The idea is that before you can accurately identify (eg) boners in an image, you must also be able to deduce the lightsources and body position and skincolor in that image... and the first (comparatively easy) step in doing this is to gather algorithms that can generate a visual representation of every possible combination of these.
The target image can then be matched to multiple hypothetical models, and the best match selected.
A simplified approach might look just for the cylindrical boner, while deducing the lightsource.
AI startingpoint: 3D model of boner plus stomach plus thighs, from every angle, at every distance, with every lighting, and every skincolor.
Once you've got that working so it can reconstruct, from just the visual output, every permutation it can generate, then you can start addressing the exceptions like clothing, hands, softies...
In explaining the chain of reasoning, they weirdly left out that the name 'Pagliacci' is explicit in the lyrics, and proposed that Watson would deduce 'tears' as a form of feelings!? (Maybe they don't want to include a database of song lyrics?)
They claim they won't use Web data, but there's no way they can compile enough databases on their own to handle Jeopardy's general knowledge. Awards, lyrics, plots, characters... the list goes on and on and on.
WolframAlpha is a recent disappointment that's spent years collecting databases and delivers almost nothing useful yet.
I'd suggest celebrity blind-items as a fun test domain that might be manageable, eg:
"Which female rocker, who was once married to a famous old-school singer, now has a penchant for young girls?" [cite]
Even textbooks on this topic don't usually spell out the very simple dependence between atmospheric depth and surface temperature: when you warm the Earth, air molecules 'bounce' higher, so the atmosphere gets deeper. When you cool it, they bounce less high. The higher they fly, the slower they move, unintuitively termed 'adiabatic cooling'.
A small percentage of the highest bouncers can be reheated by the Sun near the top of their bounces, and I assume the reported lower ionosphere is more due to a decline in this factor than to any global cooling.
In Chicago in 1981 I found it very easy to get hired to copy arcade games for the Apple, Atari, and C64 (all 6502). Roklan and Image Producers hired me to do Berzerk and Wizard of Wor (one was supposedly for Microsoft). There was no training or local expertise available, you just had to reverse engineer them. Then Atari(?) successfully sued somebody for a PacMan ripoff, and the whole bubble quickly burst...
The Apple ][ was infamous for the bizarre layout of the graphics memory (supposedly Woz chose it to save a chip, or maybe a layer on the circuit board). And if the high bit was set, all the pixels in that byte shifted, creating the other two available colors.
I found a hidden 'Hot Coffee' style easter egg in the text strings for Sierra's 'Wizard and the Princess'-- the placeholder text for the default/generic "I don't know how to **** something" reply was the f-word (never displayed)...
I was a 16yo Antioch College freshman psych major, the department visited a fridge-sized PDP off-campus and we each got to try Lunar Lander. I crashed, but I still say they had the equation for gravity wrong-- it accelerated way too fast.
I couldn't include an event record for each individual event
I'm not convinced you've really tried-- suppose the METAs described an event1, an event2, etc, and whatever Firefox extension is tasked with extracting that info could look for flags in the body that show which event is described where?
Microformats seem to be a classic 'bag taped to the side', because the logic of the semantic web was still poorly visualised when XML was selected. I'm just asking whether META doesn't deserve rehabilitation as a 'bag at the top' instead...
How much of microformats could have been done using META... None of it.
I just don't believe that. If you're describing one or more events, why can't you put most or all of those descriptions in META format?
For me, the worstcase for the waste in rejecting META is that we could have been putting Yahoo/Dmoz categories there, all this time, but haven't been because the 'cult' didn't think it was fancy enough.
Strangely enough, I just asked a question: "How much of this could have been done 5 years ago if the structured-HTML community hadn't blindly rejected META headers?"
So if anyone is on a hobbyhorse, it's Mr Coward.
It's scoped to the page not particular parts of it, and it has a plain-text content model because it uses attributes instead of child elements.
So I guess I have to ask again: How much of microformats could have been done using META, given that it's scoped to the page (which is no problem for the most important page semantics), and uses attributes?
"Examples of philosophical work relevant to AI (besides mathematical logic) include the work of Frege (sense and denotation), Gödel (modern mathematical Platonism), Tarski (theory of truth), Quine (ontology and bound variables), Putnam (natural kinds), Hintikka (formalization of facts about knowledge), Montague (paradoxes of intensionality), Kripke (semantics of modality), Gettier (examples on intensionality), Grice (conversational implicatures), and Searle (performatives)." John McCarthy
I agree. My own (universally ignored) proposal for the taxonomy problem starts with person, place, and thing as 'elements' and builds complex ideas as compounds of these: [faq]
Are they claiming that screen is the production version, or just a placeholder? Because last I heard the (specially lowcost) screen was still being developed...
Of all the random phrases the Slashdot editors might have allowed as the primary anchortext for this article, "Mother of the Internet" is about the least valid.
It's mindboggling to me how slowly this supposedly tech-savvy community is coming to terms with formulating a consistent, user-friendly policy about anchortext.
If you use the name of the magazine, it implies you're going to the homepage of the magazine, not to the article itself, so don't do that. (And we don't need a link to that homepage at all-- it just confuses things.)
My recommendation is to use a word like "article" or "interview" which describes the type of file you'll get when you click the link. This is consistent and predictable.
There used to be a kneejerk dogma that the anchortext should stand alone as a description of the content, with the justification that link-extractors could rely on this... but nobody uses automatic link extractors anyway, so I think this theory has failed.
"People laugh when I say that I think a jellyfish is one of the most beautiful things in the world. What they don't understand is, I mean a jellyfish with long, blonde hair."
My nominee was a pseudo-trackball for the Atari 2600, made by Roklan around 1982. Trackballs had been hot for arcade games since Missile Command, but were too expensive for home games. Roklan's 'ball' didn't really roll, though it was designed to look like it did. Instead, it gave the standard 2600 joystick compass-points, via the infinitely non-ergonomic semispherical controller.
"...Who does not know that a horse falling from a height of three or four cubits will break his bones, while a dog falling from the same height or a cat from a height of eight or ten cubits will suffer no injury? Equally harmless would be the fall of a grasshopper from a tower or the fall of an ant from the distance of the moon..."
They skim over the fact that when KC Munchkin lost to PacMan, there were dozens of other copycat games that were suddenly too risky to market, which contributed to the industrywide 'cold spell'.
Why don't people notice that this is all transparently implausible?
They'll have a $100 laptop if they can solve the software, the hardware, the screen, the power, the engineering, and the manufacturing... and it'll be $100 if they can solve all those issues for under $100.
Slashdot Mirror
My local station ran round one at 11am, and I've posted spoilers here: http://twitter.com/robotwisdom
Fifteen years on, I'd have thought my AI complementarity principle would be widely understood...
The idea is that before you can accurately identify (eg) boners in an image, you must also be able to deduce the lightsources and body position and skincolor in that image... and the first (comparatively easy) step in doing this is to gather algorithms that can generate a visual representation of every possible combination of these.
The target image can then be matched to multiple hypothetical models, and the best match selected.
A simplified approach might look just for the cylindrical boner, while deducing the lightsource.
AI startingpoint: 3D model of boner plus stomach plus thighs, from every angle, at every distance, with every lighting, and every skincolor.
Once you've got that working so it can reconstruct, from just the visual output, every permutation it can generate, then you can start addressing the exceptions like clothing, hands, softies...
They claim they won't use Web data, but there's no way they can compile enough databases on their own to handle Jeopardy's general knowledge. Awards, lyrics, plots, characters... the list goes on and on and on.
WolframAlpha is a recent disappointment that's spent years collecting databases and delivers almost nothing useful yet.
I'd suggest celebrity blind-items as a fun test domain that might be manageable, eg:
Even textbooks on this topic don't usually spell out the very simple dependence between atmospheric depth and surface temperature: when you warm the Earth, air molecules 'bounce' higher, so the atmosphere gets deeper. When you cool it, they bounce less high. The higher they fly, the slower they move, unintuitively termed 'adiabatic cooling'.
A small percentage of the highest bouncers can be reheated by the Sun near the top of their bounces, and I assume the reported lower ionosphere is more due to a decline in this factor than to any global cooling.
In Chicago in 1981 I found it very easy to get hired to copy arcade games for the Apple, Atari, and C64 (all 6502). Roklan and Image Producers hired me to do Berzerk and Wizard of Wor (one was supposedly for Microsoft). There was no training or local expertise available, you just had to reverse engineer them. Then Atari(?) successfully sued somebody for a PacMan ripoff, and the whole bubble quickly burst...
The Apple ][ was infamous for the bizarre layout of the graphics memory (supposedly Woz chose it to save a chip, or maybe a layer on the circuit board). And if the high bit was set, all the pixels in that byte shifted, creating the other two available colors.
I found a hidden 'Hot Coffee' style easter egg in the text strings for Sierra's 'Wizard and the Princess'-- the placeholder text for the default/generic "I don't know how to **** something" reply was the f-word (never displayed)...
I was a 16yo Antioch College freshman psych major, the department visited a fridge-sized PDP off-campus and we each got to try Lunar Lander. I crashed, but I still say they had the equation for gravity wrong-- it accelerated way too fast.
1) No one has seen the magical lowcost screen yet
2) No one knows how the mesh network will work
3) No one knows how hard the batteries will be to keep charged
4) No one knows how usable the software apps will be
5) Nobody gets a refund if any of these wishful thoughts fall through
The date on the Nasa page is 18 April.
Is that
crap really the best they've got? It makes my eyes bleed...(Since the association between the human- and machine- readable texts is wholly imaginary, why not keep the machine vesion in META?)
I'm not convinced you've really tried-- suppose the METAs described an event1, an event2, etc, and whatever Firefox extension is tasked with extracting that info could look for flags in the body that show which event is described where?
Microformats seem to be a classic 'bag taped to the side', because the logic of the semantic web was still poorly visualised when XML was selected. I'm just asking whether META doesn't deserve rehabilitation as a 'bag at the top' instead...
I just don't believe that. If you're describing one or more events, why can't you put most or all of those descriptions in META format?
For me, the worstcase for the waste in rejecting META is that we could have been putting Yahoo/Dmoz categories there, all this time, but haven't been because the 'cult' didn't think it was fancy enough.
Strangely enough, I just asked a question: "How much of this could have been done 5 years ago if the structured-HTML community hadn't blindly rejected META headers?"
So if anyone is on a hobbyhorse, it's Mr Coward.
It's scoped to the page not particular parts of it, and it has a plain-text content model because it uses attributes instead of child elements.
So I guess I have to ask again: How much of microformats could have been done using META, given that it's scoped to the page (which is no problem for the most important page semantics), and uses attributes?
How much of this could have been done 5 years ago if the structured-HTML community hadn't blindly rejected META headers?
Except for waterclocks: timeline
"Examples of philosophical work relevant to AI (besides mathematical logic) include the work of Frege (sense and denotation), Gödel (modern mathematical Platonism), Tarski (theory of truth), Quine (ontology and bound variables), Putnam (natural kinds), Hintikka (formalization of facts about knowledge), Montague (paradoxes of intensionality), Kripke (semantics of modality), Gettier (examples on intensionality), Grice (conversational implicatures), and Searle (performatives)." John McCarthy
I agree. My own (universally ignored) proposal for the taxonomy problem starts with person, place, and thing as 'elements' and builds complex ideas as compounds of these: [faq]
Are they claiming that screen is the production version, or just a placeholder? Because last I heard the (specially lowcost) screen was still being developed...
Of all the random phrases the Slashdot editors might have allowed as the primary anchortext for this article, "Mother of the Internet" is about the least valid.
It's mindboggling to me how slowly this supposedly tech-savvy community is coming to terms with formulating a consistent, user-friendly policy about anchortext.
If you use the name of the magazine, it implies you're going to the homepage of the magazine, not to the article itself, so don't do that. (And we don't need a link to that homepage at all-- it just confuses things.)
My recommendation is to use a word like "article" or "interview" which describes the type of file you'll get when you click the link. This is consistent and predictable.
There used to be a kneejerk dogma that the anchortext should stand alone as a description of the content, with the justification that link-extractors could rely on this... but nobody uses automatic link extractors anyway, so I think this theory has failed.
"People laugh when I say that I think a jellyfish is one of the most beautiful things in the world. What they don't understand is, I mean a jellyfish with long, blonde hair."
My nominee was a pseudo-trackball for the Atari 2600, made by Roklan around 1982. Trackballs had been hot for arcade games since Missile Command, but were too expensive for home games. Roklan's 'ball' didn't really roll, though it was designed to look like it did. Instead, it gave the standard 2600 joystick compass-points, via the infinitely non-ergonomic semispherical controller.
Wouldn't it be quicker to try it on the Martian meteorites that purportedly showed fossils of cells?
Galileo Galilei 1638: http://galileoandeinstein.physics.virginia.edu/tns 1.htm
They skim over the fact that when KC Munchkin lost to PacMan, there were dozens of other copycat games that were suddenly too risky to market, which contributed to the industrywide 'cold spell'.
Why don't people notice that this is all transparently implausible?
They'll have a $100 laptop if they can solve the software, the hardware, the screen, the power, the engineering, and the manufacturing... and it'll be $100 if they can solve all those issues for under $100.
Suckers!