erich@near.cs.caltech.edu (Erich Schneider) (11/01/90)
>>>>> On 31 Oct 90 17:23:10 GMT, rlp@beach.cis.ufl.edu (Trouble) said:
+> Maybe it's a trivial example, but I'll call a computer "intelligent" when
+> I can dial randomly across the radio, drop in on a song, and have it
+> identify the type of music, artist, and song, and do it as quickly as
+> a human being. Let's see how much processor power it takes to pull
+> that trick.
This is a trivial example, but I get your point. One point I would like to
make is that "processor power" would be the only thing to needed to solve
this problem. Given a list of all of the types of music/groups/songs a
human knows, along with their characteristics, it's just an algorithmic
process (i.e. a Turing machine, by Church's thesis) to perform the ID.
If, however, one could dial randomly across the radio and have the computer do
the identification, and then _without a human manually changing the program_
have the computer do something radically different (e.g. solve a complicated
differential equation or physics problem) just by showing it the problem;
I would say that program is possibly "intelligent".
Followups to comp.ai. While AI's are part of the cyberpunk literature, this
discussion has moved into a more AI-centered region.
--
erich@tybalt.caltech.edu or try erich@through.cs.caltech.edu
"Why the hell anybody plug the likes of you into a deck like that? Thing ought
to be in a museum, _you_ ought to be in grade school."
-William Gibson, _Count Zero_rlp@beach.cis.ufl.edu (Trouble) (11/02/90)
In article <ERICH.90Oct31191951@near.cs.caltech.edu> erich@near.cs.caltech.edu (Erich Schneider) writes: >>>>>> On 31 Oct 90 17:23:10 GMT, rlp@beach.cis.ufl.edu (Trouble) said: > [ my example of randomly selecting a song for an "AI" to identify ] >This is a trivial example, but I get your point. One point I would like to >make is that "processor power" would be the only thing to needed to solve >this problem. Given a list of all of the types of music/groups/songs a >human knows, along with their characteristics, it's just an algorithmic >process (i.e. a Turing machine, by Church's thesis) to perform the ID. Aaah, but sooo much processor power. Man, I'd like just a teeny bit of it on my desk...and the Hoover Dam to power it. I realize it's basically just a matter of a "brute force" algorithm to have a computer do it, but consider how fast even the average high-schooler can tell you a tune is Warrant's "Cherry Pie," or if it's just some of that "goofy classical stuff." >If, however, one could dial randomly across the radio and have the computer do >the identification, and then _without a human manually changing the program_ >have the computer do something radically different (e.g. solve a complicated >differential equation or physics problem) just by showing it the problem; >I would say that program is possibly "intelligent". I like this extension of the concept. Or how about this: the AI works on the problem, a co-worker flips on the radio, the AI says "Yuck, I hate Vanilla Ice, change to the easy listening station," and keeps working on the problem. Consider that humans do it all the time. Are we multitasking, context switching, "shadow processing" (two things running on the same processor literally at the same time [though at different priorities, perhaps], not just switching really fast from one thing to another), or something else? Disclaimer: Sorry if I've mangled any terminology or paradigms or anything. My only exposure to AI, academically anyway, was an "AI Concepts" class I took during my MBA program. Talk about feeling like a fish on a bicycle...I definitely did not dig predicate calculus at that time. So, maybe I'm just being a nay-saying gadfly. Bob -- rlp@beach.cis.ufl.edu Air: PP-SEL AMA # 541283 Road: 750 Ninja DoD # 0068 Water: NAUI OW-I <=- -=>
arh@sage.cc.purdue.edu (Eric B) (11/02/90)
In <25224@uflorida.cis.ufl.EDU> rlp@beach.cis.ufl.edu (Trouble) writes: >... Man, I'd like just a teeny bit >of it on my desk...and the Hoover Dam to power it. Makes one realize the efficiency of neurons! >I like this extension of the concept. Or how about this: the AI works >on the problem, a co-worker flips on the radio, the AI says "Yuck, >I hate Vanilla Ice, change to the easy listening station," and keeps >working on the problem. Consider that humans do it all the time. >Are we multitasking, context switching, "shadow processing" (two >things running on the same processor literally at the same time >[though at different priorities, perhaps], not just switching really >fast from one thing to another), or something else? >Bob Consider this... you're listening to music, foot tapping to the beat (assuming one has rhythm 8-), while you're doing something simple like: talking on the phone, typing a message, or reading news. Meanwhile, a whole host of essential body functions are monitored & controlled. The fact is that human brains are divided up into very specialized areas. Many are hard-wired to their corresponding body parts. When one area is damaged, its function is usually lost, although it is possible for other neurons to "rewire" themselves, gaining a partial recovery of the lost function. The idea is that brains are specialized multi-processors with dedicated hardware. A PhD. in Brain Psychology could use all the big words to describe it, but I wouldn't understand him. I would recommend taking (or auditing) a course on brain function or brain behavior. At Purdue, it's Psych 220. When I took it, I continually analyzed how the brain could be represented by silicon. Amazing stuff. I actually developed an idea for an artificial eye. I'm just waiting to apply optical computing to it. The one problem is using a {optic,wire} to neuron connection. We'll see... just spinnin' my cycles away, Eric G. Bolinger 8-) arh@sage.cc