yeah... this is just trying to replicate how the human eye cells work. To get to OSC's device you'd still need an I/O interface, which you'll note is conspicuously lacking. (and as i recall Ohaldo's left eye was the i/o socket or some such? it's been a while since i read the books)
What google is doing is not Monopolizing. They're doing vertical integration. They're not the dominating force in any of their fields excluding search technology, and their position their is a precarious one.
They are not, nor likely will they be a monopoly.
On that note, size does not indicate malevolence. Companies that continue to have drives to do things (as opposed to exclusively being interested in sucking money out of their customers) are not the sort of threat that people keep making them out to be.
It's not the size that matters, it's the corporate culture.
Anybody ever tried the humane interface?
From the conversations i've had with my friend on the dev team, the concept is that all functionality should be interoperable and sort of generalized i guess (perhaps somebody can give a better description). The root of the project as i recall came from the complaint that there is a lot of functional redundancy within a single program (spreadsheets that have calculators and word processors, wordprocessors that have spreadsheet functionality in them, etc.), and that they all functioned differently. So if one could flatten all of this & make a bunch of the functionality interoperable, it would make the learning curve for using these programs much more comfortable.
But, i haven't tried it out yet.
You'd think somebody trying to commit corporate espionage would be smarter wouldn't you? I mean, communication is a two party interaction, you'd think the coporation on the receiving end would provide a slightly more secure method for communication, if they were looking for secrecy:p
Dude, just because the first two movies sucked horribly, doesn't mean that people aren't hoping that the third one doesn't suck. You've answered your own question, people don't genuinely dislike them (as can be seen by the comments and slashdotting), they're in fact more incensed because they feel let down. If you're not defending lucas, and you're interested (as everyone else is) in how epIII will tie into epIV, then why are you so hung up about people riding down the first pair of films?
Oh well, this is Slashdot. Continue your bitching.
Most of the reasons they give for email being shunned for other mediums are pretty flimsy.
I know people who set up their AIM client so that you can't tell whether they're idle, and only respond to messages 10 hours after you've sent them, and i know people who watch their inboxes like obsessive hawks.
as for email being less "fun" than aim... I don't know, i think my gmail account is pretty cool... and conversational for that matter.
This debate is pretty silly, after all, all we're talking about is persistant electronic messaging. In terms of user experience, email and a client like ICQ aren't -drastically- different. Presumably email will get faster and friendlier, and hell, at some point probably may as well be the same as an IM system.
Microsoft is attempting to patent all intelligent systems. Their patent would essentially given them rights to any system that has a robust linguistic system and some sort of task oriented backend that explains things to people.
As a computational linguist who's interesting in making peoples lives better (through things that would fall under this patent), i think this would be a really horrible stifling idea. As a result i'm glad that the patent is so stupid, because i'm sure i could come up with counter examples.
i'll have to read more on the method, but i'm skeptical of most art classification systems. I'm also curious what exactly it is that they're learning from the peices of artwork they're generalizing over (yes yes, i'll have to RTFP), hopefully they're not doing what was done with early artificial neural networks that is, simply letting the algorithm decide what unconstrained features it found common across all the paintings.
or in other words, sounds like it's not too shabby with recall. so what's its precision?
Using a dialect to make a joke isn't a racial slur. It's an appeal to stereotype. And while linguistic stereotypes are often tied in the lay public to other sorts of stereotypes, such links are not an inherent part of language, and imo (and in the opinion of other linguists) linguistic stereotypes shouldn't be confused with racial or discriminatory stereotyping. Language use doesn't confine itself to racial borders.
And besides, you could just as easily say he's just making fun of any number of groups who play up talking like that, like rappers (some of whom you'll note are white).
Nuclear power always has to be near a water source, and always raises the water temp of the body of water that it sits on, usually by a number of degrees if i recall correctly.
This is a dramatic impact on the local habitat. No power source has no environmental impact.
Ah, so it's more of that FOAF stuff you've been talking about on grenme. I posted my comment in duplicate there, why don't we continue the conversation back at grenme?
Now, here's what all the semantic web seems to boil down to, to me:
1) Common interface and data structures for describing all objects (how ever that'll be done, i think that's a pipe dream in itself).
2) methods of combining and cross-referencing information from the aformentioned objects, whether they be user-defined or otherwise.
3) Making sure everyone complies with enough regularity that the system actually does something useful.
Now, i want to talk about a related and similarly interesting problem that i am personally more familiar with (in fact it is perfectly analogous).
I'm a computational linguist (computational psycho-linguist to be precise). Several of my professors are interested in corpus based research, which requires in some cases annotated corpora. The interesting thing about natural language corpora, is that the goal of annotation is exactly what the semantic web people want, metadata that coherently describes all the attributes of the object they're discussing.
The problem is that there isn't consensus as to what attributes human language actually has, so there are corpora which were annotated with metadata that's similar to syntactic theory A and corpora which have tags which favor theory B. So in the past couple years there have been pushes for "theory neutral annotation".
Here's the problem. If it were possible to generate an annotation scheme that could be used by adherents of theory A and adherents of theory B (that is a theory neutral annotation scheme existed), it would mean that fundamentally theory A and theory B were, in fact, compatable. Since the entire reason that the two camps have different annotation schemes is because their theories are incompatible, it seems like a futile or contradictory endevor to undertake theory neutral metadata.
The problem is the same for any system which wishes to create universal object descriptors. There is no way to do pre-theoretic object description. There are of course qualities of objects which will be uncontraversial, but ultimately some people are going to want to pay attention to certain qualities of objects, and other people aren't going to want the same attributes. While you may say "well that's fine, lets just toss in all the attributes," you'll still have the problem that some people will want to divide up the qualities that an object has one way, and someone else will want to divide up the attributes of the same object in another fundamentally incompatable manner.
So what do you do then? You're fucked. You can't enforce consistency, and so your object descriptors are dead, which means you don't have your uniform api, and ontologies won't work.
alright, having read the friggin' article, all i have to say is that they have their work cut out for them.
the problem with searching currently is that only librarians, who've had at least a year or two of graduate studies really know the ontology that libraries use. Common users bring their own concepts and ontologies to bear when they're searching for information. But if you move away from the monolithic single ontologies that libraries use, you have the problem that you have to be open to the fact that ontologies change, not just between individuals but over time, as cultures change the ontologies need to change as well. I guess the concept must be that there are a set of descriptors which are invariant, and can thus be interpreted based on the features of those objects by different ontologies.
The crazy part about trying something like that is that you have to make people define their own ontologies. Furthermore you have the problem that you need to make sure that people are describing their data in an ontologically neutral manner.
And that's the hidden third problem (the technology review article posted above has the dude citing 2 problems), getting people to behave in a sensible way when dealing with information organization. Unfortunately in so far as we know now, it's really difficult to get computers to automatically create meta-data (that doesn't mean we're not trying), but primarily humans have to be included in the decision process if you want to define what things are.
the ironic thought that pops to mind is that if you've got a set of universal descriptors, then don't you already have an ontology? And if you don't have a set of universal descriptors, how would you ever create a coherent ontology?
I'll have to rtfa to see what they propose, but just the principle of resolving contradictions is a really difficult one, and most theories of knowledge (which are essentially networks of facts) aren't terribly robust, and contradiction repair, which involves running the entire network to find invalid assumptions, and then propigating the changes is NP complete:| i'm not positive that contradiction resolution is a reasonable thing to expect out of a massive distributed network.
FYI: Mach speed is relative to air pressure... so depending on the region of orbit you're referring to, those numbers could mean vastly different things.
5. MACH
Mach is theratio of the aircraft's true airspeed to the speed of sound. For example, a speed of Mach 0.7 means the aircraft is flying at 70% of the speed of sound.
The speed of sound is variable. Depending on air temperature, relative humidity and pressure, it is approximately 1 117 feet or 340.48 meters per second at 59 degrees Fahrenheit.
By the way, Concorde crosses the Atlantic ocean with Mach 2.0 at 55 000 feet altitude.
uh, no?:p
Hitler's plan was supposed to be harmonious with catholicism.
Hitler heavily used mythologies and other superstitions to drum up the belief that the jewish and others were inherently inferior, and the germans were the aryan super-race.
Nietzsche on the other hand... (who people credit the concept of the ubermensch) his whole point was stepping "beyond good and evil." Which was a ridiculous endevour, because it a) fails to actually escape moralism and b) enabled psychopaths like the nazis.
Well... actually, if Microsoft were to be able to sucessfully do this, it would actually be news. Image recognition and object segmentation is and has been one of the biggest stumbling blocks of artificial intellegence research for... well, ever.
my bet is that they're integrating some neural network based search systems for this, and that, like current neural network systems, they won't function very well in a generalized task set.
as someone who's interested in computational neurodynamics, i've got ambivielent feelings about alot of the stuff minsky says, but i can't help but sympathize with him to some degree on this subject. Robots are cool, robots are useful, but the stuff that comes out of MIT's AI lab, has lost focus on the original goal of what he and the pioneers of the area were after.
Heading off in a different direction isn't bad, in fact, i think a lot of these devices are quite interesting, but take just one of these two paths to the neglect of the other seems kind of sucky. So, i disagree that with minsky that this stuff is useless, but i agree with him that it's a shame that the people working (loosely) on the computational problems about the mind/brain aren't in MIT engineering anymore.
Slashdot Mirror
yeah... this is just trying to replicate how the human eye cells work. To get to OSC's device you'd still need an I/O interface, which you'll note is conspicuously lacking. (and as i recall Ohaldo's left eye was the i/o socket or some such? it's been a while since i read the books)
Who is letting kids install stuff on school gear?
What google is doing is not Monopolizing. They're doing vertical integration. They're not the dominating force in any of their fields excluding search technology, and their position their is a precarious one.
They are not, nor likely will they be a monopoly.
On that note, size does not indicate malevolence. Companies that continue to have drives to do things (as opposed to exclusively being interested in sucking money out of their customers) are not the sort of threat that people keep making them out to be.
It's not the size that matters, it's the corporate culture.
Anybody ever tried the humane interface? From the conversations i've had with my friend on the dev team, the concept is that all functionality should be interoperable and sort of generalized i guess (perhaps somebody can give a better description). The root of the project as i recall came from the complaint that there is a lot of functional redundancy within a single program (spreadsheets that have calculators and word processors, wordprocessors that have spreadsheet functionality in them, etc.), and that they all functioned differently. So if one could flatten all of this & make a bunch of the functionality interoperable, it would make the learning curve for using these programs much more comfortable. But, i haven't tried it out yet.
Congrads to my friend Atul who works there ;)
You'd think somebody trying to commit corporate espionage would be smarter wouldn't you? I mean, communication is a two party interaction, you'd think the coporation on the receiving end would provide a slightly more secure method for communication, if they were looking for secrecy :p
Dude, just because the first two movies sucked horribly, doesn't mean that people aren't hoping that the third one doesn't suck. You've answered your own question, people don't genuinely dislike them (as can be seen by the comments and slashdotting), they're in fact more incensed because they feel let down. If you're not defending lucas, and you're interested (as everyone else is) in how epIII will tie into epIV, then why are you so hung up about people riding down the first pair of films?
Oh well, this is Slashdot. Continue your bitching.
Most of the reasons they give for email being shunned for other mediums are pretty flimsy.
I know people who set up their AIM client so that you can't tell whether they're idle, and only respond to messages 10 hours after you've sent them, and i know people who watch their inboxes like obsessive hawks.
as for email being less "fun" than aim... I don't know, i think my gmail account is pretty cool... and conversational for that matter.
This debate is pretty silly, after all, all we're talking about is persistant electronic messaging. In terms of user experience, email and a client like ICQ aren't -drastically- different. Presumably email will get faster and friendlier, and hell, at some point probably may as well be the same as an IM system.
9:55 EST, and it's fixed
how about... triplicate? :P
Taco appears to be "Trip"-ping.
You are very very wrong.
Microsoft is attempting to patent all intelligent systems. Their patent would essentially given them rights to any system that has a robust linguistic system and some sort of task oriented backend that explains things to people.
As a computational linguist who's interesting in making peoples lives better (through things that would fall under this patent), i think this would be a really horrible stifling idea. As a result i'm glad that the patent is so stupid, because i'm sure i could come up with counter examples.
i'll have to read more on the method, but i'm skeptical of most art classification systems. I'm also curious what exactly it is that they're learning from the peices of artwork they're generalizing over (yes yes, i'll have to RTFP), hopefully they're not doing what was done with early artificial neural networks that is, simply letting the algorithm decide what unconstrained features it found common across all the paintings.
or in other words, sounds like it's not too shabby with recall. so what's its precision?
Using a dialect to make a joke isn't a racial slur. It's an appeal to stereotype. And while linguistic stereotypes are often tied in the lay public to other sorts of stereotypes, such links are not an inherent part of language, and imo (and in the opinion of other linguists) linguistic stereotypes shouldn't be confused with racial or discriminatory stereotyping. Language use doesn't confine itself to racial borders.
And besides, you could just as easily say he's just making fun of any number of groups who play up talking like that, like rappers (some of whom you'll note are white).
Nuclear power always has to be near a water source, and always raises the water temp of the body of water that it sits on, usually by a number of degrees if i recall correctly.
This is a dramatic impact on the local habitat. No power source has no environmental impact.
Ah, so it's more of that FOAF stuff you've been talking about on grenme. I posted my comment in duplicate there, why don't we continue the conversation back at grenme?
Now, here's what all the semantic web seems to boil down to, to me:
1) Common interface and data structures for describing all objects (how ever that'll be done, i think that's a pipe dream in itself).
2) methods of combining and cross-referencing information from the aformentioned objects, whether they be user-defined or otherwise.
3) Making sure everyone complies with enough regularity that the system actually does something useful.
Now, i want to talk about a related and similarly interesting problem that i am personally more familiar with (in fact it is perfectly analogous).
I'm a computational linguist (computational psycho-linguist to be precise). Several of my professors are interested in corpus based research, which requires in some cases annotated corpora. The interesting thing about natural language corpora, is that the goal of annotation is exactly what the semantic web people want, metadata that coherently describes all the attributes of the object they're discussing.
The problem is that there isn't consensus as to what attributes human language actually has, so there are corpora which were annotated with metadata that's similar to syntactic theory A and corpora which have tags which favor theory B. So in the past couple years there have been pushes for "theory neutral annotation".
Here's the problem. If it were possible to generate an annotation scheme that could be used by adherents of theory A and adherents of theory B (that is a theory neutral annotation scheme existed), it would mean that fundamentally theory A and theory B were, in fact, compatable. Since the entire reason that the two camps have different annotation schemes is because their theories are incompatible, it seems like a futile or contradictory endevor to undertake theory neutral metadata.
The problem is the same for any system which wishes to create universal object descriptors. There is no way to do pre-theoretic object description. There are of course qualities of objects which will be uncontraversial, but ultimately some people are going to want to pay attention to certain qualities of objects, and other people aren't going to want the same attributes. While you may say "well that's fine, lets just toss in all the attributes," you'll still have the problem that some people will want to divide up the qualities that an object has one way, and someone else will want to divide up the attributes of the same object in another fundamentally incompatable manner.
So what do you do then? You're fucked. You can't enforce consistency, and so your object descriptors are dead, which means you don't have your uniform api, and ontologies won't work.
the principle of resolving contradictions *AUTOMATICALLY is a really difficult one
>:P
alright, having read the friggin' article, all i have to say is that they have their work cut out for them.
the problem with searching currently is that only librarians, who've had at least a year or two of graduate studies really know the ontology that libraries use. Common users bring their own concepts and ontologies to bear when they're searching for information. But if you move away from the monolithic single ontologies that libraries use, you have the problem that you have to be open to the fact that ontologies change, not just between individuals but over time, as cultures change the ontologies need to change as well. I guess the concept must be that there are a set of descriptors which are invariant, and can thus be interpreted based on the features of those objects by different ontologies.
The crazy part about trying something like that is that you have to make people define their own ontologies. Furthermore you have the problem that you need to make sure that people are describing their data in an ontologically neutral manner.
And that's the hidden third problem (the technology review article posted above has the dude citing 2 problems), getting people to behave in a sensible way when dealing with information organization. Unfortunately in so far as we know now, it's really difficult to get computers to automatically create meta-data (that doesn't mean we're not trying), but primarily humans have to be included in the decision process if you want to define what things are.
the ironic thought that pops to mind is that if you've got a set of universal descriptors, then don't you already have an ontology? And if you don't have a set of universal descriptors, how would you ever create a coherent ontology?
anyway, enough rambling for now
-notheory
I'll have to rtfa to see what they propose, but just the principle of resolving contradictions is a really difficult one, and most theories of knowledge (which are essentially networks of facts) aren't terribly robust, and contradiction repair, which involves running the entire network to find invalid assumptions, and then propigating the changes is NP complete :| i'm not positive that contradiction resolution is a reasonable thing to expect out of a massive distributed network.
quoted from: http://www.gg-pilot.com/ggweeklyarticle.asp?id=5
How about more NYT articles, and articles from other news outlets. How about endorsments from political figures?
i don't think it'd be that hard to convince people that this is a good idea.
the problem is attempting to drag those who have a vested interest in keeping the vote closed along.
All the people who bitch about closed voting schemes should email this article to their Senators and Representatives.
uh, no? :p
Hitler's plan was supposed to be harmonious with catholicism.
Hitler heavily used mythologies and other superstitions to drum up the belief that the jewish and others were inherently inferior, and the germans were the aryan super-race.
Nietzsche on the other hand... (who people credit the concept of the ubermensch) his whole point was stepping "beyond good and evil." Which was a ridiculous endevour, because it a) fails to actually escape moralism and b) enabled psychopaths like the nazis.
Well... actually, if Microsoft were to be able to sucessfully do this, it would actually be news. Image recognition and object segmentation is and has been one of the biggest stumbling blocks of artificial intellegence research for ... well, ever.
my bet is that they're integrating some neural network based search systems for this, and that, like current neural network systems, they won't function very well in a generalized task set.
This is a reply to the above two posts
as someone who's interested in computational neurodynamics, i've got ambivielent feelings about alot of the stuff minsky says, but i can't help but sympathize with him to some degree on this subject. Robots are cool, robots are useful, but the stuff that comes out of MIT's AI lab, has lost focus on the original goal of what he and the pioneers of the area were after.
Heading off in a different direction isn't bad, in fact, i think a lot of these devices are quite interesting, but take just one of these two paths to the neglect of the other seems kind of sucky. So, i disagree that with minsky that this stuff is useless, but i agree with him that it's a shame that the people working (loosely) on the computational problems about the mind/brain aren't in MIT engineering anymore.