While the idea would probably do some good if widely adopted what's really needed is to reduce the need for text based indexing of web sites but increasing the amount of explict semantic information about its content.
Marking up pages with information about the meaning of the terms on them is the main thrust of the work on semantic web - see http://www.daml.org/ (for DAML - the DARPA Agent Markup Language), http://www.semanticweb.org/ (One of the main information sources) and finally the new W3C activity on the subject: http://www.w3.org/2001/sw/.
How far, how fast it will go is another matter but there's certainly a lot of interest in creating a more "machine readable" web.
... has been around since the mid 80's and although it works for toy problems it's very hard to get systems of a significant scale out of it. You're basically swapping sub branches of your program around to see what works - tranversing the space of all possible programs - it takes *a lot* of random attempts to do better than a human doing it analytically. Most AI researchers believe that you need at least a little bit of knowledge to guide your program's adaptation rather than blind mutation.
The Father of GP (John Koza) may disagree with me - he runs genetic-programming.org and more or less invented the field. He's also known for his vigorous defences of GP: anybody know of real applications?
This has been touched on above but not explicitly. Mcrosoft are selling two things:
(Minor): a standard interface to an on-line address book,.. whatever.
(Major) access to their user's data (with the use's permission).
They are playing a chicken and egg game. Most users will want to store their personal data in one place (their machine or some server) and not enter this in 10,000 times. No how does my application X access Johnny's data when he logs onto my site? It's authenticates the passport and get's its from the microsoft server.
So you get access to a "service" which allows you to "serve" anybody who uses a passport. If passport becomes the worldwide standard you'd defn be getting your money's worth at $250 a pop. Let's bet that the cost won't stay there, Wait for it 1) to increase, 2) to become "per transaction" based. That's where the money really starts flowing.
However I believe that the principle behind.NET is a no-brainer good idea - people need a service like this. Clealry however it should be based top to bottom on an open standard so that anybody is able to host personal passport data (e.g. W3c P3P.
On a realistic note thought I think open ways of doing this will emerge - because many people will want a slice of this pie and people won't be held to ransom. Microsoft will obtain a big first mover advantage but also carries a lot of cost. What is critical is to check out the patent situation in the area (because you can bet your life there will be a large arsenal of them).
Roll on the OpenSource standard passport + third party providers to host that data.
No use it for sure;-). Just trying to incite the US contingent to fight against the laws that will make it illegal to export and use the technology!
Re:Get in the habit of using Crypto now...
on
FBI Wants to Tap The Net
·
· Score: 2, Interesting
...they won't crack them. They'll store them and if they think you're a suspect kindly ask you to stop using encrypted messaging followed by a supena (polite request) for your encryption keys.
Mass decryption just isn't feasible and certainly not in real time so they have to try to do one of two things:
Prevent anybody from using strong encryption ("All e-commerce stopped for evermore today when the US senate passed the...").
Find a way to force everybody to hand them their keys.
I wouldn't be so sure of your "definition" - many would argue that the is no such thing as intelligence there is only perceived intelligence. Examples:
Intelligence can depend on the environment: Is a spider intelligent? Spining a web it's amazing, stick the thing in a bath tub and it doesn't look so smart.
Intelligence can be social: is an ant intelligent? Not by itself but ant colonies perform some pretty amazing feats.
Intelligence may depend on other knowledge: A chess grandmaster may play a very strange move near the beginning of the game which looses him the game. Why? He took a calculated risk and it didn't pay off. Was he dumb? No, you say. What if it wasn't a chess grandmaster but Joe Blogss from down the street - yeah THAT was a dumb move...
Perception of intelligence is about being seen to to the right thing at the right time.
Regarding the second point - this gets to the heart of the Chinese Room Argument: can intelligence (I would distinguish "sentience") be "built" or "must" there be something more. Was deep blue intelligent? Searle would argue "no". Some would argue "Yes, In the chess domain". There was nobody on the planet it couldn't teach somethign about chess and (to an extent) explain those choices. Many AI researchers weren't happy about deep blue because it basically used very fast search and no fancy reasoning. But hey - that just shows that there's more than one way to solve a problem IMHO...
What I like most about this proposal is that we're finally going to get the definitive, use everywhere, failsafe, uncrackable security system for copy protection. Once every implements that we'll live happily and crime free for ever more.
Sorry but last time I checked this wasn't going to happen in our universe! But hey maybe it'll be based on DRM2 (slashdot).
Your real problem is that you want it both ways. You want your speech to be protected when you use email, but you also want to restrict speech for email you find disapproving. The sender won't know whether or not his email meets your criteria unless he sends you the email for your approval.
This is not what I said. I simply stated that I don't feel people have the right to send me email. Just like they don't have the right to camp outside my house and shout abuse at me every day.
Can email be a public forum? Yes it most certainly can -- in the form of mailing lists. So a spammer uses a mailing list, and it
becomes completely legal. Why? He's communicating to a large number of people using a publicly financed network. If that
doesn't sound like a public forum, I don't know what does.
I agree with the mailing list argument - in this case I willingly subscribe to a public forum. The vast majority of spam I receive is however not from any lists I'm a member of (the moderators take care of it - If there were not some control on those lists I prob wouldn't use them). I don't know if you mean that email itself is publicly financed, if you do this is erroneous - a good many people pay for connection time by the Mb. Downloading rubbish costs time and money.
Whilst I agree that many of the legislative approaches are overblown (and dangerous), expecting all users to block their own spam is (which is what the EFF is clearly advocating) is seriously unrealistic. How many people here have a hotmail, yahoo, lycos.. account - what would that account look like if those companies didn't block spam for you? I'm sure that the average user would see this as a service offered by the ISP. As long as he/she can receive mail from granny it's fine.
Most average users just want "email", they don't want the hassle of configuring 1001 spam filters. It similar to virus protection - they will just install Dr. Solomons for SPAM - or use whatever comes in the next version of XP and have Bill limit who sends them email.
The free speech argument isn't invalid its just impractical for most end users. Secondly it is being applied in the following way by the EFF:
- "ANYBODY has the right to say anything to YOU"
and not in what most people consider free speech, which is:
- "ANYBODY has the right to say anything in a public forum."
These are NOT the same thing. You get into the whole "I'm paying time and money becuase idiots keep sending me spam". Email is personal communication (uni or multi cast) it is not broadcast. If people wish to broadcast they should do so in public forums - er, like this one!
It's still an issue if an ISP blocks somebody you do want to hear from - but this is somewhat akin to the fact that millions of people around the world don't even have access to email, a telephone or even a decent postal service to even contact me in any way whatsoever.
Being black listed at least forces those areas that are to try and regulate their users. Of course Eventually this is likely to break down to requiring pretty intelligent software to determine what to block based on message content rather than sender behaviour - and even then people will still pay third parties (ISPs,M$) to perform this for them - how many pieces of software out there still use the default passwords...
One of the main counterarguments to this is that if the AI purists are "right" then it should be basically impossible to fool a human for very long with a basic pattern matcher. I.e. your going to need some real reasoning, domain knowledge and (probably) learning to fool humans (silver medal).
These bots clearly have value on their own since they can be configured to talk about particular subjects and already act as a first base customer service interface (smarter than most tech support...).
Just to be cynical for once - this sounds like a "user centric" repackaging of a whole bunch of hard AI research: learning, reactive planning, goal driven behaviour and autonomous agent work in particular.
In the end it turns out that the most complex problem arise in trying to coordinate a collection of "autonomic" (?) components. Distributed systems with unrully objects... This is what the autonomous agent community is mainly concerned with ( see the UMBC agents page or this very useful overview paper for example).
Of course IBM pushing this it might mean a kick up rear for the academic to actually get some of this potentially cool stuff working. Chances are you never want the end user to know how it works anyway.
The wired article says that ammendment ddin't get through. Interestingly (from the wired article): "We might try and block somebody," Glazier said. "If we know someone is operating a server, a pirated music facility, we could try to take measures to try and prevent them from uploading or transmitting pirated documents."
It seems unlikely that hacking the individual machines would be the best solution for this (even if the law were to allow it). The cost would be very high. Much cheaper to do what they are now doing:
Leaning on ISPs to cut off "abusing" users (without comeback - see previous slashdot stories)
Suing the larger sites (napster obviously)
Trying to stifle decryption technology.
In the long run these are likely to be 95% effective if the succeed. If their wording were to ever pass into law they would just be setting a dangerous precedent for anybody to go and explore somebody else's machine. I'm just off to RIAA's web site to "check" if they have a copy of my (copyrighted) memoires on the server...
I don't agree. I think lay people understand that there will always be ways to encrypt things which cannot be broken. The fundamental question is why are the technologies which make this as easy as sending an email?
I don't agree that one-time pads are sustainable for terrorists. Getting the same valid code book to a number of members in several countries? many of who might not know or trust each other?, regularly changing the code? using it for every messages.
At best u'd prob use one time pads to encode your daily keys for some other (faster and automatic) encryption mechanism.
Besides,in the end you will still be sending a message which makes no sense of any kind (the encrypted string). The FBI will come kocking on your door and say (prob not very politely) that they want the key. This is exactly the same result you would get if you used PGP and hadn't surrendered the key.
This is why stenography is so hot - you encode stuff in traffic which looks "innocent" so no one even knows you are sending an encrypted message.
Er, this totally ignores the massive problem with one time pads which is distribution. One time pads are uncrackable (unless you keep reusing them) but:
You have to get a copy to the person you're communicating with.
If your pad becomes compromised - somebody else gets a copy all your messages are compromised and it's much easier to size a book of codes than a private key.
Add to that lack of non-repudiation and the like and its not so hot for everyday use...
The long running Robo Cup Soccer challenge (Robocup.org) started a search and rescue competition a after one of those involved died in the Kobe earthquake:
you can build your own coordinated team of virtual robots to work in a simulated disaster zone + there are some search and rescue competitions using real robots. There's a download page for the simulators and disaster toolkits.
do I have a right to say what I like without the government knowing?
do I have a right to say what I like without another government knowing?
do I have a right to say what I like without my company knowing,
my mother, my neighour, Joe Bloggs?...
Castrating encryption software takes away all these rights in pursuit of the allowing the first (sometimes). Forcing key eschrow (a la RIP) is also unattractice but at least boss will have a bit of work to do before she knows what I'm saying about her - then you get into the interesting problem of wiretapping across national boundaries...
Seemt to me there are two important distinctions which have to be made in these privacy arguments:
Limiting available "secure technology" (e.g. strong ecryption export restrictions in the US) VS something like mandatory key eschrow.
National government having access to communications in its own territory / between its own citizens VS across national boundaries.
With the first I'm refering to attempts to criple available encryption software (e.g. through US export rules) which would basically mean that communication can only be weakly encrypted w.r.t to anybody (i.e. Joe Bloggs can crack my communications with my mother if he is determined enough). This is different to the UK's RIP laws which aim to force people to give copies of their keys to some government institution. The later solution means Old Joe is locked out whilst the government can still spy. (until somebody cracks the bank of keys...)
The second point is that it is fine and dandy for the US to wish to monitor all communications - do they expect to do this in China? in Russia? in the UK? Many of the communciations which might be involved in terrorist or illegal activity planned in one country may take in other juridictions. No doubt there is/will be inter government cooperation but it is unlikely to extend to sharing access to all communications in a given country.
Poor security technology (1. above) puts everybody at risk - governments would not be able to communicate safely, indivuduals would have their privacy compromised to anybody (not just the government). So you might be left with (at best) forced key eschrow and in the long run governments may be able to monitor their own citizens but they certainly wont have global access - those cryptologists will still have work to do.
On top of this if you want to publish where do you publish? In a high impact factor journal because your work is going to be seen and often you grant is linked to impact factor. So researchers are so desperate to get their money they give copyright of their work to journals.
I think you'll find that it is *very* rare for researchers to be paid for articles in scientific journals. Why do they publish there - because the academic establishment reuires you to publish publish publish - and precisely in this expensive journals. In some fields (medicine for example) this has lead to researchers banding together their own peer-reviewed journals.
What is needed is to break the cycle is for researchers to publish online to a respected website and to keep copyright of their work and for funding companies / governments to acknowledge these as having an impact factor (may be based on unique viewing of page I suppose ) and the libraries to stop paying them!
Agreed - but then what's a respected web site? will it archive your documents for eternity? will it classify them according to field? enable review?
One option for free publication of workshop is proceedings CEUR Sunsite at Univ Aachen. no reviewing possible there though.
Slashdot for scientific papers needed...
Slashdot Mirror
Marking up pages with information about the meaning of the terms on them is the main thrust of the work on semantic web - see http://www.daml.org/ (for DAML - the DARPA Agent Markup Language), http://www.semanticweb.org/ (One of the main information sources) and finally the new W3C activity on the subject: http://www.w3.org/2001/sw/.
How far, how fast it will go is another matter but there's certainly a lot of interest in creating a more "machine readable" web.
The Father of GP (John Koza) may disagree with me - he runs genetic-programming.org and more or less invented the field. He's also known for his vigorous defences of GP: anybody know of real applications?
A somewhat more complete description of GP can be found at Genetic-programming.com.
- (Minor): a standard interface to an on-line address book,.. whatever.
- (Major) access to their user's data (with the use's permission).
They are playing a chicken and egg game. Most users will want to store their personal data in one place (their machine or some server) and not enter this in 10,000 times. No how does my application X access Johnny's data when he logs onto my site? It's authenticates the passport and get's its from the microsoft server.So you get access to a "service" which allows you to "serve" anybody who uses a passport. If passport becomes the worldwide standard you'd defn be getting your money's worth at $250 a pop. Let's bet that the cost won't stay there, Wait for it 1) to increase, 2) to become "per transaction" based. That's where the money really starts flowing.
However I believe that the principle behind .NET is a no-brainer good idea - people need a service like this. Clealry however it should be based top to bottom on an open standard so that anybody is able to host personal passport data (e.g. W3c P3P.
On a realistic note thought I think open ways of doing this will emerge - because many people will want a slice of this pie and people won't be held to ransom. Microsoft will obtain a big first mover advantage but also carries a lot of cost. What is critical is to check out the patent situation in the area (because you can bet your life there will be a large arsenal of them).
Roll on the OpenSource standard passport + third party providers to host that data.
Mass decryption just isn't feasible and certainly not in real time so they have to try to do one of two things:
- Intelligence can depend on the environment: Is a spider intelligent? Spining a web it's amazing, stick the thing in a bath tub and it doesn't look so smart.
- Intelligence can be social: is an ant intelligent? Not by itself but ant colonies perform some pretty amazing feats.
- Intelligence may depend on other knowledge: A chess grandmaster may play a very strange move near the beginning of the game which looses him the game. Why? He took a calculated risk and it didn't pay off. Was he dumb? No, you say. What if it wasn't a chess grandmaster but Joe Blogss from down the street - yeah THAT was a dumb move...
Perception of intelligence is about being seen to to the right thing at the right time.Regarding the second point - this gets to the heart of the Chinese Room Argument: can intelligence (I would distinguish "sentience") be "built" or "must" there be something more. Was deep blue intelligent? Searle would argue "no". Some would argue "Yes, In the chess domain". There was nobody on the planet it couldn't teach somethign about chess and (to an extent) explain those choices. Many AI researchers weren't happy about deep blue because it basically used very fast search and no fancy reasoning. But hey - that just shows that there's more than one way to solve a problem IMHO...
Sorry but last time I checked this wasn't going to happen in our universe! But hey maybe it'll be based on DRM2 (slashdot).
Can email be a public forum? Yes it most certainly can -- in the form of mailing lists. So a spammer uses a mailing list, and it becomes completely legal. Why? He's communicating to a large number of people using a publicly financed network. If that doesn't sound like a public forum, I don't know what does. I agree with the mailing list argument - in this case I willingly subscribe to a public forum. The vast majority of spam I receive is however not from any lists I'm a member of (the moderators take care of it - If there were not some control on those lists I prob wouldn't use them). I don't know if you mean that email itself is publicly financed, if you do this is erroneous - a good many people pay for connection time by the Mb. Downloading rubbish costs time and money.
The free speech argument isn't invalid its just impractical for most end users. Secondly it is being applied in the following way by the EFF:
- "ANYBODY has the right to say anything to YOU"
and not in what most people consider free speech, which is:
- "ANYBODY has the right to say anything in a public forum."
These are NOT the same thing. You get into the whole "I'm paying time and money becuase idiots keep sending me spam". Email is personal communication (uni or multi cast) it is not broadcast. If people wish to broadcast they should do so in public forums - er, like this one!
It's still an issue if an ISP blocks somebody you do want to hear from - but this is somewhat akin to the fact that millions of people around the world don't even have access to email, a telephone or even a decent postal service to even contact me in any way whatsoever.
Being black listed at least forces those areas that are to try and regulate their users. Of course Eventually this is likely to break down to requiring pretty intelligent software to determine what to block based on message content rather than sender behaviour - and even then people will still pay third parties (ISPs,M$) to perform this for them - how many pieces of software out there still use the default passwords...
- US National Labs rotating Java model (doesn't show the bonds though).
- loads of static models at Rice.edu.
Nice one Mr.Buckminster...These bots clearly have value on their own since they can be configured to talk about particular subjects and already act as a first base customer service interface (smarter than most tech support...).
In the end it turns out that the most complex problem arise in trying to coordinate a collection of "autonomic" (?) components. Distributed systems with unrully objects... This is what the autonomous agent community is mainly concerned with ( see the UMBC agents page or this very useful overview paper for example).
Of course IBM pushing this it might mean a kick up rear for the academic to actually get some of this potentially cool stuff working. Chances are you never want the end user to know how it works anyway.
It seems unlikely that hacking the individual machines would be the best solution for this (even if the law were to allow it). The cost would be very high. Much cheaper to do what they are now doing:
- Leaning on ISPs to cut off "abusing" users (without comeback - see previous slashdot stories)
- Suing the larger sites (napster obviously)
- Trying to stifle decryption technology.
In the long run these are likely to be 95% effective if the succeed. If their wording were to ever pass into law they would just be setting a dangerous precedent for anybody to go and explore somebody else's machine. I'm just off to RIAA's web site to "check" if they have a copy of my (copyrighted) memoires on the server...I don't agree that one-time pads are sustainable for terrorists. Getting the same valid code book to a number of members in several countries? many of who might not know or trust each other?, regularly changing the code? using it for every messages.
At best u'd prob use one time pads to encode your daily keys for some other (faster and automatic) encryption mechanism.
Besides ,in the end you will still be sending a message which makes no sense of any kind (the encrypted string). The FBI will come kocking on your door and say (prob not very politely) that they want the key. This is exactly the same result you would get if you used PGP and hadn't surrendered the key.
This is why stenography is so hot - you encode stuff in traffic which looks "innocent" so no one even knows you are sending an encrypted message.
- You have to get a copy to the person you're communicating with.
- If your pad becomes compromised - somebody else gets a copy all your messages are compromised and it's much easier to size a book of codes than a private key.
Add to that lack of non-repudiation and the like and its not so hot for everyday use...href="http://www.r.cs.kobe-u.ac.jp/robocup-rescue
you can build your own coordinated team of virtual robots to work in a simulated disaster zone + there are some search and rescue competitions using real robots. There's a download page for the simulators and disaster toolkits.
- do I have a right to say what I like without the government knowing?
- do I have a right to say what I like without another government knowing?
- do I have a right to say what I like without my company knowing,
- my mother, my neighour, Joe Bloggs?...
Castrating encryption software takes away all these rights in pursuit of the allowing the first (sometimes). Forcing key eschrow (a la RIP) is also unattractice but at least boss will have a bit of work to do before she knows what I'm saying about her - then you get into the interesting problem of wiretapping across national boundaries...- Limiting available "secure technology" (e.g. strong ecryption export restrictions in the US) VS something like mandatory key eschrow.
- National government having access to communications in its own territory / between its own citizens VS across national boundaries.
With the first I'm refering to attempts to criple available encryption software (e.g. through US export rules) which would basically mean that communication can only be weakly encrypted w.r.t to anybody (i.e. Joe Bloggs can crack my communications with my mother if he is determined enough). This is different to the UK's RIP laws which aim to force people to give copies of their keys to some government institution. The later solution means Old Joe is locked out whilst the government can still spy. (until somebody cracks the bank of keys...)The second point is that it is fine and dandy for the US to wish to monitor all communications - do they expect to do this in China? in Russia? in the UK? Many of the communciations which might be involved in terrorist or illegal activity planned in one country may take in other juridictions. No doubt there is/will be inter government cooperation but it is unlikely to extend to sharing access to all communications in a given country.
Poor security technology (1. above) puts everybody at risk - governments would not be able to communicate safely, indivuduals would have their privacy compromised to anybody (not just the government). So you might be left with (at best) forced key eschrow and in the long run governments may be able to monitor their own citizens but they certainly wont have global access - those cryptologists will still have work to do.
"All your keys are belong to us..."
On top of this if you want to publish where do you publish? In a high impact factor journal because your work is going to be seen and often you grant is linked to impact factor. So researchers are so desperate to get their money they give copyright of their work to journals. I think you'll find that it is *very* rare for researchers to be paid for articles in scientific journals. Why do they publish there - because the academic establishment reuires you to publish publish publish - and precisely in this expensive journals. In some fields (medicine for example) this has lead to researchers banding together their own peer-reviewed journals. What is needed is to break the cycle is for researchers to publish online to a respected website and to keep copyright of their work and for funding companies / governments to acknowledge these as having an impact factor (may be based on unique viewing of page I suppose ) and the libraries to stop paying them! Agreed - but then what's a respected web site? will it archive your documents for eternity? will it classify them according to field? enable review? One option for free publication of workshop is proceedings CEUR Sunsite at Univ Aachen. no reviewing possible there though. Slashdot for scientific papers needed...