Your analogies are facile. Spam, telegraph operators... It's the big media that arranges the generation of content. The foreign correspondents? The investigative journalists? The months spent on assignment abroad? The beat writers that follow a neighbourhood for years? If you think the echo chamber of inter-referencing bloggers will be a substitute, you've got another thing coming. Commenting on content is nice. Tagging it is great. Searching through it is wonderful. But someone's got to go and spent hundreds of hours on creating it first. And someone else has got to pay for it. At the moment, the guys who had the settled models in place are being driven out of business by outfits ultimately burning through venture capitalist money; eventually, they too will get around to charging for their services, and the era of "free" (i.e. subsidised) lunch will be over, but the damage will be done. Read up on what's happened to the staffs of LA Times, Chicago Tribune, Philadelphia Inquirer, etc., count the Pulitzers those papers had won, and compare that to the output of even the most precocious blog.
Spam is irrelevant to this matter. AdBlock is indeed a good analogue. You did note that I prefaced the second paragraph with "follow your choice to its logical conclusion", right? Yes: if a large enough fraction of New York Times' site's visitors use AdBlock (or ignore the ads using more old-fashioned means;)), the company won't be able to provide it to them anymore. Now, I do not claim that this is incontrovertible evidence, because there are lots of factors at play, but it is notable that "big media" is going through an absolutely brutal time in the past 10 years trying to figure out how to make ends meet.
It's a simple bargain, at the end of the day. You can think the content is useless, and not allow the owner to sell your attention to advertisers; or you may like the content, and then you either must put up with ads, or pay for it. Liking the content, but then proceeding to pay someone else to circumvent the ads is just perverse. If everyone does it, Pandora is gone. If just you do it, you're a free-loader.
Listening to music sans ads has value to you. You are naturally willing to pay up for the privilege. What baffles me is that you prefer to hand the resources over to a third party, which would filter the ads, rather than to the *content providers*, who would then be able to eschew the ads in the first place, and continue the service you enjoy!
Follow your choice to its logical conclusion. Pandora has costs and a perfectly legitimate desire to earn a rate of return on time and capital. They persuade advertisers that people who listen to their streams will buy the shilled products. You say, "bite me", and use a filter. Advertisers realise no one pays attention to their message on Pandora, and terminate their contracts. Pandora folds. You are left with a license for an ad blocker, and static in your headphones.... profit?
The presentation is light on details, and I haven't had time to poke around the researchers' websites, but, at first blush, I wonder whether the results have much to do with psychology per se? Rather, these guys have shown, in a round-about way, that the AoM "AI" is not very strong; in particular, that it's overly cautious and "leaves a lot on the table": given available resources, it could go on the offensive sooner than it does. That's why the "aggressive" and "neurotic" agents do so well against it. Playing AoM is a very complex dynamic programming problem, and it's anyone's guess what sort of objective function its authors have constructed, but now we can see that a fairly coarse re-weighting could significantly improve it. I don't think that the general take-way here ought to be that "neurotic" agents do well in strategic games (contract to the classic "tit-for-tat" repeated-games result).
That said, from the introduction of the presentation one can see that the real goal of this effort is to create bots that *people enjoy playing against*. That's probably only loosely related to the absolute strength of the opponent, and it makes complete sense that it would be thrilling to be up against an AI that can suddenly just "take a flyer" and surprise you.
There's been a lot of great work done on characterising extremophiles, and every time a new astonishing variety is discovered, someone (often not the authors themselves, in fairness) emphasises that this would allow them to survive on Mars, hard vacuum, etc. The problem is that unless you stretch the panspermia hypothesis (life is seeded by microbe-bearing ejecta from meteor impacts onto other, life-bearing planets) a long way, isn't the barrier to overcome not "microbe with 3 billion years to evolve here could survive on Mars under horrible conditions", but rather "under said horrible conditions, enough organic chemistry is possible for life to evolve in the first place"? The extremophile nature of an indigenous Martian life-form would then be a matter of course...
Touche as far Sun goes. But the meme has spread a lot wider, I think. As I said, it seems like a sensible idea, given that bandwidth is developing faster, in relative terms, than battery or display technology, two things one wants to think about before tossing out the desktop, let alone a datacenter, and relying on a PDA for one's daily bread...
It's rather odd that the end of the datacenter is supposed to be brought about by ubiquitous, small computers, since a few years ago everyone was looking forward to "thin clients" - these very same ubiquitous, small computers that would serve as *interfaces* to... the mighty, on-demand power of the datacenter. The latter vision still makes more sense to me, at least for the foreseeable future.
I do wonder whether the authors really "expected" the distribution of the numbers of readers to be exponential... I only follow this literature for curiosity's sake, but even so I've read quite a few papers lately finding power law distributions in various human communication networks (emails, letters, social groups), social animal groups, etc. The results describing power laws in various cuts of the Internet are also very well known. As some of the studies suggest, power laws arise in "bursty" communications, when the items involved are held in a queue, which organised by priority. For instance, if you respond to emails from a few special people very promptly, handle those from most others with a bit more proscrastination, and shelve a few for a very long time, the wait times between your communications will follow a power law.
In short, I bet that people working in the field would by now consider a power law the reasonable first hypothesis, when investigating a phenomenon of this sort. The mention of the refuted expected exponential is a bit of gentle scientific sensationalism.;)
[rant] I work in the finacial industry, and I know Excel/VBA/COM all too well. I also choose to do everything but the basic data acquisition and inspection using proper tools (Python and R, in my case). In fact, I take Excel's ubiquity as yet another piece of evidence that the majority of those toiling in the finance vineyard are numerically illiterate. The fundamental problem is that a spreadsheet conflates data and analytics, the cardinal sin in anything above a throw-away script; surely you know that, if you ever had to maintain a large non-trivial sheet. Another obvious flaw is that the spreadsheet is a Flatland, with nothing but 2D arrays (sure, writing VBA/VB or tacking on something more serious using COM resolves this problem, but why drag the ball-and-chain of Excel's baroque object model around to begin with?). Moving on to the actual implementation, Excel's Frankenstein nature, with all sorts of grafts, add-ons and arbitrary limitations always terrified me. Out of the essential applications we run here, this is the one that has someone pounding the table in frustration more often than everything else put together. In summary, Excel is horrendously overused, probably because it presents a seductive shallow-learning-curve alternative and traps people in a sub-optimal situation, where they spend time cobling workarounds together once the going gets tough, instead of doing real work. [/rant]
Aha, live and learn! I did not know that "intestine" meant that; I had originally surmised that the word you had been looking for had been "internecine", which, in my defense, would have fit in the context pretty well. Very nice; "intestinal power struggle" is still funny nonetheless.:)
I know scatological humour is declase, but the phrase "intestinal power struggle" is far too amusing to pass by! I mean, just try to visualise it! "The E coli have sallied forth from their line! They're overwhelming the oposition!". Or maybe that's a florid description of what happens after too much nachos and beer? Good stuff, either way...
Funny you mention Lamont. The problem is that it is still one of the busier libraries in the Harvard system precisely because undergraduates do need special facilities, and Lamont is a convenient way to aggregate them. Course packs are put on reserve there; the language and film lab is there; and, yes, large comfortable areas for study are there. And what's wrong with that? Whyever should I jog to Kummel for a geology reserve article and then run to Cabbot for a math packet? Yes, the depth of the Lamont collection is likely insufficient in one's particular major, but a good half of coursework in an undergraduate's curriculum is devoted with something else anyway. In other words, there are good reasons to have undergraduate libraries aside from the admitedly silly atavistic stuff about the sanctity of the main stacks.
I am not going to attempt to compare MSR's achievements with those of PARC or Bell Labs, partly because the latter two have been around for a lot longer. But here's a paper on the properties of NP-hard problems, published last week in Nature. Achliopats works in MSR, and so does Naor. These gentlemen are very much for real. So here we have a counter-example to your blanket claims; what are you going to do about it?
Touche, if you own the early machines. Not at all, if you manage to ever get a glimpse of, say, MP2K (which I own, thanks be to eBay). The handwritting recognition is superb! I also have a Visor Pro, and graffiti is a sad hack compared to what the Newton can manage. I have not yet had a chance to try out Inkwell on MacOS X, but it's supposed to be the progeny of the methodology used Newton, and I can imagine it has benefited from the additional computing power too.
When "chaos theory" (better to call it "field", or "approach to complex systems", maybe?) broke upon the world in 1970s, it's not like finance people yawned and ignored it! Au contraire, there was tremendous interest for the subsequent decade, as everyone searched for power laws, fractal dimensions and attractors. But now 30 years have passed, and the conclusion that was reached after research, and not in ignorance, as Mandelbrot suggests, is that financial markets are not predominantly chaotic. In other words, the "randomness" that we see in financial series is unlikely to be generated by underlying repeated actions of some simple nonlinear system; instead, it's really stochastic, really comes from un-anticipatable non-deterministic shocks. And if you really think about how the world works, with companies cheating and countries defaulting on debt, you'll find that intuitive, as well.
Here's one overview of the current state of the matter: Barnett and Serletis, 1998 (disclaimer: found after a 5 minute search of RePEC, there are likely even better papers).
I can't help but notice... First, two passages from the recently posted NY Times review:
"Clones" takes place 10 years after "Episode I -- The Phantom Menace," and it is as thick with exposition as an undergraduate history course.
Like weary Brezhnev-era Muscovites, the American moviegoing public will line up out of habit and compulsion, ruefully hoping that this episode will at least be a little better than the last one, and perhaps inwardly suspecting that the whole elephantine system is rotten.
Now take a look at Mr. Katz's blurb, where he opens up by describing the flick thus: "increasingly elephantine Skywalker saga" and "seems more like a graduate program." Coincidence, subliminal residue from a review he probably read 15 minutes ago, or something a bit more sinister?
WRONG! GSM, broadband Internet, and most of other things mentioned by the poster do indeed exist in Easter Europe, and some of them have actually propagated to Russia as well. And if progress can get there, one has to wonder even more about the USA...
Heh, as you can see, there is definitely a vocal cohort of Bose-haters out there...:) My personal opinion is just that a company that spends so much on the marketing has to cut some corners in production. But the good thing about sound is that one can always do a side-by-side comparason and choose what sounds better: it's a matter of personal preference.
Now here's my own student-on-a-budget rig that you can have for $300 and that will serve you quite well: a set of Grado SR-80 headphones, a Creek OBH-11 headphone amp and a nice chunky no-name interconnect for $25 (anything more would probably be redundant). Bose and Sony are not quite in the same league.
The above post is right. Check out Sherry Sontag's "Blind Man's Bluff", perhaps the most intriguing account of submarine spying during the Cold War around. She devoted a chapter to previously unpublished (this was 1996) stories about tapping of Russian communications cables (not fiberoptic, though) throughout the 80s. One was in the Pacific, by Ochotsk island, and the other one was in the North Atlantic, by the great Northern port of Murmansk. In the latter operations, the ship was manned only by volunteers, and it was rigged to explode on detection. Hairy stuff... It's quite incredible to what leghts Pentagon was willing to go to get a glimpse of Soviet military machine.
The level of danger from any radiation exposure depends on at least three things:
1. dose (the article talks about that)
2. area
3. duration
The fact of the matter is, whole body exposures are far more lethal and mutagenic, than concentrated ones. So, for instance, if you irradiate a person externally over most of his body with 6 grays, he has a high (something like 35% in Chernobyl) chance of dying. In radiation therapy, patients are often given 30Gy, to one spot, and do just fine (i.e. don't keel over). Note also, that different organs react very differently to radiation. As a rule of thumb, where cells divide often, radiation is most dangerous. Thus, skeletal muscle and such can deal with much more than core organs and reproductive organs. Check out Merck's page on the subject for a good initial overview (there are more nuances in reality, as always...:))
Well, this story is clearly a joke. But what I want to remark is that in truth lifetimes of impassioned research were sometimes driven by things akin to the search for the mathematics of "a soul's path to Heaven". You only have to remember Johannes Kepler, who devised the Laws of Planetary Motion (Astronomia Nova 1600-1609), which later became the basis for much of Newton's astrophysics. The funny thing is, Kepler was only pursuing a fancy that the planets in the Solar system are arranged in proportion to the classical Pythagorian hierarchy of the 5 fundamental polygons. Of course, this was only a pipe-dream, and the purported relationships merely accidental. Nevertheless, in 30 years of work, Kepler, using primitive pre-calculus mathematics, made one of the great advances in planetary physics ever. Nothing would sound strange to me after this...
Python, I believe, is as portable as Java. Check out the list of supported platforms. Lately, a port have been made for PalmOs, so the handheld advantage of Java has just diminished. The Python scripts can run on all kinds of machines with only minor modifications.
Slashdot Mirror
Your analogies are facile. Spam, telegraph operators ... It's the big media that arranges the generation of content. The foreign correspondents? The investigative journalists? The months spent on assignment abroad? The beat writers that follow a neighbourhood for years? If you think the echo chamber of inter-referencing bloggers will be a substitute, you've got another thing coming. Commenting on content is nice. Tagging it is great. Searching through it is wonderful. But someone's got to go and spent hundreds of hours on creating it first. And someone else has got to pay for it. At the moment, the guys who had the settled models in place are being driven out of business by outfits ultimately burning through venture capitalist money; eventually, they too will get around to charging for their services, and the era of "free" (i.e. subsidised) lunch will be over, but the damage will be done. Read up on what's happened to the staffs of LA Times, Chicago Tribune, Philadelphia Inquirer, etc., count the Pulitzers those papers had won, and compare that to the output of even the most precocious blog.
Spam is irrelevant to this matter. AdBlock is indeed a good analogue. You did note that I prefaced the second paragraph with "follow your choice to its logical conclusion", right? Yes: if a large enough fraction of New York Times' site's visitors use AdBlock (or ignore the ads using more old-fashioned means ;)), the company won't be able to provide it to them anymore. Now, I do not claim that this is incontrovertible evidence, because there are lots of factors at play, but it is notable that "big media" is going through an absolutely brutal time in the past 10 years trying to figure out how to make ends meet.
It's a simple bargain, at the end of the day. You can think the content is useless, and not allow the owner to sell your attention to advertisers; or you may like the content, and then you either must put up with ads, or pay for it. Liking the content, but then proceeding to pay someone else to circumvent the ads is just perverse. If everyone does it, Pandora is gone. If just you do it, you're a free-loader.
Listening to music sans ads has value to you. You are naturally willing to pay up for the privilege. What baffles me is that you prefer to hand the resources over to a third party, which would filter the ads, rather than to the *content providers*, who would then be able to eschew the ads in the first place, and continue the service you enjoy!
Follow your choice to its logical conclusion. Pandora has costs and a perfectly legitimate desire to earn a rate of return on time and capital. They persuade advertisers that people who listen to their streams will buy the shilled products. You say, "bite me", and use a filter. Advertisers realise no one pays attention to their message on Pandora, and terminate their contracts. Pandora folds. You are left with a license for an ad blocker, and static in your headphones. ... profit?
The presentation is light on details, and I haven't had time to poke around the researchers' websites, but, at first blush, I wonder whether the results have much to do with psychology per se? Rather, these guys have shown, in a round-about way, that the AoM "AI" is not very strong; in particular, that it's overly cautious and "leaves a lot on the table": given available resources, it could go on the offensive sooner than it does. That's why the "aggressive" and "neurotic" agents do so well against it. Playing AoM is a very complex dynamic programming problem, and it's anyone's guess what sort of objective function its authors have constructed, but now we can see that a fairly coarse re-weighting could significantly improve it. I don't think that the general take-way here ought to be that "neurotic" agents do well in strategic games (contract to the classic "tit-for-tat" repeated-games result).
That said, from the introduction of the presentation one can see that the real goal of this effort is to create bots that *people enjoy playing against*. That's probably only loosely related to the absolute strength of the opponent, and it makes complete sense that it would be thrilling to be up against an AI that can suddenly just "take a flyer" and surprise you.
There's been a lot of great work done on characterising extremophiles, and every time a new astonishing variety is discovered, someone (often not the authors themselves, in fairness) emphasises that this would allow them to survive on Mars, hard vacuum, etc. The problem is that unless you stretch the panspermia hypothesis (life is seeded by microbe-bearing ejecta from meteor impacts onto other, life-bearing planets) a long way, isn't the barrier to overcome not "microbe with 3 billion years to evolve here could survive on Mars under horrible conditions", but rather "under said horrible conditions, enough organic chemistry is possible for life to evolve in the first place"? The extremophile nature of an indigenous Martian life-form would then be a matter of course ...
Touche as far Sun goes. But the meme has spread a lot wider, I think. As I said, it seems like a sensible idea, given that bandwidth is developing faster, in relative terms, than battery or display technology, two things one wants to think about before tossing out the desktop, let alone a datacenter, and relying on a PDA for one's daily bread ...
It's rather odd that the end of the datacenter is supposed to be brought about by ubiquitous, small computers, since a few years ago everyone was looking forward to "thin clients" - these very same ubiquitous, small computers that would serve as *interfaces* to ... the mighty, on-demand power of the datacenter. The latter vision still makes more sense to me, at least for the foreseeable future.
I do wonder whether the authors really "expected" the distribution of the numbers of readers to be exponential ... I only follow this literature for curiosity's sake, but even so I've read quite a few papers lately finding power law distributions in various human communication networks (emails, letters, social groups), social animal groups, etc. The results describing power laws in various cuts of the Internet are also very well known. As some of the studies suggest, power laws arise in "bursty" communications, when the items involved are held in a queue, which organised by priority. For instance, if you respond to emails from a few special people very promptly, handle those from most others with a bit more proscrastination, and shelve a few for a very long time, the wait times between your communications will follow a power law.
;)
In short, I bet that people working in the field would by now consider a power law the reasonable first hypothesis, when investigating a phenomenon of this sort. The mention of the refuted expected exponential is a bit of gentle scientific sensationalism.
[rant]
I work in the finacial industry, and I know Excel/VBA/COM all too well. I also choose to do everything but the basic data acquisition and inspection using proper tools (Python and R, in my case). In fact, I take Excel's ubiquity as yet another piece of evidence that the majority of those toiling in the finance vineyard are numerically illiterate. The fundamental problem is that a spreadsheet conflates data and analytics, the cardinal sin in anything above a throw-away script; surely you know that, if you ever had to maintain a large non-trivial sheet. Another obvious flaw is that the spreadsheet is a Flatland, with nothing but 2D arrays (sure, writing VBA/VB or tacking on something more serious using COM resolves this problem, but why drag the ball-and-chain of Excel's baroque object model around to begin with?). Moving on to the actual implementation, Excel's Frankenstein nature, with all sorts of grafts, add-ons and arbitrary limitations always terrified me. Out of the essential applications we run here, this is the one that has someone pounding the table in frustration more often than everything else put together. In summary, Excel is horrendously overused, probably because it presents a seductive shallow-learning-curve alternative and traps people in a sub-optimal situation, where they spend time cobling workarounds together once the going gets tough, instead of doing real work.
[/rant]
Aha, live and learn! I did not know that "intestine" meant that; I had originally surmised that the word you had been looking for had been "internecine", which, in my defense, would have fit in the context pretty well. Very nice; "intestinal power struggle" is still funny nonetheless. :)
I know scatological humour is declase, but the phrase "intestinal power struggle" is far too amusing to pass by! I mean, just try to visualise it! "The E coli have sallied forth from their line! They're overwhelming the oposition!". Or maybe that's a florid description of what happens after too much nachos and beer? Good stuff, either way
Funny you mention Lamont. The problem is that it is still one of the busier libraries in the Harvard system precisely because undergraduates do need special facilities, and Lamont is a convenient way to aggregate them. Course packs are put on reserve there; the language and film lab is there; and, yes, large comfortable areas for study are there. And what's wrong with that? Whyever should I jog to Kummel for a geology reserve article and then run to Cabbot for a math packet? Yes, the depth of the Lamont collection is likely insufficient in one's particular major, but a good half of coursework in an undergraduate's curriculum is devoted with something else anyway. In other words, there are good reasons to have undergraduate libraries aside from the admitedly silly atavistic stuff about the sanctity of the main stacks.
I am not going to attempt to compare MSR's achievements with those of PARC or Bell Labs, partly because the latter two have been around for a lot longer. But here's a paper on the properties of NP-hard problems, published last week in Nature. Achliopats works in MSR, and so does Naor. These gentlemen are very much for real. So here we have a counter-example to your blanket claims; what are you going to do about it?
Touche, if you own the early machines. Not at all, if you manage to ever get a glimpse of, say, MP2K (which I own, thanks be to eBay). The handwritting recognition is superb! I also have a Visor Pro, and graffiti is a sad hack compared to what the Newton can manage. I have not yet had a chance to try out Inkwell on MacOS X, but it's supposed to be the progeny of the methodology used Newton, and I can imagine it has benefited from the additional computing power too.
When "chaos theory" (better to call it "field", or "approach to complex systems", maybe?) broke upon the world in 1970s, it's not like finance people yawned and ignored it! Au contraire, there was tremendous interest for the subsequent decade, as everyone searched for power laws, fractal dimensions and attractors. But now 30 years have passed, and the conclusion that was reached after research, and not in ignorance, as Mandelbrot suggests, is that financial markets are not predominantly chaotic. In other words, the "randomness" that we see in financial series is unlikely to be generated by underlying repeated actions of some simple nonlinear system; instead, it's really stochastic, really comes from un-anticipatable non-deterministic shocks. And if you really think about how the world works, with companies cheating and countries defaulting on debt, you'll find that intuitive, as well.
Here's one overview of the current state of the matter: Barnett and Serletis, 1998 (disclaimer: found after a 5 minute search of RePEC, there are likely even better papers).
I can't help but notice... First, two passages from the recently posted NY Times review:
"Clones" takes place 10 years after "Episode I -- The Phantom Menace," and it is as thick with exposition as an undergraduate history course.
Like weary Brezhnev-era Muscovites, the American moviegoing public will line up out of habit and compulsion, ruefully hoping that this episode will at least be a little better than the last one, and perhaps inwardly suspecting that the whole elephantine system is rotten.
Now take a look at Mr. Katz's blurb, where he opens up by describing the flick thus: "increasingly elephantine Skywalker saga" and "seems more like a graduate program." Coincidence, subliminal residue from a review he probably read 15 minutes ago, or something a bit more sinister?
WRONG! GSM, broadband Internet, and most of other things mentioned by the poster do indeed exist in Easter Europe, and some of them have actually propagated to Russia as well. And if progress can get there, one has to wonder even more about the USA ...
Now here's my own student-on-a-budget rig that you can have for $300 and that will serve you quite well: a set of Grado SR-80 headphones, a Creek OBH-11 headphone amp and a nice chunky no-name interconnect for $25 (anything more would probably be redundant). Bose and Sony are not quite in the same league.
The above post is right. Check out Sherry Sontag's "Blind Man's Bluff", perhaps the most intriguing account of submarine spying during the Cold War around. She devoted a chapter to previously unpublished (this was 1996) stories about tapping of Russian communications cables (not fiberoptic, though) throughout the 80s. One was in the Pacific, by Ochotsk island, and the other one was in the North Atlantic, by the great Northern port of Murmansk. In the latter operations, the ship was manned only by volunteers, and it was rigged to explode on detection. Hairy stuff ... It's quite incredible to what leghts Pentagon was willing to go to get a glimpse of Soviet military machine.
The level of danger from any radiation exposure depends on at least three things: 1. dose (the article talks about that) 2. area 3. duration The fact of the matter is, whole body exposures are far more lethal and mutagenic, than concentrated ones. So, for instance, if you irradiate a person externally over most of his body with 6 grays, he has a high (something like 35% in Chernobyl) chance of dying. In radiation therapy, patients are often given 30Gy, to one spot, and do just fine (i.e. don't keel over). Note also, that different organs react very differently to radiation. As a rule of thumb, where cells divide often, radiation is most dangerous. Thus, skeletal muscle and such can deal with much more than core organs and reproductive organs. Check out Merck's page on the subject for a good initial overview (there are more nuances in reality, as always ... :))
Well, this story is clearly a joke. But what I want to remark is that in truth lifetimes of impassioned research were sometimes driven by things akin to the search for the mathematics of "a soul's path to Heaven". You only have to remember Johannes Kepler, who devised the Laws of Planetary Motion (Astronomia Nova 1600-1609), which later became the basis for much of Newton's astrophysics. The funny thing is, Kepler was only pursuing a fancy that the planets in the Solar system are arranged in proportion to the classical Pythagorian hierarchy of the 5 fundamental polygons. Of course, this was only a pipe-dream, and the purported relationships merely accidental. Nevertheless, in 30 years of work, Kepler, using primitive pre-calculus mathematics, made one of the great advances in planetary physics ever. Nothing would sound strange to me after this ...
Python, I believe, is as portable as Java. Check out the list of supported platforms. Lately, a port have been made for PalmOs, so the handheld advantage of Java has just diminished. The Python scripts can run on all kinds of machines with only minor modifications.