Bradley turns the Writing Center recorder on.
Bradley says, "just so I don't forget"
CosmicCat says, "Committees... electing chairs, tabling motions, having sub committees (side boards). They're all furniture salespeople in disguise!"
Bradley will edit out the top few lines of the log.
The tutor [to Bradley]: We cannot edit your paper for you. But we will read it for you and with you. Then we can talk about our mutual impressions of the paper. This way you can see the process your audience goes through when they read your paper, and you can find the places that may make them stumble.
Bradley nods at the tutor.
RobertR [to harvel]: "did you see this?"
Bradley [to the tutor]: But it's so hard to write!
Bradley smiles.
harvel [to RobertR]: yes
RobertR [to Bradley]: the tutor is our resident AI *8^)
Bradley grins
Bradley says, "Ah, local examples!"
GregS wonders if the tutor will respond to plagiarism.
GregS is tossing out key words, fishing for responses. ;-)
Bradley says, "One of my projects for the summer is improving our Turing bots on MOOville"
RobertR [to GregS]: I have forgotten many of the tutor's keywords
Bradley says, "and of course my project, the ideology of ease, has a lot to do with AI, the demands folks place n it"
RobertR [to Bradley]: OK, let's start
RobertR says, "Perhaps we can begin with introductions"
Bradley is Bradley Dilger from the University of Florida, Gainesville, and chair of C&W Online 2001.
Bradley says, "co-chair, really, with the fine committee we've established"
CosmicCat is Andrew Lee (mostly at work (in the UK) so apologies if I get whisked away by some real-life person thing in the next 2 hours)
RobertR is Robert Royar the presenter for this session
GregS is Greg Siering, Ball State University
harvel is Michael Slone, a colleague (former student) of RobertR
RobertR says, "I believe CosmicCat is presenting on the panel about AI at the F2F conference"
RobertR says, "I hope CJ will be able to make it some time today"
Bradley says, "Great, I'll have to check that out."
CosmicCat says, "Yes, I'll be there (Hi Robert, pleased to meet you!)"
James ducks into the wc
RobertR says, "My interest in AI began with the film 2001 (1968 I think)"
James waves to everyone
RobertR says, ":wave to James"
RobertR ooops
CosmicCat waves back
[ 9:18 am ]
CosmicCat says, "What was it about the film that caught your imagination with AI?"
James grins at RobertR, does that too
Bradley, weirdo, hasn't seen the film but is still interested in AI.
James thinks the 2001 era is interesting-----that's when NASA was adding mechanical appendages to mice too----late 60s
CosmicCat says, "It's been re-re-re-re-released here in the UK and available in all it's cinematic glory, hooray! (Now all I need to do is find a cinema that shows it...)"
RobertR says, "Actually, I can remember in the forth grade (1966) thinking how wonderful it would be to have a machine which would store all the books for my class and have a ham radio connction to my glasses which would magically display whatever book I needed when I took a test. That is not actually AI, but I also wanted it to help me find what I needed to read."
CosmicCat says, "In a lot of ways, that's happened - perhaps not in the same sense you envisioned it, but we have PDAs that hold books for reading. "
CosmicCat says, "(A bit noticeable in a test though... all that pen tapping going on)"
James says, "well, and as i understand your location of AI in collaborative agency between humans and machines, the book-storing machine and ham radio isn't too far from AI, right?"
James says, "errrr-----aren't too far . . . . .""
James learns to make subjects and verbs agree
RobertR says, "2001 was beyond me then, but I felt as though it showed a wonderful future. Then a few weeks after the film came out Walter Cronkite had one of his 21st Century programs to fucus on computers. In that he showed the HAL shutdown scene as HAL began to sing Bicycle Built for Two more slowly. The scene in the Cronkite show ended with an actual 1968-69 era computer singing "Daisy, Daisy...""
The tutor [to RobertR]: It is as though I lived in a computer. I eat, breathe, drink, and sleep computers. Why is that? They don't pay me enough, but they buy new hardware every year. No wonder "hey mon" gets so upset about our supplies. Sometimes the hardware frightens the clients. I caught a computer virus suriv retupmoc a thguac last year.
James grins
CosmicCat says, "Way back a few decades ago, that kind of thing was "wow".. these days we get some really quite nifty bits of software published and don't bat an eyelid. "Oh look, another voice recognition system, just like the last one, ho-hum..."
RobertR says, "No, they aren't too far. "
RobertR realizes the effect of lagtime
James nods to RobertR, was battling that this week and is very sympathetic
Bradley [to RobertR]: Do the best you can with the net you've got... :/
CosmicCat says, "It's like we're getting numbed to progress so when something spectacular happens, we may not even notice"
RobertR says, "It's a T3 on a local fiber ATM, and it is too slow."
James nods to Bradley------we'll hang in here with you, RobertR----no worries
Bradley nods to RobertR. "We have a big pipe, but it's mismanaged. Ah well."
RobertR shows the slide "slide1" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * Artificial Intelligence and Writing Instruction * * Artifice and Techne * * ----------------------------------------- * * Presented by Robert D. Royar (aka RobertR) * * Morehead State University * * ------------------------------------------ * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *RobertR shows the slide "slide2" on the episcope.
* * * * * * * * * * * * * * * * * * * * ********************************************************** * * * An On-Line Discussion in Preparation for CW2001 * * Session C.1 (Friday 18 May @ 2:15 in RB 105) * * Artificial Intelligence: Cultural Perceptions, * * History, Theory, and Practice * * * * Read the program at * * http://www.bsu.edu/cw2001/program.htm * * * ********************************************************** * * * * * * * * * * * * * * * * * * * *Bradley notes the time.
RobertR shows the slide "slide3" on the episcope.
* * * * * * * * * * * * * * * * * * * * ********************************************************** * * * VISIT THE COMPANION WEB SITE * * at * * http://www.kcte.org/etc/ai/index.html * * * ********************************************************** * * * * * * * * * * * * * * * * * * * *[ 9:28 am ]
RobertR reminds folks that there is a Ceilidh forum for this presentation at the KCTE site
RobertR shows the slide "slide4" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * What if the locus for AI were between the * * human and the machine, rather than * * in the machine? * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *RobertR says, "Slide 4 is the crux of the biskit for this presentation."
Bradley says, "Yeah, that's part of the problem with traditional conceptions of AI, right?"
RobertR says, "I believe so, that the intelligence is a shared reality."
Bradley says, "It's sort of a service-for-man thing... the computer washes the dishes, does the laundry, makes dinner, whatever..."
The tutor [to Bradley]: It is as though I lived in a computer. I eat, breathe, drink, and sleep computers. Why is that? They don't pay me enough, but they buy new hardware every year. No wonder "hey mon" gets so upset about our supplies. Sometimes the hardware frightens the clients. I caught a computer virus suriv retupmoc a thguac last year.
RobertR says, "I mean it is one, but it is not usually discussed that way."
RobertR remebers that c*mputer is a bot attention getter
James says, "would you say a little more about how the shared locus can be conceived as "artificial" intelligence?"
Bradley says, "I agree. I'm doing a lot of work with interfaces, and most writing on the subject seems to tread the intelligence of the c0mputer as a black box"
Bradley says, "when clearly it's not -- as our two different ways of spelling comput3r demonstrate"
RobertR says, "part of what I mean by that shared locus (or of its importance) is that we may be able to achieve more by making the user more computer savvy."
The tutor [to RobertR]: It is as though I lived in a computer. I eat, breathe, drink, and sleep computers. Why is that? They don't pay me enough, but they buy new hardware every year. No wonder "hey mon" gets so upset about our supplies. Sometimes the hardware frightens the clients. I caught a computer virus suriv retupmoc a thguac last year.
RobertR hushes the tutor.
Bradley [to RobertR]: That's a tough one. Many people are very resistant.
CosmicCat says, "Having to locus of the AI between Human and Machine sounds like the use of a spell checker - the machine is vaguely using AI by recognising that a word isn't "quite right" (which means the word is not in the dictionary) and the human intelligence does the real intelligence by saying whether the word is correct or incorrect."
Bradley thinks "The computer thinks for me" is an all-too-common defintion of AI.
James thinks------hmmmm
James says, "so artificial intelligence is a shared intelligence?"
CosmicCat says, "(And that can then be applied to any bit of software - they all do something "clever-ish", it's the human that applies the intelligence to make the software useful)"
RobertR says, "yes, the speller is a good example--but we need to extend the speller to make it more useful by no longer assuming ignorance on the user's part."
Bradley [to James]: Definitely
James nods to Bradley
James says, "would cookies, then, be an instance of human-computer interaction, but in terms of the way computers react to the packets of information in the cookies?"
RobertR says, "imagine if our text editor could "remember" our error patterns, our word frequency, phrases we use often and that it could help us find ways to improve our writing based on analyses that are "aware" of our text personae""
[ 9:38 am ]
RobertR says, "cookies could be an example of ways that computers track their "knowledge" of our persona"
Bradley [to James]: Sure, albeit a lot less complex than the 'remembering' RobertR just mentioned
James says, "i dunno why i'm so full of questions----sorry. i think it's just that i'm really interested in your presentation, robertr"
James says, "what about clippy in ms word?"
CosmicCat says, "We have grammar checkers that go halfway towards the error patterns - they can spot some of the common grammatical errors but not remember them (I remember some of them have a report assessing the "correctness" but doesn't do anything more about it)"
Bradley [to RobertR]: I think that's called 'intelligent agents' by some -- right?
Bradley grins, Clippy!
RobertR says, "yes, and clippy is a cutely annoying example."
GregS [to James]: perhaps if Clippy started recognizing your style, rather than relying on its own generic rules? And I prefer the dog assistant over clippy.
James . o O (Would you like to say something in the MOO? Would you like help? Would you like to work without help? Clippy loves us.)
CosmicCat says, "(Sad news for all... Microsoft will be getting rid of Clippy in the next version of Office...)"
RobertR says, "I wish I had a clippy bot in XEmacs. I think that would be a great agent to program. And there is another point I would like to make. We need tools that let us actually program intelligent agents if we are to make them act intelligently."
Bradley laughs, "Poor Clippy. But we need stuff like him to move forward to a more sophisticated agent."
James grins, imagines the board meeting: Hi, I'm Clippy. Would you like help removing me from your software? Would you prefer to remove me on your own?
Bradley [to James]: Clippy learns to downsize. Great.
James woos, hopes not!
James [to Bradley]: i was meaning to remove itself
RobertR shows the slide "slide5" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * ASSERTION (LOCUS) * * --------- * * Artificial intelligence is attainable with current * * systems if we (a) accept the ersatzness of artifice * * and (b) learn to see ourselves as the interface * * between the software and our task. * * ROYAR * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *Bradley [to RobertR]: You know, yeah, what are the generic frameworks for agents?
GregS [to RobertR]: So if we had some way of extending/personalizing Clippy in some way?
Bradley hasn't seen an 'artificial intelligence agent developers kit' on freshmeat.net...
RobertR says, "well, i would want to be able to completely rewrite clippy"
James [to RobertR]: what sorts of rewrites would you do?
CosmicCat says, "With these intelligent agents, the intelligence is limited to doing one specific thing - asking, say a search engine web crawler, to solve a temporal logic problem gets a "huh?" response. The same would go for a hyper-Clippy - it's great for doing well defined simple stuff, but asking it to do more than it's expected gets it stuck (one of the reasons why people find Clippy so annoying...)"
Bradley says, "I agree with this assertion 100%. But I think most folks would refuse it because it places a lot of weight on the user."
RobertR says, "the agent needs to be able to tap into a store of information beyond rules it has given. It needs the ability to learn from stores (DBs) which have a different structure than was originally intended."
Bradley nods at CosmicCat.
James nods to Bradley
Bradley [to CosmicCat]: However, we should be able to grow that role. Heck, all other software seems to be expanding beyond a single task...
Bradley smiles ruefully.
RobertR says, "ComisCat's point is important. I have begun to think that writing a paper about a David Foster Wallace novel is a restricted domain--similar in scope to the restricted domain that diagnostic programs work within in medicine."
RobertR says, "Why do we see a novel as less artifical, more alive than we do unix?"
[ 9:49 am ]
Bradley nods at RobertR.
Bradley says, "Yeah, McLuhan asked the same question about the fixed game shows in the late fifties -- he wondered why controlling outcomes there was so different than in movies."
RobertR shows the slide "slide6" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * All acts of artificing are mediated by the inter- * * face between a designer and a design space. * * ALEXANDER * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *James [to RobertR]: because many more people have read novels than operated in unix, thus making the novel more "real" for people in general?
James [to RobertR]: within specific discourse communities, etc
RobertR says, "we have developed the software we need to make the novel live. "
Bradley wonders about the "reality" of the novel.
James hmms
James says, "i don't mean that the novel itself is more "real"------i mean that "novel" as a concept is more mainstream"
CosmicCat says, "The novel is also aimed at appealing to us emotionally whereas Unix is seen as a tool for getting a computer to do something (um, okay, *some* people find Unix emotionally appealing...!)"
RobertR says, "I imagine that many of us find that novels become part of our day-to-day existence. I find that I think as though I am in many of the novels I read as I am reading them--Infinite Jest is a good example."
RobertR admits that he sometimes thinks of the real world as a file system.
James [to RobertR]: interesting! so what shape, form, and role do humans take in that file system?
RobertR says, "yesterday in a workshop I caught myself looking underneath a piece of paper to see if I could discover the code that created the formatting I was seeing."
Bradley grins at RobertR.
harvel says, "Did you find it?"
RobertR [to harvel]: unfortunately, I did not.
harvel says, "I find that Unix can become part of my day-to-day existence just as novels can."
RobertR says, "That relation between the paper and its formatting is rich in that the original concept of looking behind text on the screen for its underlying source grew out of desktop and print metaphors."
Bradley says, "Many novels are a pretty complete world, too. To me computer operating systems don't have that organic totality -- even ones like MacOS where the interface is wedded very closely to the hardware."
Bradley isn't using 'organic' to mean 'wholesome', but is still thinking in a McLuhanite sense.
James nods to Bradley, agrees
harvel says, "Also, novels tend to give the reader little chance for interactivity; it's just a transcript. If a reader looked at a transcript of a session with a computer, e might find the computer to be much more alive"
RobertR [to harvel]: would you say more about how the unix systme can be similar to novels?
Bradley nods at harvel.
CosmicCat says, "Also Unix (or any other operating system) has a limited vocabulary to perfrom well defined tasks, whereas novels can explore a rich seam of language/emotion/concept"
[ 9:59 am ]
RobertR says, "what if language/emotion/concept were a file structure? what would the child of concept be and what would be language's parent?"
harvel [to RobertR]: It is not so much similar to novels as reading a novel is to participating in a conversation. When I read most novels, I feel as though I am a mostly nonparticipating observer of the events described. When I use Unix, it feels as though I am conversing with the computer in order to get what I want done done
Bradley says, "/dev/anger"
RobertR [to harvel]: the difference is similar to Socrates' distinction between dialectic and rhetoric--the computer being dialectic.
CosmicCat says, "Hmm.. parent concepts of language/emotion/concept... basic human needs (love/acceptance/contribution/growth and so on). I might come up with a concept to explain to other people to want to be accepted by them, or to add to human knowledge (contribution) or to push my thinking further (growth)..."
RobertR has disconnected.
RobertR has connected.
RobertR oops got bumped
RobertR shows the slide "slide7" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * ASSERTION (Poesis) * * --------- * * Well-designed technology invites and facilitates * * interactivity between the technology and the user. * * Poorly-designed technology creates a barrier * * between the two. * * ROYAR * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *GregS excuses himself... was on the phone for awhile with a faculty client, and now must run across campus to attend to a probematic website. Thanks, Robert... sorry I have to run.
A DoubleCola that harvel borrowed from the Writing Center has become stale and has recycled itself.
RobertR wave bye to Greg
GregS has disconnected.
Bradley says, "another fine assertion"
CosmicCat says, "Oop, so where were we... we had a difference between Unix and a Novel - how does that relate to the implications of AI in teaching?"
CosmicCat says, "(And moving the locus of intelligence from machine to human)"
Bradley says, "Of course this way of thinking poesis, again, takes a lot more effort than just building (or assuming) a black box."
CosmicCat says, "Yes, once we know what we want to build, building it becomes easier."
RobertR says, "to an extent the main reason we don't see the interaction with software to be as rich as the one we feel between ourselves and novels is that we are not yet familiar enough with software. As we become more used to manipulating it ourselves, the familiarity should make us better able to create that intelligent "locus""
GregS's friends arrive to cart him off to bed.
[ 10:09 am ]
RobertR says, "we have resisted the effort. Look at the way we often teach computers to people--as though we wished the computer to go away. We don't teach reading books that way--instead we encourage active engagement."
Bradley grins at RobertR.
CosmicCat says, "Is that familiarity with the interface to software or familiarity with the construction of software to build the sorts of systems we're aiming to create?"
RobertR says, "of course books have a long history that ties them to the sacred"
RobertR [to CosmicCat]: I believe it is the familiarity with the construction and an understanding of what the machine is doing when it "thinks" and of how the software is structured--its flow of instructions.
Bradley says, "That history of books is really critical here -- perhaps that sacredness is sort of a zero-sum game"
RobertR says, "If I were to take the role of computer apologist, I would argue that we cannot xpect software to be intelligent so long as we want it to be invisible."
CosmicCat says, "On the computer programming side, there is a lot of work with the construction and understanding of what the machine is doing, so in a lot of ways we have that aspect covered. A lot of the problems that arise stem from not knowing what we want the computer to do and the inability to specify those requirements to get a computer to be able to perform the task"
Bradley says, "I'm wondering if, given that books are still held up in the way they are (reading is fundamental, etc) there's room for computers to take a more important role."
CosmicCat says, "We can get a computer to "pretend" to have intelligent behaviour but not have any intelligence at all - Playing Chess was a problem considered to need lots of intelligence, now we have software that doesn't need to "understand" chess, just have a database of possibile positions and whether they lead to a high probability of winning."
CosmicCat says, "Oh, I think computers already are reaching that sacred spot in that people are awed at the potential of the computer without realising it's really still a mchine that pushes ones and zeroes around"
RobertR says, "using books again as the analogy--or more generally, writing as the analogy, there was a time when the sea of stories was new that the magic of words was a given. We needed technicians to create them, and they were not able to make them do many things in written form that could be done in speaking. But at some point the written word became alive and (like the virus it is *8^) ) directed our own development."
Bradley says, "Well, but I don't think awe is the kind of sacredness folks hold for books."
CosmicCat says, "And even before that, there were storytellers who would walk from town to town and be greeted like kings for the ability to enchant people with oral stories (the writing just made the story readily available when the storyteller wasn't there, perhaps devaluing the magic of the story since it was instantly reproduceable)"
[ 10:20 am ]
CosmicCat concedes, awe was not the best of words to use in conjunction with books. Although awhen books were introduced, I suspect they held the same level of awe as perhaps the computer does now (history repeating itself? We may be able to learn something from that)RobertR says, "there is a critical mass of users/practicianers that once we reach it everything changes. Once we realized that writing was just another technology and that the written word was not foundational, I think it paradoxically began to live on its own. Once we decide that intelligence is really just the skin of the onion, the appearance of intelligence is as good as it gets. (Perhaps close enough for government work.)"
Bradley nods to CosmicCat, "Definitely -- that is true."
CosmicCat says, "Since we're not exactly sure what intelligence is, there may not be any way to tell whether something is intelligent or just has the appearance of intelligence. Otherwise we'd be calling a thermos flask intelligence since it "knows" when the contents are cold and then keeps it cool, and it "knows" when the liquid is hot and changes behaviour to keep it hot."
RobertR shows the slide "slide8" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * ASSERTION (Intelligence) * * --------- * * Artificial Intelligence is a form of test. It is * * an experimental context. * * ROYAR * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *RobertR shows the slide "slide9" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * If context and performance interact, there are * * almost certainly important features of the * * situation which won't agree between experiment * * and mundane circumstances. (90) * * LAVE * * ---- * * "What's Special about Experiments * * as Contexts * * for Thinking?" * * --------------------------------- * * The Quarterly Newsletter of the Laboratory of * * Comparative Human Cognition * * October 1980 (2.4) * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *Bradley says, "Robert, I'm really liking these assertions."
CosmicCat says, "Experimental or experiential context...? We see something as intelligent if it does something unexpected and requiring some sort of distinction that changes it's behaviour. I'm wondering whether the thermostat was considered intelligent when it was first invented, but now it's so common that we don't even consider it as intelligent."
RobertR says, "A similar question may be whether humans will be considered intelligent as we learn more about brain structure and processes."
CosmicCat says, "Once we've experienced something that could once have been considered intelligent, is it then no longer considered intelligent and we move on to wanting a more advanced behaviour (yow, we're becoming intelligence addicts)"
RobertR says, "It is difficult to watch crows and not think they are highly intelligent--and they know how to fly which we didn't until a short time ago <g>."
CosmicCat says, "Ooh yes, if in a few years time we find out that we're just very good at applying rules to respond to the world - the appearance of intelligence versus actual intelligence"
Bradley wonders what the classical definition for intelligence is...
RobertR shows the slide "slide10" on the episcope.
* * * * * * * * * * * * * * * * * * * * ************************************************************ * * * [T]he AI must be interactive - that is, it cannot * * simply take some input data, such as a research * * paper, and regurgitate it, such as by producing a * * summary. * * * * Ben Garney * * http://kcte.org/etc/ai/41003d6500A-4126-1328+00.htm * * * ************************************************************ * * * * * * * * * * * * * * * * * * * *[ 10:32 am ]
CosmicCat says, "An AI that wasn't interactive still has elements of intelligence - just that the intelligence is limited to such a small domain that adding interactivity would confuse it immensely. "
Bradley says, "Actually, summarizing is a pretty difficult skill -- I think that's a bad example to support that assertion"
RobertR says, "There will likely be some effect from our interacting with computers in many endeavors. Think about how we react to dogs and horses. They become friends--part of our lives, but it was not also so. Computers may someday be friends in this way. Then we will need to see them as intelligent--at least as we see other animals as being intelligent--because we will share common experiences."
RobertR [to Bradley]: I had hoped Mr. Garney, who made that assumption, would attend.
Bradley nods at RobertR.
RobertR shows the slide "slide11" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * We are continually shaping a common perspective * * when we speak a common language and so are active * * participants in the communality of our experience * * of the world. (110) * * GADAMER * * ------- * * Reason in the Age of Science * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *Bradley says, "It seems ok to me up to the example. I'll post back to him on the forum."
RobertR [to CosmicCat]: what are example of current computer languages that AI researchers like to use?
CosmicCat says, "That reminds me of some interactive scultpure I read about. In 1960-something in Paris, they had something that looked like a crane attached to a speaker and some position sensors. If someone cooed softly at it, the movements would be slow and gentle, if the sounds were loud ("threatening") the movements were fast. Visitors to the exhibit were reported to say they saw the lumps of metal as alive because it conformed to what they expected as behaviour from an animal"
CosmicCat says, "These days, there are still projects done in list based languages, but not that many commercially (things like Lisp and Prolog). Most of the projects seem to be done in C++ these days (perhaps because of higher availability of C++ programmers compared to other languages)"
RobertR says, "There is also the little heads in the MIT labs that respond to stimuli in similar ways. People react to them as though they were alive."
harvel [to RobertR]: "Such as Cog?
CosmicCat says, "Reacting to machines as if they are alive fools people into thinkin ghtere is some intelligence in there somewhere, when the machine is just applying rules that are known to get certain responses from people (now *that's* artificial artificial intelligence...)"
RobertR [to harvel]: yes, I think Cog was one of the two I read about.
RobertR shows the slide "slide12" on the episcope.
* * * * * * * * * * * * * * * * * * * * ******************************************************** * * * ASSERTION (Artificial Intelligence) * * --------- * * Context and knowledge of prior events both * * determine and undermine our current concepts of * * knowledge and of intelligence. How we have * * interpreted events (and in this case events can * * be texts or other software) governs the way we * * react to and approach the human-computer inter- * * face. * * ROYAR * * * ******************************************************** * * * * * * * * * * * * * * * * * * * *Bradley wonders about the availability of metatools, code that writes code...
RobertR says, "compilers do that, but tat is not exactly what you mean I imagine."
[ 10:43 am ]
CosmicCat says, "Metatools don't seem to be doing too well these days. I think people got a bit disillusioned with the 4GLs that were touted as saviours of the programming world a few years ago. There is one system I've read about that is meant to take a limited natural language input and churn out a program."
Bradley says, "Right, I'm thinking of a tool that would actually extend the code base -- more than just building upon a given code base in different ways."
Bradley says, "There are the translators that have been written in support of the DeCSS lawsuits -- I find those incredibly interesting."
RobertR says, "Isn't one of those expressible as a 128 digit prime number?"
RobertR [to harvel]: is 128 digits far too large?
Bradley doesn't know.
harvel [to RobertR]: 128-digit primes in binary certainly aren't too large
Bradley says, "let me find a link -- it's a tool that translates C to English and back, if I remember correctly"
harvel [to RobertR]: One of what is expressible as a 128-digit prime number?
Bradley says, "the point being there's a bridge between "human" and "machine" intelligence (though clearly both are human in origin)."
RobertR says, "10,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000"
harvel [to RobertR]: The 37th Mersenne prime---2^3021377 - 1---has more than 128 digits :)
CosmicCat says, "There's this .net initiative from Microsoft which is meant to be the new way of doing things. .Net downloads small programs that perform specific tasks and communicate with each other so it's like a distributed software system. This may accelerate the pace of practical AI development by adding capabilities a bit at a time until a more complete intelligent system can be evolved"
Bradley says, "http://www.mit.edu/~ocschwar/C_English.html"
RobertR [to harvel]: on program to decrypt dvds is expressible as a single 128 digit prime number which means that if the movie industry is correct, it is the only prime number that is illegal to display or transmit to others.
CosmicCat says, "*Eek* Did we just break the law or do we have to print out all the digits first...? :)"
[ 10:55 am ]
harvel [to CosmicCat]: All the digits would have to be printed out first
harvel [to RobertR]: There is a 1401-digit (in decimal) prime that corresponds to a DeCSS program
CosmicCat says, "*phew*! The only laws I want to break are the laws of physics..."
RobertR [to harvel]: that is what I was thinking of. I could not remember the size of the number.
Bradley grins. "Laws, woo!"
RobertR says, "we are nearing the end of our appointed time. I want to thank the attendees and invite you to the f2f presentation in May. Also, I invite you to participate in the forum at http://www.kcte.org/etc/ai/"
CosmicCat applauds. Thank you for arranging the session Robert!
Bradley will post a note to Garney's assertion, which looks interesting.
Bradley [to RobertR]: Thanks much -- I'll definitely be at the panel in Muncie
CosmicCat says, "What have we learnt from the session (leading into a summary...)"
RobertR [to CosmicCat]: You're welcome. It was both fun and helpful.
RobertR says, "we may have learned that judging the difference between artificial intelligence and artificial artificial intelligence is tough."
Bradley nods to RobertR. "Also, I think there's a lot of work to be done to unpack common assumptions regarding the role of 'interface' in determining the bounds of 'artificial'.
RobertR [to Bradley]: Yes
CosmicCat says, "And the pretense of intelligence is virtually indistinguishable from real intelligence (mainly because we're fuzzy on what intelligence actually is)"
[ 11:06 am ]
Bradley says, "Any other summaries? I need to move along to an appointment, and the log moves with me...."
CosmicCat says, "... and the newness of Computers and Intelligence now is akin to the newness of the printed word when it first came available."
CosmicCat says, "(perhaps!)"
CosmicCat has no more summaries that immediately spring to mind
RobertR does not have any either.
Bradley says, "Thanks again for a great session."
CosmicCat says, "Yes, thank you all!"
CosmicCat says, "I'll see you all in Muncie next month to perhaps go even further with this (if we're not spellbound with all the other things that will be happening...)"