why are they stopping the V2 security updates half a year after v3 was released?
Because they have finite resources and want to concentrate on keeping v3 fully secured.
The beauty is, since FF is open source, this potentially opens up a market for some third party to continue patching FF2 where Mozilla left off (if in fact there is any sizeable contingent of users who just cannot bear to upgrade). That's much less likely to happen with a closed-source browser simply because of the code being proprietary.
I can conceive of no situation in which a human could recognize the correct layout of a chessboard but a computer couldn't. We've already mentioned the spatial transformations. Lighting is not an issue--haven't you ever used photoshop? In fact computers are better at lighting in some cases because they are immune to certain optical illusions.
Shakiness is easily corrected with image stabilization techniques.
The thing to worry about is not whether the computer can make sense of the relation between the external world and its internal model, but whether it can infer what the model is with no preconception of what the rules are. But then, there is still debate about whether we cando thatourselves.
Most vision systems I've seen use very specific models to compare images to. Yes, such as a model of a generalized chess board (roughly square, eight-by-eight fields, with a specific initial position which can be used to infer what each of the pieces look like based on their locations).
The computer would have to be able to capture images of the board in near real-time to be able to determine when the opponent moved, and how it moved. Again, a web camera is perfectly capable of doing this (although you'd probably want something with better resolution). When looking at something as static as a chess board, detecting and interpreting movements as changes in the image is fairly easy (just detect if any pixel values have changed significantly, and if so determine which square they correspond to).
It would also need sufficient accuracy in image acquisition, interpretation, and AI to determine where exactly the robotic arm needs to grab the piece without having it slip or tilt - that is an enormous task of physics simulation. It would need to be calibrated, yes, just like any other mechanical arm needs to be calibrated so that external actions correspond to internal representations. This happens to us as well as our nervous systems are developing. But we can do it a lot faster in a computer because it's a more controlled process. We're not teaching it how to swing a baseball bat, just to move a little arm at constant velocity and stop it at the right place.
Next is realtime spacial awareness, the system would have to create a 3D model of the board, pieces, and the robotic arm in space, and compute distances & movements for pick up and placement all without hitting other pieces, accidently dropping the piece above the board, or pushing the piece through the board. That's what 3D coordinates and physics models are for. Ever played Quake?
... we could find ourselves surrounded by technology we do not understand...
For the (vast?) majority of humankind this is already true.
There is an important issue of control though. We might possibly create beings with above-human intelligence which are nevertheless not above humans in power (or freedom or whatever you wish to call it). That is, they might be capable of immense feats of intelligence, but all of their capability will be transparent to us by virtue of our having complete access to and control of the contents of their "minds". That is not to say we will automatically understand everything they think or do, but at least we have the data to try and understand. Also, while they may have great freedom in the intellectual realm, we can maintain tight control over their control of the external world--simply by keeping them relatively disembodied. The one thing to worry about is the possibility of computers discovering social engineering. When the machine figures out how to brainwash the operator, then we're in trouble;-p (incidentally, that would make an awesome AI movie).
Imagine the size of big blue if it had to actually see the board and physically move the pieces. Actually, those two tasks are fairly easy compared to the task of winning a game of chess. This is because chess boards all have a fairly consistent look, and even if presented with a strange, unfamiliar-looking chess board, a decent "chess vision" algorithm should have relatively little trouble inferring what the pieces are etc. Similarly, good robotics coupled with a good 3D world model should take care of moving the pieces relatively easily. So, your home computer with a webcam and a nice USB robotic arm attached could take care of those two tasks. Now, to deal with the 10^123 game-tree complexity...
Here's an idea: create one UI that is flexible enough to be customized however you like--even customized to have fewer settings to customize;-o (lots of programs have a choice between "basic" and "advanced" settings anyway). That way you not only have choice, but you can have a "best of both worlds" interface with your favorite features from one together with your favorite features from the other.
Why don't they just do what they normally do in these situations? Create a new project at www.wikimath.org and have the proofs live there. The Wikipedia pages can then reference the proof pages, just as they often do for wikibook, wikiquote, etc.
Is it possible that the chimps selected for this experiment just have an exceptionally good memory for chimps, maybe even photographic memories? It may be that we're comparing the Stephen Hawkings of chimps with a random sample of college students. We should have Stephen Hawking take the test to make it fair.
I agree with what many have said above. It appears that the author doesn't really know what computer science is--it is the study of the theory of computation. As such, it is entirely reasonable if not necessary to describe characteristics of computation (complexity of algorithms etc.) mathematically. As for 'process expression' versus 'algorithm', he seems to be splitting hairs just to make an (unconsequential) point. It's not like a theory of 'process expression' isn't already a large part of CS, it's just not the only large part. (And for the record, I *do* hope that my operating system is able to terminate, and cleanly too;-P). I think either this is a marketing ploy to get his company's name in the news, or he is just sore about getting bad grades in his math classes.
I wonder how Fant will manage to describe his new theory of 'process expression' without using any mathematics.
This is just a special case of capitalism: paying wages for work is not necessary because all workers automatically benefit from the common good--they essentially pay themselves. Meanwhile, non-workers also benefit but *do* have to pay for the benefit--whether this payment is a price, a fine, or non-monetary punishment is immatterial.
This theory would seem to work, except that the common good is arbitrarily defined. The article even alludes to the fact that this type of system is supposedly used by the mafia to achieve what is a common good to the contributors but a common problem to the rest of society.
Slashdot Mirror
54Mbps should be enough for anybody.
http://en.wikipedia.org/wiki/Homomorphic_encryption:
http://en.wikipedia.org/wiki/Malleability_(cryptography):
Have you tried turning it off and back on again?
As if we needed another reason.
why are they stopping the V2 security updates half a year after v3 was released?
Because they have finite resources and want to concentrate on keeping v3 fully secured.
The beauty is, since FF is open source, this potentially opens up a market for some third party to continue patching FF2 where Mozilla left off (if in fact there is any sizeable contingent of users who just cannot bear to upgrade). That's much less likely to happen with a closed-source browser simply because of the code being proprietary.
Yes, but trees currently cost less than solar cells.
How many solar cells do you need to power a pizza oven, anyway?
How about two sticks and some kindling?
What a horrible horrible waste of power consumption.
Man does not fly. The machine does.
Not the man. Not the machine. Mind is flying.
You can't prove things in physics.
No, really.
You are exactly right, but to paraphrase:
"All models are wrong, but some are useful."
Someone will *have* to be the loser.
You're assuming the game actually ends. But we know that computers are prone to infinite loops :-p.
I can conceive of no situation in which a human could recognize the correct layout of a chessboard but a computer couldn't. We've already mentioned the spatial transformations. Lighting is not an issue--haven't you ever used photoshop? In fact computers are better at lighting in some cases because they are immune to certain optical illusions. Shakiness is easily corrected with image stabilization techniques.
The thing to worry about is not whether the computer can make sense of the relation between the external world and its internal model, but whether it can infer what the model is with no preconception of what the rules are. But then, there is still debate about whether we can do that ourselves.
... and it will take an extra 5 seconds to checkmate you.
... we could find ourselves surrounded by technology we do not understandFor the (vast?) majority of humankind this is already true.
There is an important issue of control though. We might possibly create beings with above-human intelligence which are nevertheless not above humans in power (or freedom or whatever you wish to call it). That is, they might be capable of immense feats of intelligence, but all of their capability will be transparent to us by virtue of our having complete access to and control of the contents of their "minds". That is not to say we will automatically understand everything they think or do, but at least we have the data to try and understand. Also, while they may have great freedom in the intellectual realm, we can maintain tight control over their control of the external world--simply by keeping them relatively disembodied. The one thing to worry about is the possibility of computers discovering social engineering. When the machine figures out how to brainwash the operator, then we're in trouble
A Beowulf cluster!
Here's an idea: create one UI that is flexible enough to be customized however you like--even customized to have fewer settings to customize ;-o (lots of programs have a choice between "basic" and "advanced" settings anyway). That way you not only have choice, but you can have a "best of both worlds" interface with your favorite features from one together with your favorite features from the other.
Posts like this make me wish moderation scores went above +5.
Why don't they just do what they normally do in these situations? Create a new project at www.wikimath.org and have the proofs live there. The Wikipedia pages can then reference the proof pages, just as they often do for wikibook, wikiquote, etc.
Is it possible that the chimps selected for this experiment just have an exceptionally good memory for chimps, maybe even photographic memories? It may be that we're comparing the Stephen Hawkings of chimps with a random sample of college students. We should have Stephen Hawking take the test to make it fair.
I agree with what many have said above. It appears that the author doesn't really know what computer science is--it is the study of the theory of computation. As such, it is entirely reasonable if not necessary to describe characteristics of computation (complexity of algorithms etc.) mathematically. As for 'process expression' versus 'algorithm', he seems to be splitting hairs just to make an (unconsequential) point. It's not like a theory of 'process expression' isn't already a large part of CS, it's just not the only large part. (And for the record, I *do* hope that my operating system is able to terminate, and cleanly too ;-P). I think either this is a marketing ploy to get his company's name in the news, or he is just sore about getting bad grades in his math classes.
I wonder how Fant will manage to describe his new theory of 'process expression' without using any mathematics.
This is just a special case of capitalism: paying wages for work is not necessary because all workers automatically benefit from the common good--they essentially pay themselves. Meanwhile, non-workers also benefit but *do* have to pay for the benefit--whether this payment is a price, a fine, or non-monetary punishment is immatterial.
This theory would seem to work, except that the common good is arbitrarily defined. The article even alludes to the fact that this type of system is supposedly used by the mafia to achieve what is a common good to the contributors but a common problem to the rest of society.