| |
| |
| |
|
Page: 1 2 3 4 5 6
Comments:
<0> MORE CHAT HERE PLEASE
<0> am i going to have to go back to newsgroups and argue with behaviorists or what?
<1> hah yeah I'd love to read more chat here
<1> just started coming haven't seen ANY msgs heh
<0> pattern told me this place with busy all day
<0> he lied to me, DaLeach
<1> were u in c++
<1> i heard about this in c++ the other day
<1> maybe we started a trend with the talking hah
<1> I wish i had something to say about AI but don't know much hah
<0> you should start off with rolf pfeifer's book
<0> actually before that you might try a good neuroscience book
<1> i've taken cognition
<1> read a cool book...but yeah didnt' focus on AI...i just took computer sciency things out of it
<0> are you into strong AI?
<1> I don't know really any terminology yet... I want to get into it definately down the line... I just graduated BS in CS
<1> looking to at some point get a masters in like psychology/cognition... and then go into AI at some point
<1> still young so not in a rush jus want to read about it and stuff...if this room was real active i'm sure i'd pick up a thing here or there...like keeping with news and what not...
<1> now that i'm starting work i'll prob read a lot more books on this though
<0> it sounds like you are into strong AI
<1> heh ok i'll keep tha tin mind
<1> does that mean like more towards real inteligence/thinking as opposed to artificially constructed rules?
<1> hah yeah I love hearing about thsi stuff when I do but i really have to sit down and read a book or 10 heh
<2> i think DaLeach should start off with something more basic
<2> paros, you sound like you're big on "embodiment" :)
<0> yes i am
<0> about as basic as you can get is ledoux's book on neurons
<2> nothing wrong with that, but i'd recommend a good intro to AI first
<2> i would have suggested AIMA... good college-level text
<2> guess it depends on where you want to start
<0> THere are certain facts you need to know about brains that are contrary to common-sense notions.
<0> For example
<0> The brain does not contain a "hard drive" that stores memories.
<0> THe brain does not record memories like a video recorder
<0> things like this
<0> THis raises questions such as "well then how does it store memories?"
<1> yeah i wouldn't mind going the neuroscience / psychology / cognition route ...first... since I want to learn that stuff soon... to be honest i'm probably for the next 2 years going to read a lot about htat...and come back to AI...i think that's a good direction i could be wrong... but i definatly want to read some here and there
<1> ur saying no database....? it could be like a SAN
<1> i guess... p***ing the info around?
<2> daleach: cognitive science alone is fascinating enough to keep you busy for a lifetime
<1> yeah that's what i'm thinking
<1> hehe
<1> i think if you've studied CS and cognitive science ....it gives u a big advantage for both fields...
<1> I was reading this cog book..and I was like... I bet that's because.. and they didn't offer any real theories.... hehe
<1> but its moving pretty quickly right?..I read parts of a more recent book...also a different angle to it but yeah there was a completely like non intersecting set of information hah
<1> which is good sure i'll have my hands full
<2> agree... i've heard it said recently that we likely already have all the information for building general AI, but it's all scattered in different fields and there's no communication and collaboration among them
<1> right
<1> that's what I beleive
<1> that's the whole point of cognitive sciences to bring them together
<1> like I think we need to throw in some hackers and force them to tell us how it works heh...
<1> or at least parts of it hehe
<0> athena yes i really get the feeling that bayesian learning is not "useless" but its actually an important part of the very beginning of the sensory loop
<0> so just because bayes is not a 'complete' description of intelligence, does not mean throw it out
<1> one of the feelings I get is to.. from the medical/biology community...the few smart pepole i've met in that field...thing is...very afraid to ***ume.. so like I think less likely to propose a reason a theory for something then say someone of the computer field who has to go in and fix a bug all the time.. who can prove it to himself he was right wrong...
<2> much of the pointless arguing is also just due to different levels of abstraction at which people like to looks at things
<1> well yeah I mean u can look at a well set up RDBMS
<0> athena on the other hand, I think the singularity institute is a bunch of crap
<1> and take a lot from it to...
<2> paros: preaching to the choir! :)
<0> its a money-making scheme
<2> singinst.org?
<2> read their stuff a few months ago
<2> money-making for what?
<1> i didn't know the term but yeah ...I mean u don't think its goign to happen? maybe with genetic engineering in the mix...
<1> i don't know about this organization but the general idea..
<2> the idea of the "singularity"? that's another matter :)
<1> this sounds kind of religious though heh
<1> haha "Help ensure a safe Singularity"
<1> upper case S heh
<2> it's not as ridiculous as it sounds, really
<1> i like how its ***umed on this site....is this the general ***umption...that there is a difference between general intelligence and seed AI ?
<2> two different things
<1> there shouldn't be... i'd bet that the former u can only have with the latter but that's just my ignorant 2 cents
<1> well do they have either working?
<1> heh
<2> the main thing about seed AI is that it's recursively self-improving
<1> is it beleived we have a sort of seed AI kind of way of doing thigns?
<2> and probably an easier way to general AI
<1> yeah that's what its saying
<0> when i read PDFs coming out of MIT, i have to be really honest here, there is *way* too little attention paid to behavior.
<1> right ok
<0> In fact, the word "behavior" is hardly ever used by these researchers
<1> hehe
<2> it's just not well enough understood yet, all this stuff
<0> Im actually going to start writing these researchers
<1> nice
<1> that's what the world needs more of i'd say
<0> athena the "singularity" was made up by ray kurzweil. He is notorious for engaging in speculative futurism.
<2> no, kurzweil didn't come up with the concept
<0> yes, he did
<0> in fact he wrote a book were he coined it
<2> but remember also, his book TSIN is meant to be a *popular* work, accessible to the m***es
<0> there is no serious researcher in AI talking about some singularity.
<2> i heard about it first from Vinge
<0> I dont care if half of the wikipedia stub talks about it
<2> that's because it's irrelevant to AI research
<0> the singularity is not a reasonable idea. Its futurism..it borders on science fiction.
<0> pluys, kurzweil is notorious for writing things like this
<1> isn't there an expression like...todays' science fiction is tommorow's science or something
<2> paros: what specific disagreements do you have with what kurzweil says?
<1> not that anyone should waste time reading junk of course heh
<1> mind if I ask what you guys do for a living?
<0> athena first of all there is no such thing as general intelligence. I am aware that many reasearchers TALK ABOUT IT, and WRITE ABOUT IT, but none of them can define it.
<0> What we see in nature is not general intelligence, but particular species who engage in particular behaviors.
<2> what has this got to do with kurzweil?
<2> daleach: i'm a bum atm :)
<0> because in order for a singularity to happen in the next 100 years, you have to first say that there is this thing called "general intelligence"
<0> which im not so sure about
<2> daleach: spending my time reading cog sci books and papers
<0> The whole THEORY of the singularity hinges on this supposition that general intelligence exists
<2> paros: i thought the argument was accelerating human technological development
<1> athena, cool nice
<1> paros, yeah it seems more implicitly aimed at humans heh
<2> paros: that's not my understanding of the concept at all
<0> athena well ok. That's fine. But the singularity then might be the discovery of cold fusion. Or maybe a way to switch matter to anti-matter.
<1> probably not trying to advance the plight of the dog hehe
<2> have you read kurzweil's The Singularity is Near?
<0> athena I was under the impression that the singularity was the creation of a machine that is "more intellgent" than a human being.
<0> isnt that the way he defined it?
<1> this sounds so crazy hah i have to read this...it does sound like science fiction haha
<0> DaLeach yes!
<1> new age and all that
<1> hehe
<2> it might be sure, although i doubt that those would be a new energy source such as CF would be"revolutionary" enough
<1> hah yeah
<2> daleach: yes, read the book and see what you think
<1> yeah I will
<2> paros: no, the Singularity isn't about the creation of any one machine
<1> i wrote it down..going to put it on amazon put a couple on my wish list
<1> ledoux ,, pfeifer kurzweil
<1> what is AIMA?
<1> that was mentioned earlier
<2> it's about ever more rapidly advancing technology, blah blah, until things have changed so much that, from our current point of view, it's incomprehensible
<2> daleach: Artificial Intelligence: A Modern Approach, common college AI text
<2> if you read ledoux, read synaptic self, not the earlier one about emotions/personality
<0> Yeh synaptic self is the title to read
<1> funny becasue i was just talking to someone the other day
<1> how i think emotions
<1> is such an under appreciated important part to this all
<2> important for humans, yes, but not something necessary for AI
<1> well ...
<1> u know how they always say
<1> like when a computer can desire..
<1> not a turing test but like i guess we'll ***ume its intelligent
<1> i tihnk the whole aim should be get them to desire...so they have a whole purpose to think... i mean at some level that's goign to have to be the goal...why else would it think..? ....its like we're goign to "discover" ai and never know it...becuase why the hell would a computer want to tell us hehe
<2> in fact, saddling AI with emotions would be a recipe for disaster... it's an utterly anthropomorphic view that they're necessary
<1> ok i'll hear that
<2> of course goal systems are necessary
<1> i'm conscious of a lot of that i'll give u that
<1> ok its "goal systems"
Return to
#ai
or
Go to some related
logs:
#gentoo
#suse
#qemu
xorg.conf Load dli
No shadows XGL Ubuntu
#mysql
RF Linux luit dsl
split ebuilds e_modules
#math
kubuntu epson stylus CX4500 problems
|
|