video software that can track 3-dimensional objects using a 2-dimensional video image. This is a known solved problem.
A second layer of software (that uses the first as input) that distinguishes static and moving objects. Static objects form a "background" which can largely be ignored except for collisions with moving objects and except for specific human-input exceptions.
A rule database. The initial rule database must have many rules about default object behavior and interactions.
A learning system that detects heretofore unseen interactions, alerts a human user and asks for a new rule or validation of the anomaly as something that should cause an alert.
Moving objects must be dynamically tracked and their behavior somehow segmented into steps. This is arbitrary in that there can be a theoretically infinite variety of such segments for any given macro behavior. E.g., I can say "Mary handed the book to Tom." or I can say "Mary grasped the book, Mary extended her arm in the direction of Tom, Tom extended his hand in the direction of Mary and grasped the book, Mary un-grasped the book, Tom retracted his arm (with the book)." Both describe the same action, but the second has a finer segmentation of behavior.
All the pieces are available in freeware. It's essentially a classic AI expert system. Since they're getting millions for known technology (and software), I expect to see a freeware version of this available soon!8-))
video software that can track 3-dimensional objects using a 2-dimensional video image. This is a known solved problem.
A second layer of software (that uses the first as input) that distinguishes static and moving objects. Static objects form a "background" which can largely be ignored except for collisions with moving objects and except for specific human-input exceptions.
A rule database. The initial rule database must have many rules about default object behavior and interactions.
A learning system that detects heretofore unseen interactions, alerts a human user and asks for a new rule or validation of the anomaly as something that should cause an alert.
Moving objects must be dynamically tracked and their behavior somehow segmented into steps. This is arbitrary in that there can be a theoretically infinite variety of such segments for any given macro behavior. E.g., I can say "Mary handed the book to Tom." or I can say "Mary grasped the book, Mary extended her arm in the direction of Tom, Tom extended his hand in the direction of Mary and grasped the book, Mary un-grasped the book, Tom retracted his arm (with the book)." Both describe the same action, but the second has a finer segmentation of behavior.
In summary I don't see that there is much new here: certainly all the pieces are available. It's essentially the classic AI expert system. Since they're getting millions for known technology (and software), I expect to see a freeware version of this available within the next 6 months!8-))
Wipe the drive with software. Do it several times with different programs if you're paranoid. Set up an assembly line to do it if you have many, with each individual responsible for a separate step. Test drives prior to re-release.
People are so badly mistaken about how recoverable disk data is: they believe the same way they believe in Bigfoot or the Loch Ness monster. What a waste of good work.
Since we're just hearing about these, my assumption is that production is long since underway and that we have already offered them to Israel (if not supplied them) a year or so ago, hoping that they'll use them in their Iran air raids instead of nukes.
But nukes will work while something less may not. And these MOBs are new and untested on the battlefield. Israel can try the new American bombs on the hope that they _may_ work or instead they may simply nuke Iran's facilities with their own bombs guaranteeing success. I would choose the latter and be done with it: close membership in the nuclear club while you still can.
It isn't necessary to use cookies, but they are easier to understand and use (initially) than the alternatives. Enter the Universal Incompetent Programmer (UIP) who seizes on the quickest, easiest solution and voila! - an Internet brimming over with cookies.
You could say cookies are a standard, and they are. But cookies also violate certain more important underlying HTTP conventions (e.g., REST). I know, you don't know what REST is.
In the end, laziness and intellectual sloppiness is what has brought us here. Most UIPs, like most people, are not in the top 10% of their group. It takes some smarts to read an RFC or a book. Most UIPs skip that to read Slashdot instead. Slashdot is OK, but sometimes you need to read something else, something more enlightening technically. Sometimes you should study; but most UIPs don't.
Personally I don't get it. Why is it so hard to accept? Reliance on academic authories has its pitfalls of course, but a certain point you need the humility to accept that there is no debate over this particular point among experts.
Perhaps because no academic authority has proven consistently successful in predicting the economy or the weather 5 years into the future?
Follow the money! The Chinese have so many dollars they're practically burning them. This guy was undoubtedly paid a couple million under the table by the Chinese for his "contribution". It's obvious he made a bundle because he is willing to go to jail for it.
All the crap about sharing with graduate students and not reading the regulations is ridiculous. When all is said and done the professor is set for life and we've lost our lead in another technological area. We should change the laws so that we shoot agents of industrial espionage rather than imprison them for a few years.
An organism interacts with it's environment by slightly modifying it's behavior. That behavior alters the environment, sometimes radically. Sometimes a positive feedback loop is established between organism and environment that causes unusully rapid evolutionary change. Man is the most extreme case of niche construction . See Niche Construction for details.
This is the best idea yet presented. And it's possible to do it.
The only correct solution is a jet engine that can handle birds. The OP's proposal is both smart and evolutionary (it's an extension of existing jet engine technology). So I suggest the OP get a patent ASAP!
Since Deja News, I've felt that the NNTP technical newsgroups were worth paying for - even more than for YouTube.
But back to the major topic. Yes, YouTube technical articles, e.g., "How to disassemble a Microwave", etc. ) are an excellent learning tool and worth paying for.
I would expect Google to pay authors for providing content on a per-view basis.
The D.E. Shaw's and hedge funds are crushing the "old school" traders as trading becomes more about speed (esp. algorithmic trading) and liquidity and less about connections.
They crushed them and then they went bankrupt (or soon will - it's hardly over) and in one fell swoop lost (or will soon lose) more money than the sum of their accumulated gains in prior decades.
Granted, the current crisis was caused by over-reliance on models, but that happened because most traders and managers did not understand the models and their limitations. To rectify that, there will be an even greater need for those trained in financial mathematics.
No. Taleb is saying that the process is not amenable to existing models. Without a model you can't manage risk and without managing risk you can't wisely invest. Continuing to rely in existing models and their limitations will bring about another financial disaster.
But his thesis is that such events are fundamentally unpredictable. If he made a fortune, it means _he_ was able to predict it, well enough to profit for it. Which argues not that the events are unpredictable, but rather that his model is better.
Not correct. What Taleb recommends is avoiding investments where risks cannot be estimated with confidence. He recommends a "dumbbell" strategy, investing
most (~90%) in safe instruments: T-bills, cash, etc.,
the remainder in high-risk long shots (biotech, nanotech, high tech,etc.).
Most of the latter investments will fail, but those that win should win big - enough to more than make up for the many little losses. Taleb speaks at length how this is very difficult to do for humans. IIRC he also speak of buying call options, since in that case losses can be limited but wins can be big.
Taleb says the stock market is much more risky than we have been led to believe. His book "The Black Swan" is fascinating, amusing, and chockfull of insights into human behavior. It is an important book to read at this time in financial history. The timing of the book's publication is almost a "Black Swan" event, since Taleb's main concerns are so pertinent to the recent financial collapse. It's important to note that he wrote and published the book before TSHTF.
I hate to say it but the fact that our vision system has complete coverage over the visual field is so f**ing obvious and has been shown so often before that there should be little need to do yet more research on that subject.
What is really valuable and novel about this research?
I see nothing illegal here. If a buyer is willing to buy a PC for $599 without realizing that there is an included service plan charge, then he must believe it is a good deal. Of course, $519 (the original price minus $80 service plan) would be a _better_ deal, but the original deal is enough for the buyer to buy.
So a deal is made, some profit is made, the user buys an unnecessary service plan but nonetheless is very happy with the price. Aren't we getting a little Victorian about the profit motive here? After all, the goal of business is to profit and survive, not to give up all your profits in service to the customer. That's the road to bankruptcy.
Another way to look at it is to ask if you would criticize the deal were the additional $80 pure profit.
You might be able to combine the above 2 methodologies:
Case-Based Reasoning - Define problem parameters, e.g., number of items, sorted array of item sizes, etc. and use those parameters as indexes into a database of past solutions. You'll have to try various parameters to find which work best. Over time gather a database of successful (problem, solution) pairs.
Genetic Algorithm - Given a new problem, compare it's parameters with those of other (problem, solution) pairs stored in your database. Select one or more "closest matches" and evaluate them to see if they solve the new problem. If they are inadequate use genetic algorithms(GA) to modify the closest matches. Store any new and useful (problem, solution) pairs in the database.
This is convenient because it learns on it's own to some degree but, if the GA won't solve the problem sufficiently after some N generations then a human can intervene, define a solution and store that into the database. So it can learn from humans or from it's own exploration.
We've successfully cured cancer lots of times. The problem is that there's millions of versions of cancer; heck, you could say everybody who gets cancer gets their own, personalized version. A person can get cancer, completely separate, unrelated cancers, multiple times.
You are the most renowned general of an undefeated army. The Emperor calls you to court and tells you: "We have received intelligence that a mighty army, intent on invasion, is gathering at the frontier. Go forth, engage and defeat them."
You return from the frontier 6 months later to tell the Emperor: "Indeed there is a mighty army. There are Teutons, Magyars, Turks, Armenians, Slavs, Mongols, Franks, Sarmatians, Goths (whom are composed of Visigoths, Ostrogoths, Western Goths) and Huns, to name a few. I have devised a taxonomy of the barbarian invaders..."
The Emperor interrupts "Have you defeated the enemy?"
...doctors may be one step closer to treating a multitude of diseases, including Alzheimer's, diabetes and Parkinson's.
To which I say "Horseshit!" The day that American medicine finds a cure for cancer, Alzheimer's, diabetes and Parkinson's is the day that American medicine goes out of business. Doctors, HMOs, big pharma and hospitals are too busy making money off the sick to fix these problems.
We've been paying for a war on cancer for over 50 years and don't have a cure; surgery remains butchery; antibiotics are losing their effectiveness; no significant inroads against viruses have been made. The American medical establishment needs a radical dollar-ectomy in the form of
reduced requirements for an M.D.,
More use of computerized diagnostic systems,
increased specialization to the point of effectiveness,
tracking of success rates and costs of individual doctors.
"One step closer" doesn't mean anything when you're miles from home.
Slashdot Mirror
To do what AISight does one needs:
Moving objects must be dynamically tracked and their behavior somehow segmented into steps. This is arbitrary in that there can be a theoretically infinite variety of such segments for any given macro behavior. E.g., I can say "Mary handed the book to Tom." or I can say "Mary grasped the book, Mary extended her arm in the direction of Tom, Tom extended his hand in the direction of Mary and grasped the book, Mary un-grasped the book, Tom retracted his arm (with the book)." Both describe the same action, but the second has a finer segmentation of behavior.
All the pieces are available in freeware. It's essentially a classic AI expert system. Since they're getting millions for known technology (and software), I expect to see a freeware version of this available soon!8-))
To do what AISight does one needs:
Moving objects must be dynamically tracked and their behavior somehow segmented into steps. This is arbitrary in that there can be a theoretically infinite variety of such segments for any given macro behavior. E.g., I can say "Mary handed the book to Tom." or I can say "Mary grasped the book, Mary extended her arm in the direction of Tom, Tom extended his hand in the direction of Mary and grasped the book, Mary un-grasped the book, Tom retracted his arm (with the book)." Both describe the same action, but the second has a finer segmentation of behavior.
In summary I don't see that there is much new here: certainly all the pieces are available. It's essentially the classic AI expert system. Since they're getting millions for known technology (and software), I expect to see a freeware version of this available within the next 6 months!8-))
Wipe the drive with software. Do it several times with different programs if you're paranoid. Set up an assembly line to do it if you have many, with each individual responsible for a separate step. Test drives prior to re-release.
People are so badly mistaken about how recoverable disk data is: they believe the same way they believe in Bigfoot or the Loch Ness monster. What a waste of good work.
being certain to include all members of the "XXVIIth General Assembly of the International Astronomical Union." Humanity must be saved.
C++ is a disaster.
Since we're just hearing about these, my assumption is that production is long since underway and that we have already offered them to Israel (if not supplied them) a year or so ago, hoping that they'll use them in their Iran air raids instead of nukes.
But nukes will work while something less may not. And these MOBs are new and untested on the battlefield. Israel can try the new American bombs on the hope that they _may_ work or instead they may simply nuke Iran's facilities with their own bombs guaranteeing success. I would choose the latter and be done with it: close membership in the nuclear club while you still can.
Yes, I'm probably talking about you.
It isn't necessary to use cookies, but they are easier to understand and use (initially) than the alternatives. Enter the Universal Incompetent Programmer (UIP) who seizes on the quickest, easiest solution and voila! - an Internet brimming over with cookies.
You could say cookies are a standard, and they are. But cookies also violate certain more important underlying HTTP conventions (e.g., REST). I know, you don't know what REST is.
In the end, laziness and intellectual sloppiness is what has brought us here. Most UIPs, like most people, are not in the top 10% of their group. It takes some smarts to read an RFC or a book. Most UIPs skip that to read Slashdot instead. Slashdot is OK, but sometimes you need to read something else, something more enlightening technically. Sometimes you should study; but most UIPs don't.
Incompetence is the key.
Perhaps because no academic authority has proven consistently successful in predicting the economy or the weather 5 years into the future?
his universally-shared complaint!
The subject line says it all.
I appreciate your insights into hiring but they would be more useful in a discussion about hiring. Here we're discussing renting textbooks.
Follow the money! The Chinese have so many dollars they're practically burning them. This guy was undoubtedly paid a couple million under the table by the Chinese for his "contribution". It's obvious he made a bundle because he is willing to go to jail for it. All the crap about sharing with graduate students and not reading the regulations is ridiculous. When all is said and done the professor is set for life and we've lost our lead in another technological area. We should change the laws so that we shoot agents of industrial espionage rather than imprison them for a few years.
An organism interacts with it's environment by slightly modifying it's behavior. That behavior alters the environment, sometimes radically. Sometimes a positive feedback loop is established between organism and environment that causes unusully rapid evolutionary change. Man is the most extreme case of niche construction . See Niche Construction for details.
And we can evolve faster than ever thanks to that!
This is the best idea yet presented. And it's possible to do it.
The only correct solution is a jet engine that can handle birds. The OP's proposal is both smart and evolutionary (it's an extension of existing jet engine technology). So I suggest the OP get a patent ASAP!
Since Deja News, I've felt that the NNTP technical newsgroups were worth paying for - even more than for YouTube.
But back to the major topic. Yes, YouTube technical articles, e.g., "How to disassemble a Microwave", etc. ) are an excellent learning tool and worth paying for.
I would expect Google to pay authors for providing content on a per-view basis.
They crushed them and then they went bankrupt (or soon will - it's hardly over) and in one fell swoop lost (or will soon lose) more money than the sum of their accumulated gains in prior decades.
No. Taleb is saying that the process is not amenable to existing models. Without a model you can't manage risk and without managing risk you can't wisely invest. Continuing to rely in existing models and their limitations will bring about another financial disaster.
Not correct. What Taleb recommends is avoiding investments where risks cannot be estimated with confidence. He recommends a "dumbbell" strategy, investing
Most of the latter investments will fail, but those that win should win big - enough to more than make up for the many little losses. Taleb speaks at length how this is very difficult to do for humans. IIRC he also speak of buying call options, since in that case losses can be limited but wins can be big.
Taleb says the stock market is much more risky than we have been led to believe. His book "The Black Swan" is fascinating, amusing, and chockfull of insights into human behavior. It is an important book to read at this time in financial history. The timing of the book's publication is almost a "Black Swan" event, since Taleb's main concerns are so pertinent to the recent financial collapse. It's important to note that he wrote and published the book before TSHTF.
They don't want to hire them, they want to catch them.
Anyone stupid enough to show an interest will be repaid by having their background and their "back" proctoscoped by the Feds.
I hate to say it but the fact that our vision system has complete coverage over the visual field is so f**ing obvious and has been shown so often before that there should be little need to do yet more research on that subject.
What is really valuable and novel about this research?
I see nothing illegal here. If a buyer is willing to buy a PC for $599 without realizing that there is an included service plan charge, then he must believe it is a good deal. Of course, $519 (the original price minus $80 service plan) would be a _better_ deal, but the original deal is enough for the buyer to buy.
So a deal is made, some profit is made, the user buys an unnecessary service plan but nonetheless is very happy with the price. Aren't we getting a little Victorian about the profit motive here? After all, the goal of business is to profit and survive, not to give up all your profits in service to the customer. That's the road to bankruptcy.
Another way to look at it is to ask if you would criticize the deal were the additional $80 pure profit.
You might be able to combine the above 2 methodologies:
Case-Based Reasoning - Define problem parameters, e.g., number of items, sorted array of item sizes, etc. and use those parameters as indexes into a database of past solutions. You'll have to try various parameters to find which work best. Over time gather a database of successful (problem, solution) pairs.
Genetic Algorithm - Given a new problem, compare it's parameters with those of other (problem, solution) pairs stored in your database. Select one or more "closest matches" and evaluate them to see if they solve the new problem. If they are inadequate use genetic algorithms(GA) to modify the closest matches. Store any new and useful (problem, solution) pairs in the database.
This is convenient because it learns on it's own to some degree but, if the GA won't solve the problem sufficiently after some N generations then a human can intervene, define a solution and store that into the database. So it can learn from humans or from it's own exploration.
You are the most renowned general of an undefeated army. The Emperor calls you to court and tells you: "We have received intelligence that a mighty army, intent on invasion, is gathering at the frontier. Go forth, engage and defeat them."
You return from the frontier 6 months later to tell the Emperor: "Indeed there is a mighty army. There are Teutons, Magyars, Turks, Armenians, Slavs, Mongols, Franks, Sarmatians, Goths (whom are composed of Visigoths, Ostrogoths, Western Goths) and Huns, to name a few. I have devised a taxonomy of the barbarian invaders..."
The Emperor interrupts "Have you defeated the enemy?"
Moral: Current medicine is B Ship material.
To which I say "Horseshit!" The day that American medicine finds a cure for cancer, Alzheimer's, diabetes and Parkinson's is the day that American medicine goes out of business. Doctors, HMOs, big pharma and hospitals are too busy making money off the sick to fix these problems.
We've been paying for a war on cancer for over 50 years and don't have a cure; surgery remains butchery; antibiotics are losing their effectiveness; no significant inroads against viruses have been made. The American medical establishment needs a radical dollar-ectomy in the form of
"One step closer" doesn't mean anything when you're miles from home.
The title says it all.