If you're judging SaaS by the performance of M$ or if your opinions are driven by sensational media coverage and highly visible outages like Skype then you're incapable of sound judgment. Even if this weren't a worthless ad hominem attack, it would be an overly broad generalization from the premises.
There have been constant small and spectacular meltdowns by IS shops all over the planet but they don't get noticed by the press. I'd much rather trust my stuff to the grid and the "Googleplex" than the average IT shop. It's like more people are killed by lightning than by tsunamis, tornadoes, hurricanes, volcanoes, and earthquakes but the most attention goes to the latest big disaster. There's more to trust than just availability- like, oh, security. But go ahead and trust the grid- there's no one snooping, sniffing, or sending your information to the NSA.
The collective stability of the net/grid and Google's 500,000+ servers are going to bring irresistible disruptive change. Those who do not adjust to this trend will become extinct relics. Why buy a CD when you can just tap the grid for an MP3? Because I want the whole album/ lyric sheets/ art/ backup media.
Why make a trip to the store to get a DVD when you can just download it from the net in seconds? Because I don't have that kind of broadband/ live in the country/ want backup media.
Why go to the movie theatre when you can just watch it on your own home theatre which has better quality video and audio? Because I can't afford home theater/ can't fit all of the people I want to watch it with into my apartment/ actually like the communal experience of some movies. Most vividly, I remember an awesome moment in Se7en when the entire theater screamed, and that's something that I will not likely experience at home.
Why bother with installing and constantly upgrading software when you can just use it on the Net? Because what constitutes a real upgrade depends on who you talk to. The content publisher might like a new ribbon interface that requires you to relearn where everything is. The content publisher might like to make it impossible for you to use what you bought in any way but the way they would prefer, because they're in bed with other content distributors.
Why worry about backups when Google makes it routine and does it for you so you don't even have to think about it? Because if/when they do fail, everyone else will be screwed too. Do you want to be dealing with a disaster at the same time that everyone else is?!?! Imagine your exciting world where everyone's totally jacked into the net. Cool, huh? Except now, when the big guys fail, it's utterly catastrophic. The thousands of minor failures experienced all over the world are and will be a multitude of cuts and abrasions on society, but if everyone is dependent on one thing that fails, that's a gunshot wound. Which would you rather try to recover from?
It's appalling that so many people lack vision and are blinded by the here and now. Visions differ. Just because someone doesn't see what you see, doesn't mean they're not looking. I imagine that by distributing effort rather than centralizing it, you give up some efficiency for robustness, and that's worth it sometimes.
A few years ago, I wrote an album using sounds generated within Matlab. The idea was to produce an album that was as entirely original as I could- not using any recorded sounds, and not using synthetic sounds that I had not created myself with my own algorithms.
When it came to mixing the album, I adjusted things as best I could, but I had no background along those lines. I got feedback from my friends that the loud portions were too loud and the quiet portions were too quiet. But I didn't know to what degree the audio should be compressed. I was at square one.
I took a cross-section of tracks from my ripped CD library and measured their peak level and RMS level. Having this information would tell me what people would be used to. Unfortunately, the only consistent pattern that I found was that the higher the RMS level, the later the release date of the CD.:(
One day far in the future, we will start up our chess programs and they will immediately announce "Mate in 326". A "good" move will be one that hastens the loss by as little as possible.
How do I know? Because the exact same maths apply to a different domain, and we'd already have seen developments there if this was true.
Decompilation uses exactly the same abstract mathematical concepts as shape recognition (of which facial recognition clearly is a subset). Just replace "vertices" with assembly-language instructions and the "shapes" to which they may belong with program structures (for / while loops, subroutines &c). Let's pretend for the moment that decompilation used the same math as facial recognition (and I find that idea suspect).
Same math + different problem sizes = differing feasibility given fixed computing resources.
You can resample images down to a feasible resolution prior to applying a face recognition algorithm. You cannot do that same trick to a program.
It is not possible to describe such things (let alone voltage islands, voltage scaling) in an HDL language and they must either be a special feature built into synthesis (with an extra set of constraints) or done by hand at the transistor/gate level.
Verilog is perfectly capable of describing designs at the gate level. Just last semester I designed a simple 16-bit RISC processor at the gate level (with the exception of flip-flops, given to us as black boxes) using Verilog.
Verilog is a huge language. While it may be the case that some of its constructs assume a clock (I can't say- I worked with a tiny subset), the language itself does not assume a clock and is capable of dealing with asynchronous circuits.
But can anybody explain to me why this is any different from, say, somebody who loves playing soccer, or playing piano?
From a diagnostic perspective: The DSM consistently takes the perspective that a behavior is disordered if it is causing problems in other areas of your life.
From a biological perspective: Not all pleasure is created equal. There are reward circuits in our brains that are dedicated to helping us be social and bond with others. Occasionally, some people will experience pleasures (e.g. heroin) that start to hijack those circuits so they no longer respond to social stimuli- over time, those circuits respond preferentially, and then exclusively, to the new, addictive stimulus. This is not the way that most people respond to most pleasurable stimuli (e.g. playing piano).
I am not trying to imply that video games are indeed such an addictive stimulus. I mainly mean to point out that there are a few different (but phenomenologically convergent) ways of recognizing qualitative differences between pleasurable activities- differences between "addictive" and "non-addictive" stimuli. IMHO, for some people, video games do hijack the social circuitry, but I don't think it works that way for everyone.
I come from a large family. I have 4 siblings with Crohn's. I have 5 siblings with diabetes. There is one sibling with neither.
I just recently got Crohn's to even out the teams.
If Crohn's and diabetes were not genetically related, it would be very weird to not have any overlap between the group of siblings with Crohn's and the siblings with diabetes (if the probabilities had been independent, then someone should've gotten both). So it makes sense to me to hear that Crohn's and diabetes would share some genetic basis. Perhaps, then, different alleles of (an)other gene(s) could control which disease ends up expressed in the phenotype.
The thrust of Anderson's argument is that we should limit what is available to the general population because certain unstable or anti-social personalities might be pushed over the edge by them. He readily admits the influence is marginal and insignificant for the vast majority of the population (even he isn't contesting this--I'm not sure why you are.)
I was not making any statement specific to video games or violence. I'm making a statement about what can be sensibly said about accounting for variance. It makes no sense to speak of 90% of the variance in behavior, unless you are talking about a specific measure of behavior (but you were explicitly not doing so, by mentioning seemingly unrelated behavior such as accents).
In order to make a claim about behavior as broad as yours, one would have to have access to a meta-analysis (and it would be hard to imagine a meta-analysis of the scope required). Since no such meta-analysis was referenced or even hinted at in the original comment, I concluded that you were not using a meta-analysis as the basis of your statement.
On a note more specific to the content of your claims, to say that genes and peer groups take 90% of the behavior-variance-cake requires us to enter those into the analysis first. It is not hard at all to imagine that parenting practices or at least parental SES (which would affect what schools the child attends) would have an influence on the child's peer groups. The correlated-predictors problem would then rear its head when attempting to use genes and peer groups to explain variance in behavior data.
First of all, this is precisely my point with regard to Anderson's research. The predictor variables are linked for aggression. The kind of person who likes Myst (and will therefore play it a lot if assigned it) is different from the kind of person who plays Quake (and will play it a lot if assigned it.) They will also react to the games in different ways. Again, the conclusion is that people who enjoy Myst are different from those who enjoy Quake.
I'm not sure about the entirety of Anderson's work, but the paper linked to in the article (Anderson and Dill, 2000), contained two studies- one correlational and one experimental. The experimental study was described by Slate in a way that may have misled people- "...when the students went on to play a second game, the Wolfenstein 3-D players were more likely to behave aggressively toward losing opponents." Slate makes it sound as though the experimenters allowed players to choose whether they would play a second game or not, and those people that did play a second Wolfenstein game were more aggressive than those who played a second Myst game. But reading the original paper shows that all of the players (regardless of their predispositions or preferences) played two games- and the ones that played the Wolfenstein exhibited the more aggressive behavior on a later competitive task. So no, there is no issue of correlated predictors here.
Lastly, I apologize for the terrible formatting of my above post. It was my first at slashdot, and was made without previewing first.
As someone about to finish their Psych and CS double-major, I have long been irritated by this "debate". Everyone I've talked to in our Psych department thinks "Well, duh, of course videogames are a risk factor for violence." and everyone I've talked to from CS says "Well, duh, of course videogames don't make people violent- you don't see me going off and killing everyone, do you?" It's not a discussion as long as people are married to their own viewpoints.
Of course, the *real* question is how much this matters. Even if there were a well-controlled, randomized study showing that the amount of game time played directly correlated with the likelihood of a violent crime, is that enough cause to ban games?
My own viewpoint is that we should go out there and find the truth. A lot of different interests are going to interpret the truth differently depending on what kinds of decisions they have power over. For example, one might look at this issue and say "as a voter, I want politicians to lay off my video games, how ever this turns out." But there are other filters through which people might decide the relevance of the issue- for example, I want to have kids someday, and if GTA is going to make it more likely that they'll have behavior problems, then I'd like to know that.
In general society, there may have been something other than the prevalence of videogames that has changed between then and now. That's what experiments are for- to try to set up an environment that ensures that the only systematic differences between the different groups you want to compare are those predictors that are interesting to you.
Recent research into human behaviour finds too main causal factors: genetic predisposition (measured in twin studies), and peer influence (for example, why do children speak with the accent of their peers and not that of their parents.) These probably account for as much as 90% of variance. The remaining 10% includes parent, teachers, life experience, and all media. So how much influence is left for video games? Not a lot. Behavior is not a scalar quantity, so it does not have scalar variance. One can devise scalar measures for particular aspects of behavior, and it may then be possible to account for the variance in those measures, but it is completely nonsensical to speak of accounting for 90% of variance in behavior overall. This much, any psychology researcher worth their salt could tell you.
But let's pretend you have a scalar measure of behavior and you want to start accounting for the variance in your measurements to develop explanations for where the variance "comes from". It's a lot trickier than you think. Let's say we have a data set with a bunch of variables. How much variance in, say, lung capacity does alcohol consumption account for? That depends. You could just look at their correlation (r squared) and take the square root. Or you could "partial out" other variables first (like cigarette smoking) using partial correlations or stepwise regression. The amount of variance accounted for by a particular variable depends on what other variables you're taking into account. When our predictor variables are linked, the order of predictors in the regression makes a big difference (what if we had data on time spent at bars, which would be correlated with secondhand smoke and with alcohol consumption?)
In short, to say that two variables account for 90% of the variance in behavior is simplistic at best and nonsensical at worst.
Slashdot Mirror
Touché.
Hmm... while many programmers are powered by java, all life on Earth is powered at least indirectly by the Sun.
A few years ago, I wrote an album using sounds generated within Matlab. The idea was to produce an album that was as entirely original as I could- not using any recorded sounds, and not using synthetic sounds that I had not created myself with my own algorithms.
:(
When it came to mixing the album, I adjusted things as best I could, but I had no background along those lines. I got feedback from my friends that the loud portions were too loud and the quiet portions were too quiet. But I didn't know to what degree the audio should be compressed. I was at square one.
I took a cross-section of tracks from my ripped CD library and measured their peak level and RMS level. Having this information would tell me what people would be used to. Unfortunately, the only consistent pattern that I found was that the higher the RMS level, the later the release date of the CD.
One day far in the future, we will start up our chess programs and they will immediately announce "Mate in 326". A "good" move will be one that hastens the loss by as little as possible.
Decompilation uses exactly the same abstract mathematical concepts as shape recognition (of which facial recognition clearly is a subset). Just replace "vertices" with assembly-language instructions and the "shapes" to which they may belong with program structures (for / while loops, subroutines &c). Let's pretend for the moment that decompilation used the same math as facial recognition (and I find that idea suspect).
Same math + different problem sizes = differing feasibility given fixed computing resources.
You can resample images down to a feasible resolution prior to applying a face recognition algorithm. You cannot do that same trick to a program.
The business case would become irrelevant if use of equipment like this was mandated for ISPs by governments.
Even then, you can turn those sticky keys off by pressing shift 5 times.
Verilog is perfectly capable of describing designs at the gate level. Just last semester I designed a simple 16-bit RISC processor at the gate level (with the exception of flip-flops, given to us as black boxes) using Verilog.
Verilog is a huge language. While it may be the case that some of its constructs assume a clock (I can't say- I worked with a tiny subset), the language itself does not assume a clock and is capable of dealing with asynchronous circuits.
From a diagnostic perspective: The DSM consistently takes the perspective that a behavior is disordered if it is causing problems in other areas of your life.
From a biological perspective: Not all pleasure is created equal. There are reward circuits in our brains that are dedicated to helping us be social and bond with others. Occasionally, some people will experience pleasures (e.g. heroin) that start to hijack those circuits so they no longer respond to social stimuli- over time, those circuits respond preferentially, and then exclusively, to the new, addictive stimulus. This is not the way that most people respond to most pleasurable stimuli (e.g. playing piano).
I am not trying to imply that video games are indeed such an addictive stimulus. I mainly mean to point out that there are a few different (but phenomenologically convergent) ways of recognizing qualitative differences between pleasurable activities- differences between "addictive" and "non-addictive" stimuli. IMHO, for some people, video games do hijack the social circuitry, but I don't think it works that way for everyone.
I come from a large family. I have 4 siblings with Crohn's. I have 5 siblings with diabetes. There is one sibling with neither.
I just recently got Crohn's to even out the teams.
If Crohn's and diabetes were not genetically related, it would be very weird to not have any overlap between the group of siblings with Crohn's and the siblings with diabetes (if the probabilities had been independent, then someone should've gotten both). So it makes sense to me to hear that Crohn's and diabetes would share some genetic basis. Perhaps, then, different alleles of (an)other gene(s) could control which disease ends up expressed in the phenotype.
Hmmph. Apparently hardcore Battlelords fans are being left out of this.
I was not making any statement specific to video games or violence. I'm making a statement about what can be sensibly said about accounting for variance. It makes no sense to speak of 90% of the variance in behavior, unless you are talking about a specific measure of behavior (but you were explicitly not doing so, by mentioning seemingly unrelated behavior such as accents).
In order to make a claim about behavior as broad as yours, one would have to have access to a meta-analysis (and it would be hard to imagine a meta-analysis of the scope required). Since no such meta-analysis was referenced or even hinted at in the original comment, I concluded that you were not using a meta-analysis as the basis of your statement.
On a note more specific to the content of your claims, to say that genes and peer groups take 90% of the behavior-variance-cake requires us to enter those into the analysis first. It is not hard at all to imagine that parenting practices or at least parental SES (which would affect what schools the child attends) would have an influence on the child's peer groups. The correlated-predictors problem would then rear its head when attempting to use genes and peer groups to explain variance in behavior data.
First of all, this is precisely my point with regard to Anderson's research. The predictor variables are linked for aggression. The kind of person who likes Myst (and will therefore play it a lot if assigned it) is different from the kind of person who plays Quake (and will play it a lot if assigned it.) They will also react to the games in different ways. Again, the conclusion is that people who enjoy Myst are different from those who enjoy Quake.
I'm not sure about the entirety of Anderson's work, but the paper linked to in the article (Anderson and Dill, 2000), contained two studies- one correlational and one experimental. The experimental study was described by Slate in a way that may have misled people- "...when the students went on to play a second game, the Wolfenstein 3-D players were more likely to behave aggressively toward losing opponents." Slate makes it sound as though the experimenters allowed players to choose whether they would play a second game or not, and those people that did play a second Wolfenstein game were more aggressive than those who played a second Myst game. But reading the original paper shows that all of the players (regardless of their predispositions or preferences) played two games- and the ones that played the Wolfenstein exhibited the more aggressive behavior on a later competitive task. So no, there is no issue of correlated predictors here.
Lastly, I apologize for the terrible formatting of my above post. It was my first at slashdot, and was made without previewing first.
As someone about to finish their Psych and CS double-major, I have long been irritated by this "debate". Everyone I've talked to in our Psych department thinks "Well, duh, of course videogames are a risk factor for violence." and everyone I've talked to from CS says "Well, duh, of course videogames don't make people violent- you don't see me going off and killing everyone, do you?" It's not a discussion as long as people are married to their own viewpoints.
Of course, the *real* question is how much this matters. Even if there were a well-controlled, randomized study showing that the amount of game time played directly correlated with the likelihood of a violent crime, is that enough cause to ban games?My own viewpoint is that we should go out there and find the truth. A lot of different interests are going to interpret the truth differently depending on what kinds of decisions they have power over. For example, one might look at this issue and say "as a voter, I want politicians to lay off my video games, how ever this turns out." But there are other filters through which people might decide the relevance of the issue- for example, I want to have kids someday, and if GTA is going to make it more likely that they'll have behavior problems, then I'd like to know that.
In general society, there may have been something other than the prevalence of videogames that has changed between then and now. That's what experiments are for- to try to set up an environment that ensures that the only systematic differences between the different groups you want to compare are those predictors that are interesting to you.