On Thursday, October 24, 2013 11:09:40 PM UTC-4, chris peck wrote: > > *>> The alien might be completely confident in his judgement, having a > brain made of exotic matter. He would argue that however complex its > behaviour, a being made of ordinary matter that evolved naturally > could not possibly have an understanding of what it is doing.* > > Aliens don't matter. They can be wrong about us being thoughtless and we > can be right that computers are thoughtless. > > There seem to be two points of view here: > > 1) Whether a machine is thinking is determined by the goals it achieves > (beating people at chess, translating bulgarian) > > 2) Whether a machine is thinking is determined by how it trys to achieve a > goal. How does it cognate? >
My view is 3) Whether a machine is thinking is determined by the extent to which it understands and cares about the content of its thought. As long as we assume that who and the why of consciousness can be reduced to the what and how of logic, we have no chance of understanding it. We cannot learn about what makes the Taj Mahal special by studying masonry. > > I find myself rooting for the second point of view. A machine wouldn't > need to beat kasperov to convince me it was thinking, but it would have to > make mistakes and successes in the same way that I would against kasperov. > > In developmental psychology there is the question of how children learn > grammar. I forget the details; but some bunch of geeks at a brainy > university had developed a neural net system that given enough input and > training began to apply grammatical rules correctly. What was really > interesting though was that despite arriving at a similar competence to a > young child, the journey there was very different. The system outperformed > children (on average) and crucially didn't make the same kind of mistakes > that are ubiquitous as children learn grammar. The ubiquity is important > because it shows that in children the same inherent system is at play; the > absence of mistakes between computer and child is important because it > shows that theses systems are different. > > At this juncture then it becomes moot whether the computer is learning or > thinking about grammar. It is a matter of philosophical taste. It certainly > isn't learning or thinking as we learnt or thought when learning grammar. > The way we cognate is the only example we have of cognition that we know is > genuine. Do AI systems do that? The answer is obviously : No they don't. > Are computers brainy in the way we are? No they are not. You can broaden > the definition of thought and braininess to encompass it if you like, but > that is just philosophical bias. They do not do what we do. > I agree, but to me the interesting part is *why* AI systems are different than we are. It's not so much about passing a test by sprinkling human-like errors into a computer to rough it up around the edges, it's about seeing that the entire cosmos is fundamentally based on absolute improbability and that logical truth is actually derived from that. From the local perspective, absolute improbability looks like error or probabilistic coincidence, but that is because our expectation is cognitive rather than emotional or intuitive, and therefore it is specialized for virtual isolation and alienation from the Absolute. Thanks, Craig > Regards > > > From: stat...@gmail.com <javascript:> > > Date: Fri, 25 Oct 2013 13:11:47 +1100 > > Subject: Re: Douglas Hofstadter Article > > To: everyth...@googlegroups.com <javascript:> > > > > On 25 October 2013 12:31, Craig Weinberg <whats...@gmail.com<javascript:>> > wrote: > > > > >> You could say that human chess players just take in visual data, > > >> process it in a series of biological relays, then send electrical > > >> signals to muscles that move the pieces around. This is what an alien > > >> scientist would observe. That's not thinking! That's not > > >> understanding! > > > > > > > > > Right, but since we understand that such an alien observation would be > in > > > error, we must give our own experience the benefit of the doubt. > > > > The alien might be completely confident in his judgement, having a > > brain made of exotic matter. He would argue that however complex its > > behaviour, a being made of ordinary matter that evolved naturally > > could not possibly have an understanding of what it is doing. > > > > > The > > > computer does not deserve any such benefit of the doubt, since there > is no > > > question that it has been assembled intentionally from controllable > parts. > > > When we see a ventriloquist with a dummy, we do not entertain > seriously that > > > we could be mistaken about which one is really the ventriloquist, or > whether > > > they are equivalent to each other. > > > > But if the dummy is autonomous and apparently just as smart as the > > ventriloquist many of us would reconsider. > > > > > Looking at natural presences, like atoms or galaxies, the scope of > their > > > persistence is well beyond any human relation so they do deserve the > benefit > > > of the doubt. We have no reason to believe that they were assembled by > > > anything other than themselves. The fact that we are made of atoms and > atoms > > > are made from stars is another point in their favor, whereas no living > > > organism that we have encountered is made of inorganic atoms, or of > pure > > > mathematics, or can survive by consuming only inorganic atoms or > > > mathematics. > > > > There is no logical reason why something that is inorganic or did not > > arise spontaneously or eats inoragnic matter cannot be conscious. It's > > just something you have made up. > > > > > > -- > > Stathis Papaioannou > > > > -- > > You received this message because you are subscribed to the Google > Groups "Everything List" group. > > To unsubscribe from this group and stop receiving emails from it, send > an email to everything-li...@googlegroups.com <javascript:>. > > To post to this group, send email to > > everyth...@googlegroups.com<javascript:> > . > > Visit this group at http://groups.google.com/group/everything-list. > > For more options, visit https://groups.google.com/groups/opt_out. > -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To post to this group, send email to everything-list@googlegroups.com. Visit this group at http://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/groups/opt_out.
