First off, I admit I can very easily be proved wrong in this.
But I've been downloading trans-oceanic quicktimes for the past couple of hours (at 28.8) and I still think it's a hoax.
Depictions of robots in sci-fi movies have always had the problem that they skip over the 'intermediate forms' between obviously-not-human and much-too-realistically human-with-a-few-small-flaws. And as far as I can see Honda's materials commit the same sin-- where are the intermediate forms???
If you see a demo that in every detail resembles a person crouched in a special suit... what does Occam's razor say you should conclude?
Why did they need to give it a humanoid torso and arms? Why do these also move exactly like the human versions? Why is the head concealed in a visor, if not to give their actor a way to look around? Why do the P3 quicktimes show only the cliched, anonymous box-with-blinking-lights inside, instead of showing off the real high-tech in the legs?
Some time ago I read an article on bipedal robotics that said that the
guys at honda used data from movements of real people to make their
robots look as if they moved like a human. It is not a inherent
smoothness that results from the way this thing is constructed.
Humans actually use much less energy for the same movements
because they use the natural swing of their legs. But it looks cool
anyway.
This would explain why it looks like a total hoax, with a crouching guy in there. (I still want to see proof that it's not.)
But if it's just 'playback' then it's not even AI-- it can't possibly adapt to changed terrain... right?
"...a page designed with a CSS2 style sheet can have multiple
media types, the cellphone section would describe how to render
the page on a cellphone, while the browser sections would identify
how to display it under a browser..."
Um, will you at least put the "cellphone section" first, then... with an automatic cutoff that keeps the rest from flooding my bandwidth? ;^/
The theory here was supposed to be that you have exactly one section, that's just rendered differently for each device. As a theory that's very elegant, but as an interface in the real world it's turning out, over and over, to be a hollow dream.
McCluhan taught us in the sixties that the medium is the message-- you can't put the message in one file and the 'styling' in another. This problem with mobile devices and bandwidth is just the tip of the iceberg for CSS's fundamental wrongheadedness, imho.
The right longterm solution, I've been arguing forever, is to embed the document's styles in exactly the way WordPerfect used to do it-- with codes for BOLD and ITALIC, etc. Display devices can render those however they choose.
But in these earliest days of wireless, when every byte counts, the only way to go is optimised custom pages-- it shouldn't even be HTML, because the tags are too fat.
And when the bandwidth improves enough to handle HTML, it will work best with the same basic HTML that lynx likes. CSS just never solves any real problem along this evolutionary path!
I see this as just another top-down, ivory-tower W3C boondoggle, like XML. Those guys couldn't design a usable interface to save their lives-- look at how unreadable all their specs are! TimBL is the guy who thought up "aitch-tee-tee-pee-colon-slash-slash" for crying out loud...
This may already be a cliche, but if it isn't it should be:
The best games are ones that define a limited number of abstract 'dimensions' and then milk every possible drop of playability out of them.
When any game makes the jump to 'realistic' representation of 3-D space, it becomes impossible to do this, and the designer is frustrated by having too many degrees of freedom.
My experience is totally atypical, but I think it shows how inadequate the categories of 'mystic' and 'hacker' are...
I started both with mysticism and computers as a teen in the 60s, and I actually walked away from my first programming job because I realized very vividly that it was going to distract me from the search for self-knowledge (which was very big in the early 70s).
I explored mystics like Gurdjieff and psychological systems like Fritz Perls' gestalt psych, and I put a lot of effort into meditation as a way of seeing what was really going on in my psyche/subconscious.
By the mid-70s this range of 'gurus' had narrowed to one who was especially austere and conceptually challenging-- J Krishnamurti. From him I came to see that soiritual freedom was about living without preconceptions or attachments, which is a common theme of a lot of mystical schools (especially Zen), but one that Krishnamurti handled with greater honesty, I thought.
At a certain point then I turned away from the whole mystical 'quest' and just started living, and at that point sociobiology became very inspiring, as did (non-sci-fi) literature.
I formulated a philosophy I call 'Robot Wisdom' that demands-- ultimately-- computer models of human behavior, that fit the requirements of sociobiology but also have the humane dimension of literature.
So my whole thrust has been to wrestle with how to model human emotions on a computer, eg in video games. And I look to great literature (especially James Joyce) for inspiration about the laws of human emotion.
I still feel a lot of sympathy for mystical points of view, because they deal with real psychological phenomena that materialist viewpoints seem to deny, but I don't practice any form of organised worship.
There's a TuneServer in Germany that works on the principle of whether the pitch rises or falls with each note. It claims to analyse wav files of whistling (a lot simpler than mp3s of full instrumentation).
It's based on a book published by G Spencer Brown, the mathematical logician (Laws of Form).
Who gave the W3C the right to publish HTML standards on behalf of the community anyway?
Simson Garfinkel wrote an eye-opening backgrounder that explores this question. A quote:
Cargill says he thinks companies have stopped sending people to meetings because they realize that the General Assembly's Advisory Council Committee
merely rubber-stamps what Berners-Lee wants to do.
TimBL, remember, is the guy who invented 'http://' and who dictated that two 'P's in a row should dispay the same as one (absurdly forcing everyone to add non-breaking spaces).
It's just basic design common-sense that you don't create top-down 'standards' groups who dictate the rules of human factors without ever testing the standards, and without having the slightest understanding of what human-factors is all about.
the real interest in this device wouldn't be in simply disposing of extraterrestrial garbage, it would be in destroying the enemy's satellites.
A challenge for evil geniuses: what's the cheapest way to bring down the most satellites in one pop, assuming you're just going to use the threat for extortion, and don't care if you make satellite-space unusable forever?
You'd probably want something that spews spacedust in all directions over a range of altitudes...?
The kind of project that interests me is a system that uses XML formats to provide clear semantics on the web
That's the problem, all right, but I don't see anyone in the AI world except Cycorp really wrestling with it, and they've mostly given up on human psychology, which is where all the useful web-semantics lie.
For example, suppose you want to find a Drew Barrymore fanpage that discusses how she stays so upbeat all the time-- how do you represent that with a finite set of XML labels?
Nobody knows, and solving it will take a sort of subtle psychological analysis that's the real, gaping abyss in current AI curricula.
I've been arguing since 1979 that the process of building really-intelligent software will require that we first come to deeply understand our own psyches-- robot wisdom. So limiting the research process in order to avoid 'evil robots' seems excessively alarmist.
I saw on a science show where one of the leading DNA experts collected DNA from lots of Egyptian mummies, to explore (among other things) how much inbreeding the royal line really did. That was several years back and I never heard any results so I think it must have been inconclusive.
I'm getting pretty disgusted with the unexamined assumption here, that more speed, more storage, bigger monitors, flatter panels, etc etc etc are ever worth getting excited about, in themselves.
If people put their energy into writing better code instead, we'd be a lot farther along. Gizmo-porn is just a distraction, imho,
"I wonder if Rob and Co will now disclose their stock holdings in the companies they report on (like most respectable journalists)."
To me this seems like the critical question-- if a negative story about VA will cost Rob&c money, then we have to expect they'll spin it in the other direction.
These lines: "I almost quit when I heard about the merger... Since then I have been personally reassured by Larry Augustin..." also give me pause, because *persuasion* is what makes capitalism run, and there's no way to draw a line between a little 'safe' persuasion and a whole lot of utter-crap persuasion.
I never registered but for the last few days the NYT site has been letting me in anyway. I dunno if it's a bug or a quiet change of policy-- I hope the latter.
Chris Byron has the best analysis in this week's NY Observer, I think.
My version (related but not identical):
AOL's value is nowhere near what the market claims
Time-Warner would be nuts to give away so much for so little
What Time-Warner is really doing is milking the bubble while the milking is good, and when it collapses Time-Warner will still hold all the cards
The stockholders can take advantage of the bubble if they're smart, but will lose their shirts if they don't get out in time
The only other explanation is that TW knows that they won't have to open their cable access to others, and can thus force AOL on all their customers (ie, the fix is in).
But I think all the metaphors about strangleholds and big players squeezing out little guys are naive propaganda-- the Web works in exactly the opposite way in the long run, and even a nebbish like Drudge can get a half-million hits a day with no advertising budget.
"I was reading the alt.best.of.usenet or whatever it was called --at the moment they turned on their sewage pipe. Immediately, the newsgroup filled with humorous postings was awash with crap."
C'mon, tell it right...
This was 1993 or '94 he's talking about, and the group was alt.best.of.internet, where people were supposed to repost good news postings from other groups.
That group went thru regular longwave cycles where a few people would use it correctly, and it would draw an audience, so various idiots would try to take advantage of the attention, and people would go away... rinse, repeat.
AOL finally got around to offering newsgroups at that point, featuring a few 'starter' groups as everyone's default, the first of which-- in alphabetical order-- was aboi.
So AOL people started deluging into the group, asking totally random things like "Where can I get soy fertilizer in Ames, Iowa?" or "Anyone in Duluth want to make hot monkey love tonight?"
Now, the first response of aboi regulars was, disgracefully enough, to play sadistic games with the newcomers, making the situation much worse. (Shooting fish in a barrel was the image that came to my mind.)
It was actually me who realized what the situation was at AOL's end, and started trying to get them a message to change the default for the sake of aboi.
They didn't respond until Weemba posted a message with the subjectline "AOL top brass joins Satanic cult in ritual sacrifices" or the equialent, within hours of which the spigot went off.
I was just thinking this morning how I've never enjoyed programming on anything else the way I enjoyed 6502 on the Apple ][... but I *assume* that someday someone will write a GUI that's equally elegant, and gives even beginning hackers the same sense of transparency. Or is this a pipe dream?
When I see "This is an editorial I wrote noting a few points about what Linux has to do before *it moves totally mainstream.* It talks about both the small and the large issues currently slowing the widespread use of Linux." I really have to wonder whether to click the 'it moves totally mainstream' part or the 'Read More...' link. (It would have been even worse if there was another choice like "an editorial I wrote in *this magazine*".)
My philosophy is that you shouldn't ever link to the magazine's homepage (it will be linked from the article) because it's just another opportunity for confusion. And the highlighted text should be unambiguous-- go with "editorial" so I know where I'm going when I click.
(For all I knew, the 'totally mainstream' link might have been to a Corel download page or something.)
Most people have the ability to buy a subscription to one of their academic journals (Science magazine leaps to mind) that would amply inform them of recent events.
"Science News" ought to do them fine, and is a lot easier to read.
Wisdom Tree did a version of Wolfenstein called Super 3D Noah's Ark. Quote from an old review:
"The game designers at Wisdom Tree bought permission from id to remake and slightly alter Castle Wolfenstein 3D to design a game called Super 3D Noah's Ark. In the Wisdom Tree game, players assume the role of Noah trying to get his animals back in their pens, as opposed to a soldier trying to hunt down Adolph Hitler. They also did away with the grenades and machine guns and replaced them with a single slingshot that players used to shoot food at animals. Instead of having soldiers that charged at you, then fell in a bloody pool when shot, Super 3D Noah's Ark has little goats that try to kick you with their hind legs and fall asleep when you shoot them with food."
Slashdot Mirror
But I've been downloading trans-oceanic quicktimes for the past couple of hours (at 28.8) and I still think it's a hoax.
Depictions of robots in sci-fi movies have always had the problem that they skip over the 'intermediate forms' between obviously-not-human and much-too-realistically human-with-a-few-small-flaws. And as far as I can see Honda's materials commit the same sin-- where are the intermediate forms???
A real example of an intermediate form
If you see a demo that in every detail resembles a person crouched in a special suit... what does Occam's razor say you should conclude?
Why did they need to give it a humanoid torso and arms? Why do these also move exactly like the human versions? Why is the head concealed in a visor, if not to give their actor a way to look around? Why do the P3 quicktimes show only the cliched, anonymous box-with-blinking-lights inside, instead of showing off the real high-tech in the legs?
PROVE ME WRONG, PLEASE!
My page of robotics links if you're interested.
This would explain why it looks like a total hoax, with a crouching guy in there. (I still want to see proof that it's not.)
But if it's just 'playback' then it's not even AI-- it can't possibly adapt to changed terrain... right?
Um, will you at least put the "cellphone section" first, then... with an automatic cutoff that keeps the rest from flooding my bandwidth? ;^/
The theory here was supposed to be that you have exactly one section, that's just rendered differently for each device. As a theory that's very elegant, but as an interface in the real world it's turning out, over and over, to be a hollow dream.
McCluhan taught us in the sixties that the medium is the message-- you can't put the message in one file and the 'styling' in another. This problem with mobile devices and bandwidth is just the tip of the iceberg for CSS's fundamental wrongheadedness, imho.
The right longterm solution, I've been arguing forever, is to embed the document's styles in exactly the way WordPerfect used to do it-- with codes for BOLD and ITALIC, etc. Display devices can render those however they choose.
But in these earliest days of wireless, when every byte counts, the only way to go is optimised custom pages-- it shouldn't even be HTML, because the tags are too fat.
And when the bandwidth improves enough to handle HTML, it will work best with the same basic HTML that lynx likes. CSS just never solves any real problem along this evolutionary path!
I see this as just another top-down, ivory-tower W3C boondoggle, like XML. Those guys couldn't design a usable interface to save their lives-- look at how unreadable all their specs are! TimBL is the guy who thought up "aitch-tee-tee-pee-colon-slash-slash" for crying out loud...
The best games are ones that define a limited number of abstract 'dimensions' and then milk every possible drop of playability out of them.
When any game makes the jump to 'realistic' representation of 3-D space, it becomes impossible to do this, and the designer is frustrated by having too many degrees of freedom.
I started both with mysticism and computers as a teen in the 60s, and I actually walked away from my first programming job because I realized very vividly that it was going to distract me from the search for self-knowledge (which was very big in the early 70s).
I explored mystics like Gurdjieff and psychological systems like Fritz Perls' gestalt psych, and I put a lot of effort into meditation as a way of seeing what was really going on in my psyche/subconscious.
By the mid-70s this range of 'gurus' had narrowed to one who was especially austere and conceptually challenging-- J Krishnamurti. From him I came to see that soiritual freedom was about living without preconceptions or attachments, which is a common theme of a lot of mystical schools (especially Zen), but one that Krishnamurti handled with greater honesty, I thought.
At a certain point then I turned away from the whole mystical 'quest' and just started living, and at that point sociobiology became very inspiring, as did (non-sci-fi) literature.
I formulated a philosophy I call 'Robot Wisdom' that demands-- ultimately-- computer models of human behavior, that fit the requirements of sociobiology but also have the humane dimension of literature.
So my whole thrust has been to wrestle with how to model human emotions on a computer, eg in video games. And I look to great literature (especially James Joyce) for inspiration about the laws of human emotion.
I still feel a lot of sympathy for mystical points of view, because they deal with real psychological phenomena that materialist viewpoints seem to deny, but I don't practice any form of organised worship.
It's based on a book published by G Spencer Brown, the mathematical logician (Laws of Form).
Simson Garfinkel wrote an eye-opening backgrounder that explores this question. A quote:
TimBL, remember, is the guy who invented 'http://' and who dictated that two 'P's in a row should dispay the same as one (absurdly forcing everyone to add non-breaking spaces).It's just basic design common-sense that you don't create top-down 'standards' groups who dictate the rules of human factors without ever testing the standards, and without having the slightest understanding of what human-factors is all about.
I have an old rant about this.
A challenge for evil geniuses: what's the cheapest way to bring down the most satellites in one pop, assuming you're just going to use the threat for extortion, and don't care if you make satellite-space unusable forever?
You'd probably want something that spews spacedust in all directions over a range of altitudes...?
Nuh-uhn.
Even Finnegans Wake has grammatically well-formed sentences underlying its puns.
That's the problem, all right, but I don't see anyone in the AI world except Cycorp really wrestling with it, and they've mostly given up on human psychology, which is where all the useful web-semantics lie.
For example, suppose you want to find a Drew Barrymore fanpage that discusses how she stays so upbeat all the time-- how do you represent that with a finite set of XML labels?
Nobody knows, and solving it will take a sort of subtle psychological analysis that's the real, gaping abyss in current AI curricula.
I've been arguing since 1979 that the process of building really-intelligent software will require that we first come to deeply understand our own psyches-- robot wisdom. So limiting the research process in order to avoid 'evil robots' seems excessively alarmist.
I saw on a science show where one of the leading DNA experts collected DNA from lots of Egyptian mummies, to explore (among other things) how much inbreeding the royal line really did. That was several years back and I never heard any results so I think it must have been inconclusive.
You guys are slipping! My weblog had Reuters' story about this last Thursday. (It says next year.)
I've been collecting ideas under the name DecentOS
Briefly: HyperCard + emacs + Netscape + Frontier + ResEdit (Mac)
If people put their energy into writing better code instead, we'd be a lot farther along. Gizmo-porn is just a distraction, imho,
"I wonder if Rob and Co will now disclose their stock holdings in the companies they report on (like most respectable journalists)."
To me this seems like the critical question-- if a negative story about VA will cost Rob&c money, then we have to expect they'll spin it in the other direction.
These lines: "I almost quit when I heard about the merger... Since then I have been personally reassured by Larry Augustin..." also give me pause, because *persuasion* is what makes capitalism run, and there's no way to draw a line between a little 'safe' persuasion and a whole lot of utter-crap persuasion.
I never registered but for the last few days the NYT site has been letting me in anyway. I dunno if it's a bug or a quiet change of policy-- I hope the latter.
My version (related but not identical):
The only other explanation is that TW knows that they won't have to open their cable access to others, and can thus force AOL on all their customers (ie, the fix is in).
But I think all the metaphors about strangleholds and big players squeezing out little guys are naive propaganda-- the Web works in exactly the opposite way in the long run, and even a nebbish like Drudge can get a half-million hits a day with no advertising budget.
C'mon, tell it right...
This was 1993 or '94 he's talking about, and the group was alt.best.of.internet, where people were supposed to repost good news postings from other groups.
That group went thru regular longwave cycles where a few people would use it correctly, and it would draw an audience, so various idiots would try to take advantage of the attention, and people would go away... rinse, repeat.
AOL finally got around to offering newsgroups at that point, featuring a few 'starter' groups as everyone's default, the first of which-- in alphabetical order-- was aboi.
So AOL people started deluging into the group, asking totally random things like "Where can I get soy fertilizer in Ames, Iowa?" or "Anyone in Duluth want to make hot monkey love tonight?"
Now, the first response of aboi regulars was, disgracefully enough, to play sadistic games with the newcomers, making the situation much worse. (Shooting fish in a barrel was the image that came to my mind.)
It was actually me who realized what the situation was at AOL's end, and started trying to get them a message to change the default for the sake of aboi.
They didn't respond until Weemba posted a message with the subjectline "AOL top brass joins Satanic cult in ritual sacrifices" or the equialent, within hours of which the spigot went off.
I was just thinking this morning how I've never enjoyed programming on anything else the way I enjoyed 6502 on the Apple ][... but I *assume* that someday someone will write a GUI that's equally elegant, and gives even beginning hackers the same sense of transparency. Or is this a pipe dream?
My philosophy is that you shouldn't ever link to the magazine's homepage (it will be linked from the article) because it's just another opportunity for confusion. And the highlighted text should be unambiguous-- go with "editorial" so I know where I'm going when I click.
(For all I knew, the 'totally mainstream' link might have been to a Corel download page or something.)
Apropos of nothing, Vladimir Nabokov used the image of a red top hat to imply someone beheaded by a guillotine.
"Science News" ought to do them fine, and is a lot easier to read.
"The game designers at Wisdom Tree bought permission from id to remake and slightly alter Castle Wolfenstein 3D to design a game called Super 3D Noah's Ark. In the Wisdom Tree game, players assume the role of Noah trying to get his animals back in their pens, as opposed to a soldier trying to hunt down Adolph Hitler. They also did away with the grenades and machine guns and replaced them with a single slingshot that players used to shoot food at animals. Instead of having soldiers that charged at you, then fell in a bloody pool when shot, Super 3D Noah's Ark has little goats that try to kick you with their hind legs and fall asleep when you shoot them with food."
They would have had to start in 1995, or switch over their whole database at some more recent date.