I remember downloading a zmodem client over SuperKermit on a 2400 baud modem.
I was so ignorant and innocent back in those days... some of my friends did not have Internet access, so we all shared the same account and voluntarily did not read each other's email (although that sometimes happened accidently if we were not careful with mailx).
And the joys of figuring out for the first time how to use rm on a file named '-'... Wow, I could go on and on about the old days...
And I'm sure some folks here can tell even older stories.
You obviously haven't read any technical research papers. You always cite other work you borrowed techniques from.
And I remember specific classes at Georgia Tech where I was either told (as a student) or I told students (as a TA) to write down in their HW the names of people that they discussed it with.
Considering this occurred in Fall 2001, "off-week" must have meant weekend or a holiday. Assignments are posted at least seven days in advance. Hence, if you start early, you will *always* have a chance to talk to TAs or the professor. However, if you wait till the last minute, you may not be able to get help: TAs are not on-call paramedics.
What's so hard about starting an assignment early?
And the CS13xx courses have newsgroups for asking questions and have tons of TAs. There are recitations and labs and office hours. There is plenty of a chance for students to ask for help and get help. Unfortunately, too many students are lazy bastards and don't want to put forth the effort of doing the assignment honestly and getting help when they need it.
People are taking the Georgia Tech policy way out of context and way too harshly. The policy is simple: students may talk (and are encouraged to talk) about high-level issues, but when it comes down to writing code, they have to do separate work. Unfortunately the average student fears getting caught for cheating and interprets this rule way too harshly.
In some of my grad classes at UW, we have a similar policy, the Gilligan's Island policy. You may talk about the assignments as much as you want, but before you work on it yourself, watch an episode of Gilligan's Island.
Regardless, if he would have just *cited* what he had borrowed from other people, he would not be guilty of academic misconduct. The graders may not have given him credit for creating that portion of the code, but he would have been honest about what was his own original work.
Those policies are really only for the introductory courses. Face it, coding is something that takes time. It is applied. You cannot be tested in just an hour on coding abilities. The homework assignments for CS13xx serve as a form of test. Once the students "pass" this test and take later CS courses, most of the projects are collaborative in nature from the sheer magnitude of what has to be coded. But at some point, people have to be judged on their ability to code. Find me a better way to judge and I'll be all ears.
Whew, it's been a long time since I've taken algebra... Froebenius's result just applies to associative division algebras. Octonions are the only other finite-dimensional real division algebra.
http://mathworld.wolfram.com/DivisionAlgebra.htm l
All these reports indicate one thing: while it may take millions of monkeys hammering away at keyboards to get what you want, use only a few thousand pigeons and get it faster and better.
Your argument that biological intelligence does not use symbolic manipulation is hollow. It's like saying computer intelligence does not use symbolic manipulation because it's just a bunch of transistors.
There has been plenty of great research in reinforcement learning. Isn't that what you mean by rewards and punishments?
People have also tried to make neurologically inspired models for AI. Note that the transistor is even abstractly similar to the neuron. The AI community is in much better shape than you think.
I think you are missing the point. A well written machine learning program can be incredibly short and simple. Even in less than ten lines of MATLAB code. But it requires a lot more than elite hacking skills to be able to prove mathematically that an algorithm, however so simple, will perform optimally in the Bayesian sense.
While you may not consider that to be a contribution, the same model may be used for more complicated things, such as piloting a spacecraft. I saw a talk the other day who trained a program to hover a remote-controlled helicopter in place. It performed better than the leading human controllers.
There's a lot more to AI than symbol manipulation. Knowledge representation is a very small subset of the field. Some researchers choose to concentrate on the small subfields of AI, and those fields have prospered, providing great advances to datamining, graphics and vision, theory, etc. Some researchers are very much interested in neurologically and biologically inspired computing. Calling AI researchers clueless for ignoring these areas is just revealing your ignorance about AI and your fanatacism against AI..
I know very few programmers outside of AI or mechanical engineering who can write a program to perform optimal cruise control given sensor/motor noise and unknown road conditions. There's more to it than "Read the speedometer. Accelerate if too slow, decelerate if too fast. Repeat."
What if "optimal" were defined as some weighted combination making the ride smooth and conserving gas? Is that so trivial a program to write? AI researchers have solved many important problems over the years, and I don't consider it at all a disappointment that we are not even close to approaching human intelligence.
Cruise control could be very well formulated as an AI problem. There is sensor noise from the speedometer. There are uphills and downhills and different road conditions. In this case, it probably boils down to "just" a Kalman filter, but a Kalman filter easily qualifies as machine learning.
The classic blockworld problem was creating a plan to stack three blocks which required some work to be "undone" before reaching the goal.
The catch though is it was a very rigorous treatment, and it was a very elegant paper because it distilled one of the difficulties in planning to its simplest case.
Game AI can focus on problems that are as simple, but the goal is different -- the goal is simply to make the game enjoyable for the player. So the game AI can cheat (have access to more information) or be hard-coded.
Slashdot Mirror
But don't the black ones have bigger disks? ;)
I remember using lynx in 1994.
Nothing beats:
Would you like to quit? (y/n)
Excellent!
I remember downloading a zmodem client over SuperKermit on a 2400 baud modem.
I was so ignorant and innocent back in those days... some of my friends did not have Internet access, so we all shared the same account and voluntarily did not read each other's email (although that sometimes happened accidently if we were not careful with mailx).
And the joys of figuring out for the first time how to use rm on a file named '-'... Wow, I could go on and on about the old days...
And I'm sure some folks here can tell even older stories.
> Did you even read the article?
Dude, I went to the friggin' school. The student in the article is spouting a load of crap and you're believing every cent of it as 100% truth.
You obviously haven't read any technical research papers. You always cite other work you borrowed techniques from.
And I remember specific classes at Georgia Tech where I was either told (as a student) or I told students (as a TA) to write down in their HW the names of people that they discussed it with.
Considering this occurred in Fall 2001, "off-week" must have meant weekend or a holiday. Assignments are posted at least seven days in advance. Hence, if you start early, you will *always* have a chance to talk to TAs or the professor. However, if you wait till the last minute, you may not be able to get help: TAs are not on-call paramedics.
What's so hard about starting an assignment early?
Right, so for that portion of the code, write a comment, "This part really confused me, so I asked Billybob for help and he gave me some advice."
And the CS13xx courses have newsgroups for asking questions and have tons of TAs. There are recitations and labs and office hours. There is plenty of a chance for students to ask for help and get help. Unfortunately, too many students are lazy bastards and don't want to put forth the effort of doing the assignment honestly and getting help when they need it.
People are taking the Georgia Tech policy way out of context and way too harshly. The policy is simple: students may talk (and are encouraged to talk) about high-level issues, but when it comes down to writing code, they have to do separate work. Unfortunately the average student fears getting caught for cheating and interprets this rule way too harshly.
In some of my grad classes at UW, we have a similar policy, the Gilligan's Island policy. You may talk about the assignments as much as you want, but before you work on it yourself, watch an episode of Gilligan's Island.
Regardless, if he would have just *cited* what he had borrowed from other people, he would not be guilty of academic misconduct. The graders may not have given him credit for creating that portion of the code, but he would have been honest about what was his own original work.
Those policies are really only for the introductory courses. Face it, coding is something that takes time. It is applied. You cannot be tested in just an hour on coding abilities. The homework assignments for CS13xx serve as a form of test.
Once the students "pass" this test and take later CS courses, most of the projects are collaborative in nature from the sheer magnitude of what has to be coded. But at some point, people have to be judged on their ability to code. Find me a better way to judge and I'll be all ears.
Whew, it's been a long time since I've taken algebra... Froebenius's result just applies to associative division algebras. Octonions are the only other finite-dimensional real division algebra.
m l
http://mathworld.wolfram.com/DivisionAlgebra.ht
Big Yellow :)
Sure, and by the Celebrated Theorem of Froebenius, there is no such thing in a higher dimension. :)
Here's an algebraic topology version of the problem:
Given a simply connected tetrahedral mesh, show that the mesh can be collapsed by topologically invariant operations to a single tetrahedron.
Without reading the preprint, I cannot say (not that I could understand it anyway :) ). But it wouldn't surprise me if the proof was just for 3.
R^3 is kind of a magical place. R^2 might not have enough wiggling room, but R^4 might have too much. There exists a cross product in only R^3.
All these reports indicate one thing: while it may take millions of monkeys hammering away at keyboards to get what you want, use only a few thousand pigeons and get it faster and better.
Your argument that biological intelligence does not use symbolic manipulation is hollow. It's like saying computer intelligence does not use symbolic manipulation because it's just a bunch of transistors.
There has been plenty of great research in reinforcement learning. Isn't that what you mean by rewards and punishments?
People have also tried to make neurologically inspired models for AI. Note that the transistor is even abstractly similar to the neuron. The AI community is in much better shape than you think.
I think you are missing the point. A well written machine learning program can be incredibly short and simple. Even in less than ten lines of MATLAB code. But it requires a lot more than elite hacking skills to be able to prove mathematically that an algorithm, however so simple, will perform optimally in the Bayesian sense.
While you may not consider that to be a contribution, the same model may be used for more complicated things, such as piloting a spacecraft. I saw a talk the other day who trained a program to hover a remote-controlled helicopter in place. It performed better than the leading human controllers.
There's a lot more to AI than symbol manipulation. Knowledge representation is a very small subset of the field. Some researchers choose to concentrate on the small subfields of AI, and those fields have prospered, providing great advances to datamining, graphics and vision, theory, etc. Some researchers are very much interested in neurologically and biologically inspired computing. Calling AI researchers clueless for ignoring these areas is just revealing your ignorance about AI and your fanatacism against AI..
I know very few programmers outside of AI or mechanical engineering who can write a program to perform optimal cruise control given sensor/motor noise and unknown road conditions. There's more to it than "Read the speedometer. Accelerate if too slow, decelerate if too fast. Repeat."
What if "optimal" were defined as some weighted combination making the ride smooth and conserving gas? Is that so trivial a program to write? AI researchers have solved many important problems over the years, and I don't consider it at all a disappointment that we are not even close to approaching human intelligence.
Cruise control could be very well formulated as an AI problem. There is sensor noise from the speedometer. There are uphills and downhills and different road conditions. In this case, it probably boils down to "just" a Kalman filter, but a Kalman filter easily qualifies as machine learning.
AI has to be real-time also in many applications. Think robots.
The 0th law today would be to protect copyright for greedy corporate executives.
character recognition software that reads zipcodes in the post office
natural language translation from french to english
diagnosis and treatment of disease
datamining
texture synthesis
making a helicopter hover still in the air
Robotics is interesting in that it is the holistic (Rod Brooks) view of AI: a robot needs sensory systems, control systems, a planner, etc.
Academic AI can be very simple too.
The classic blockworld problem was creating a plan to stack three blocks which required some work to be "undone" before reaching the goal.
The catch though is it was a very rigorous treatment, and it was a very elegant paper because it distilled one of the difficulties in planning to its simplest case.
Game AI can focus on problems that are as simple, but the goal is different -- the goal is simply to make the game enjoyable for the player. So the game AI can cheat (have access to more information) or be hard-coded.
SPEC
see the 1st quarter 2002 results for CPU2000