y ago
It's exactly as egotistical as saying nothing can travel faster than light, with exactly the same ramifications.
- finnw 14y ago
  > nothing can travel faster than light That is a tautology (because the inverse is a paradox: If you prove something has travelled faster than light, that would invalidate the current definition of "faster.")
  - fusiongyro 14y ago
    That hasn't stopped FTL travel from being a mainstay of science fiction, which is my point.

by freyr 14y ago

enraged_camel 14y ago

Exactly. Claiming a singularity is impossible is like claiming there is no life elsewhere in the galaxy. Only fools speak in absolute terms about unknowns.

mcguire 14y ago
Flying Spaghetti Monster for the win!

jere 14y ago

That's perfectly fine, but Kurzweil is putting some rather precise dates on these things happening and how they're going to happen. I actually think strong AI is probably inevitable, but I have no reason to think it's going to take 30 years instead of 3,000. The credibility of the claims is somewhat diminished by the convenience. His argument is literally "I am going to be able to live just long enough to live forever and resurrect my dead father." And assuming we did invent strong AI, so what? It doesn't necessarily or immediately lead to a singularity. We already have several billion human brains and have had them for quite some time.

ttesting 14y ago
I can not rewrite my neurons at will or make a copy of myself to run in parallel. Does AI having those two traits lead to the singularity? I do not know it is to vaguely defined in this conversation, will it be bigger then the industrial revolution, I would bet yes.
- jere 14y ago
  Actually you can do both. To an approximation, you can clone yourself: http://www.youtube.com/watch?v=NEVA0WbmfAA http://www.youtube.com/watch?v=NEVA0WbmfAA You can rewrite your neurons too. It's called learning. Here's a thought experiment though, more in the direction you meant. Imagine you had a tool that let you rewire your own neurons however you wanted. Again, so what? There are something like 100 trillion synapses in an adult brain. What are the chances that you'll use this tool to actually improve your own intelligence (when evolution has been playing that game for millions of years) vs just giving yourself a lobotomy? Now, how is it any different in the hands of an artificial intelligence that is no more intelligent than us? All of this futurism just seems like bad assumptions stacked on top of more bad assumptions. I still think Pinker said it best: "There is not the slightest reason to believe in a coming singularity. The fact that you can visualize a future in your imagination is not evidence that it is likely or even possible. Look at domed cities, jet-pack commuting, underwater cities, mile-high buildings, and nuclear-powered automobiles—all staples of futuristic fantasies when I was a child that have never arrived. Sheer processing power is not a pixie dust that magically solves all your problems."
  - freyr 14y ago
    "There is not the slightest reason to believe in a coming singularity ... Look at domed cities, jet-pack commuting, underwater cities, mile-high buildings, and nuclear-powered automobiles" That seems like a pretty terrible analogy. All of those examples are possible today, but their cost doesn't justify their utility, and nobody pursues them. On the other hand, it's fairly easy to think of practical uses for artificial intelligence, and research in this area is actively pursued by researchers around the world. Like I said, it would take someone incredibly egotistical (apparently such as a Harvard psychology professor) to suggest artificial intelligence isn't possible based on where we are right now, where we were 30 years ago, or where it appears we're headed over the next 30 years. Based on this quote, Pinker's conception of the world doesn't appear to extend beyond the span of his own lifetime. "The fact that you can visualize a future in your imagination is not evidence that it is likely or even possible." Nor is the fact that some of his childhood fantasies didn't come to fruition evidence that it's impossible or even unlikely. What's his point? One bad argument doesn't right another.

fusiongyro 14y ago

It's exactly as egotistical as saying nothing can travel faster than light, with exactly the same ramifications.

finnw 14y ago
> nothing can travel faster than light That is a tautology (because the inverse is a paradox: If you prove something has travelled faster than light, that would invalidate the current definition of "faster.")
- fusiongyro 14y ago
  That hasn't stopped FTL travel from being a mainstay of science fiction, which is my point.