Comic Discussion > QUESTIONABLE CONTENT
WCDT 22-26 August 2011 (1996-2000)
Is it cold in here?:
If I understood pwhodges right, genetics comes into it because he was talking about the prospects for artificial life as opposed to artificial intelligence. There is a point of view which holds that you have to have an unintelligent base for an intelligence to work from, and that it is senses rather than logic that are the basis for cognition. This is diametrically opposed to the belief in AI research that AI is an easier problem than artificial life. Nobody can prove either one at our current state of knowledge.
Then there's the matter of our choice about what we will call "alive". If I understand pwhodges right, he would not call Charlotte a faithful emulation of a living being, while Charlotte's opinion is a matter of record. For my part I would say "It's life, Jim, but not as we know it".
Then there's a distinction to be made about the limits of algorithmic processing. We can be absolutely confident about some limits of algorithms, but solving the halting problem is not a precondition for intelligence. We have no idea whether Turing equivalence is adequate for matching human capabilities. The Church Hypothesis posits that it is adequte, but computability theory does not help us decide whether it's true.
EDIT: Additional AnthroPC discussion
pwhodges:
If the ability to make some decisions is all that's required for AI, we're not so far off; if the ability to function as recognisable robots in society is the aim, then the QC robots appear to have reached it; but if you want the decisions to be backed up by the whole depth of human experience, then I think we're hardly started.
Can you picture your robot sales assistant going to the park for a lunch break, visiting the art gallery after work, climbing a tree at the weekend, spending six months doing VSO in Africa? All things which richen the human experience, and feed into the ability to make decisions in matters that are not rather trivial. Or if not those activities, some other mind (AI) broadening activity with equivalent effect. Maybe R Daneel Olivaw was up to these things, but I don't imagine Charlotte is.
gangler:
Whoa, I haven't done any of that either. Be right back. I have some business to attend to.
Boradis:
--- Quote from: Carl-E on 27 Aug 2011, 07:41 ---
A friend of mine in gradd school was in Nuclear Engineering. He was constantly upset that the only work that recieved grants was for power generation and weapons, when all he wanted to work on was a propulsion device.
--- End quote ---
But what kind of nuclear-powered propulsion device can be built that wouldn't irradiate the launch site and cause massive fallout? Reactors work great on sea-going vessels where they generate electricity to spin the propeller, but that won't move you through a vacuum.
Skewbrow:
Is it cold in here? gave a nice summary IMHO. I apologize for any possible confusion created by pointing at provably undecidable questions (such as the halting problem). That is my knee-jerk reaction to any claim/expectation that computers/AI will soon be solve any problem.
Re: Nuclear propulsion. Any idea what the mechanism would be? If a nuclear reaction could eject used fuel (or ballast) at relativistic speeds that would make a fine rocket, but is that at all feasible? @Boradis: I think that most if not all commercial nuclear reactors generate electricity by first turning water into steam that then turns turbines. I may be wrong about this as I only read about this somewhere last spring during the Fukushima incident. I don't know how the reactors at nuclear powered submarines do it.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version