Yeah, but then we have the quite reasonable fallback: extraordinary claims require extraordinary evidence.
A completely physically-naive person would look at hard drive capacities from 1960-1990 and extrapolate a trend to infinite capacity. Physicists and the hyper-specialized niche of hard-drive boffins know better.
Now, which one does Kurzweil sound more like?
The thing about stat 101 is that anyone can be wrong based on it! Statistics should be called "numerical skepticism".
Singularitarians, being primarily hopeful, lose out on the selling point of the nearly-defunct Christianity: there's no skin off my back if I'm wrong. Since I'm a technical person, I'm contributing to the singularity if it's going to happen; and merely making a living if it's not. There's no hell for me to go to, so I don't see the need to believe in a Miracle.
(In case the last paragraph is misleading, I do what I do for the love of it, not for the money.)
When I come up with an idea for an inequality, or even just a probabilistic model, I can whack together a few lines of R and see how it works. Usually almost instantly! Basic programming (i.e. work) literally replaces "genius", and I don't need to think like Hardy and Littlewood, or Fisher, or Feynman to get really good hints as to why some ideas of mine are good, or bad. They say Feynman solved PDEs in his head by visualizing the solutions. For the rest of us, there's a computer. That's a "force multiplier" of possibly inconceivable magnitude.
Of course, most people don't do this with a computer. That's probably a pity. But the point is, the computer (along with basic programming skills) turns motivation, curiosity, and insight into results without the need for the mystical catalyst of "genius".
On the other hand, these results are infinitely more modest than The Singularity and, honestly, will likely stay that way. Really, since we can't even agree as a species whether we cause global warming, I'm not holding out hope that we're looking at technological utopia in a few decades.
If I were to believe in its feasibility, I'd rather take my fantasy tempered by reality... I'd go with Hugo de Garis, who predicts a gigadeath war between AI and its supporters, against the reactionaries. There would be a hell of a lot of people who, once computers start replacing us, will have a rather dark (but not totally unjustified) idea about what would come next...
I think the idea is that people with $25K go to Singularity University in order to "learn" how to spend their money on more singularitarian bullshit.
Any place of learning, from high school through community college and up to grad school, is Singularity University. Hint: take math and science classes. I think I'd rather take linear algebra and diff eq. at a community college than pay $25K to hear a blowhard's dream for the future. Hell, if you take a decent statistics class you can outsmart these guys by learning about what's wrong with extrapolating a fitted curve past its support is not valid...
What I have noticed is that these days, any mention made of an income disparity at all receives a knee-jerk accusation of communism. Even when the mentioner explicitly disdains the idea of any social programs or income adjustment, and is quite willing to accept his struggle.
This is extremely disturbing to me... The end of civilization comes when the elites are too weak and ashamed to even accept the fact that there is an under-class.
Did you know that the Brothers Karamazov was originally written as a series for a newspaper over the course of months? I wondered for a while, why the chapters seemed to start with a recap of the previous one.
Anyway, agreed, it's great. I've noticed though, that in contrast my favorite science fiction tends to be very weak on characterization (e.g. Vinge) and more focused on entire societies and their conflicts and trajectories. A "systems view" if you will.
Haven't watched Lost myself, it sounds like gratuitous mysteries for commercial utility. At least David Lynch did it for pure, honest self-indulgence.:)
Wow, I've been making this drink for myself for years. Quite enjoyable as you say, although I wouldn't use anything but three buck chuck, or box wine. I didn't know it had a name.
C128: You have to be in 80-column mode (which was really impressive to me back then...) to use fast; the (separate) chip for 40-column couldn't operate at 2MhZ, and thus would disable. (The "slow" mode was 1MhZ btw.)
Turbo: I think that the turbo button was there for compatibility with old software. Software wasn't always written with explicit timing routines, so usually a game written for an 8MhZ machine would be unplayable at 25MhZ. There were some hacks to slow down software by introducing interrupts, but this of course wasn't as reliable as getting the hardware speed closer.
As software started to be written with evolving hardware in mind, the turbo button became superfluous. (Today's turbo button is "installing a graphics card";-)
Your description is vague (perhaps intentionally so), but I'm skeptical nonetheless.
The persistence-of-vision hurdle is easily jumped, by tuning a decay function to interpolate across the animated gif so that it looks like the appropriate single frame. Note, this only has to be done once.
This leaves the optical illusions. Again, there are really only so many of these, and they can be pattern-recognized and classified as whatever they represent. You can stick them together in any combination but this just adds a segmenting problem. Both of these problems have already been solved for standard captchas (where there is the extra problem of those lines connecting letters, in order to make the segmentation harder).
These kinds of considerations aren't anything new, and injecting them into soft sci-fi like Back to the Future is a waste of time. BttF is enjoyable, though, and does make a great accidental (?) satire of the American dream and hubris. For science, read some Larry Niven or Stanislaw Lem instead.
For example, Vernor Vinge did something like this, involving teleportation. A teleporter could control both the outcome position and velocity, but velocity was "harder" and took effort proportional to the difference in velocity.
Therefore, long distance teleports were only feasible along a longitude, and to the opposite latitude, since you had to match momentum or die by either being crushed or flung off into space. The earth's spin matches at lat X long Y, and lat -X long Y, but nowhere else.
As a result, one of the world's superpowers controlled both semi-polar regions, alternating by season; while the other stuck to the equator.
Unless the kind of man who sticks his peen into multiple women, tends to find different partners than those who don't...
What's most noteworthy is that there are two settings for male behavior in the simulator. Each of them is totally unrealistic, and they give completely different results.... So, what am I supposed to learn here?
Not to mention that there's nothing too new about the results, and somehow people in the 60s understood them without a java applet. Basically what happens is that promiscuous women implicitly quarantine the men who find them. It just so happens, if I understand correctly, that there is a portion of the phase-space, where increasing the rate of promiscuity in women serves to reduce the overall rate of disease by concentrating the disease among the promiscuous while keeping men from sleeping with the unpromiscuous women.
Now, if that sounds like a realistic description of the world to you...
In re your first paragraph: I can only suppose you're a Christian. Would you actually be happier if they specifically invoked your lord Christ for the sole purpose of getting you to spend money on... petty material frivolities? Seriously?
In re your second paragraph: The situation is not as symmetric as you make it seem. Hint: schools don't have solstice parties, and most students and faculty likely wouldn't even know what it was. That's just a strawman you pulled out of a hat.
Whatever your opinion is of the legitimacy of tax revenue, you can treat them as "gross revenue" just like any corporation and the computations all work out. It's not like economic models are perfect for non-state corporations anyway.
The S language is a free spec developed at Bell Labs. According to my professor, the statistics community was generally shocked and appalled to find that they had entered an exclusive licensing deal with Insightful Corp., who make the commercial implementation S-Plus. Before this, S was mostly open-source although the term didn't exist at the time.
So in a way, R is just putting things back to the way they were originally, about 30 years ago.
If they don't, it means that use and development of advanced statistical methods in finance, are as much smoke-and-mirrors as the rest of the finance "industry".
No free RPro for linux, and only binary installers for Windows and Mac. Please explain to me how you aren't violating the GPLv2, which is R's license, by doing the latter. (The former is just annoying, not illegal.) Since you refer to your product as an "enhanced distribution of open-source R", I can only assume you didn't recode it from scratch (which would be quite an undertaking).
Since it's open source, I suppose what I need to do to get a linux version, is find someone who has bought the "server" version and get a copy from them... surely you wouldn't have a problem with this, yes?
I remember my intern officemate once talking disrespectfully about some girl at a stripjoint. I don't have anything against this in-and-of-itself, as it is part of the implied service a stripper provides. However, he was in dire need of an Outlook Adjustment for a few other reasons...
So I told him how much the strippers at that establishment make (most of it under the table) for what is basically a part-time job, and he got real quiet. Then I asked him, since I knew he was under the same "patent transfer" arrangement I was, what exactly the difference was between selling your body and selling your mind, apart from the fact that you'll never get your ideas back, ever. He didn't talk to me much after that. Some people can't handle the truth I guess.
Back on-topic (sort of): if Steven Levitt (the Freakonomics guy) is to be trusted, most "foot soldiers" in the drug trade actually do pull a second job at McDonalds. He tells the story that they are similarly undignified jobs, except that one of them incurs a much larger probability of dying: http://www.ted.com/index.php/talks/steven_levitt_analyzes_crack_economics.html
If you plot those points, you'll still see a trend. If you want to ignore your senses and call it an illusion, you can try computing the regression yourself and find that p=0.005, r=-0.81. Even if you take out Matt Smith you still have p=0.03, r=-0.72.
The trend is pretty obvious if you just plot the data. I took the opportunity to learn a little bit of google's painful chart API (no criticism please; I see how it could be prettier and I don't care):
Nonetheless, there is an overall trend to younger doctors. Correlation of age with order of appearance: r=-0.7, with each doctor an average of 2 years younger than the last. (p-value: 0.01)
Without Matt Smith, that goes down to r=-0.6, and 1.7 years younger. (p-value: 0.04)
Slashdot Mirror
Yeah, but then we have the quite reasonable fallback: extraordinary claims require extraordinary evidence.
A completely physically-naive person would look at hard drive capacities from 1960-1990 and extrapolate a trend to infinite capacity. Physicists and the hyper-specialized niche of hard-drive boffins know better.
Now, which one does Kurzweil sound more like?
The thing about stat 101 is that anyone can be wrong based on it! Statistics should be called "numerical skepticism".
Singularitarians, being primarily hopeful, lose out on the selling point of the nearly-defunct Christianity: there's no skin off my back if I'm wrong. Since I'm a technical person, I'm contributing to the singularity if it's going to happen; and merely making a living if it's not. There's no hell for me to go to, so I don't see the need to believe in a Miracle.
(In case the last paragraph is misleading, I do what I do for the love of it, not for the money.)
The hell it isn't.
When I come up with an idea for an inequality, or even just a probabilistic model, I can whack together a few lines of R and see how it works. Usually almost instantly! Basic programming (i.e. work) literally replaces "genius", and I don't need to think like Hardy and Littlewood, or Fisher, or Feynman to get really good hints as to why some ideas of mine are good, or bad. They say Feynman solved PDEs in his head by visualizing the solutions. For the rest of us, there's a computer. That's a "force multiplier" of possibly inconceivable magnitude.
Of course, most people don't do this with a computer. That's probably a pity. But the point is, the computer (along with basic programming skills) turns motivation, curiosity, and insight into results without the need for the mystical catalyst of "genius".
On the other hand, these results are infinitely more modest than The Singularity and, honestly, will likely stay that way. Really, since we can't even agree as a species whether we cause global warming, I'm not holding out hope that we're looking at technological utopia in a few decades.
If I were to believe in its feasibility, I'd rather take my fantasy tempered by reality... I'd go with Hugo de Garis, who predicts a gigadeath war between AI and its supporters, against the reactionaries. There would be a hell of a lot of people who, once computers start replacing us, will have a rather dark (but not totally unjustified) idea about what would come next...
I think the idea is that people with $25K go to Singularity University in order to "learn" how to spend their money on more singularitarian bullshit.
Any place of learning, from high school through community college and up to grad school, is Singularity University. Hint: take math and science classes. I think I'd rather take linear algebra and diff eq. at a community college than pay $25K to hear a blowhard's dream for the future. Hell, if you take a decent statistics class you can outsmart these guys by learning about what's wrong with extrapolating a fitted curve past its support is not valid...
What I have noticed is that these days, any mention made of an income disparity at all receives a knee-jerk accusation of communism. Even when the mentioner explicitly disdains the idea of any social programs or income adjustment, and is quite willing to accept his struggle.
This is extremely disturbing to me... The end of civilization comes when the elites are too weak and ashamed to even accept the fact that there is an under-class.
If that's what you got out of the parent post, it says a hell of a lot more about you than the poster.
Did you know that the Brothers Karamazov was originally written as a series for a newspaper over the course of months? I wondered for a while, why the chapters seemed to start with a recap of the previous one.
Anyway, agreed, it's great. I've noticed though, that in contrast my favorite science fiction tends to be very weak on characterization (e.g. Vinge) and more focused on entire societies and their conflicts and trajectories. A "systems view" if you will.
Haven't watched Lost myself, it sounds like gratuitous mysteries for commercial utility. At least David Lynch did it for pure, honest self-indulgence. :)
Wow, I've been making this drink for myself for years. Quite enjoyable as you say, although I wouldn't use anything but three buck chuck, or box wine. I didn't know it had a name.
Credit bureaus have long tentacles...
What's interesting to me, is that the three of them have slight discrepancies between them, which means that they actually "compete" in some way.
C128: You have to be in 80-column mode (which was really impressive to me back then...) to use fast; the (separate) chip for 40-column couldn't operate at 2MhZ, and thus would disable. (The "slow" mode was 1MhZ btw.)
Turbo: I think that the turbo button was there for compatibility with old software. Software wasn't always written with explicit timing routines, so usually a game written for an 8MhZ machine would be unplayable at 25MhZ. There were some hacks to slow down software by introducing interrupts, but this of course wasn't as reliable as getting the hardware speed closer.
As software started to be written with evolving hardware in mind, the turbo button became superfluous. (Today's turbo button is "installing a graphics card" ;-)
Your description is vague (perhaps intentionally so), but I'm skeptical nonetheless.
The persistence-of-vision hurdle is easily jumped, by tuning a decay function to interpolate across the animated gif so that it looks like the appropriate single frame. Note, this only has to be done once.
This leaves the optical illusions. Again, there are really only so many of these, and they can be pattern-recognized and classified as whatever they represent. You can stick them together in any combination but this just adds a segmenting problem. Both of these problems have already been solved for standard captchas (where there is the extra problem of those lines connecting letters, in order to make the segmentation harder).
These kinds of considerations aren't anything new, and injecting them into soft sci-fi like Back to the Future is a waste of time. BttF is enjoyable, though, and does make a great accidental (?) satire of the American dream and hubris. For science, read some Larry Niven or Stanislaw Lem instead.
For example, Vernor Vinge did something like this, involving teleportation. A teleporter could control both the outcome position and velocity, but velocity was "harder" and took effort proportional to the difference in velocity.
Therefore, long distance teleports were only feasible along a longitude, and to the opposite latitude, since you had to match momentum or die by either being crushed or flung off into space. The earth's spin matches at lat X long Y, and lat -X long Y, but nowhere else.
As a result, one of the world's superpowers controlled both semi-polar regions, alternating by season; while the other stuck to the equator.
You take it as a foregone conclusion that video games influence crime, but then ask "Were they simply the victim of horrible parenting?"
So, which one is it? It's good to doubt both sides as long as reasonable, but as long as we do that, let's not come to any conclusions.
Or, it is worth that much but the money is going to the wrong people...
The GPLv2 license, unless modified, includes a clause allowing re-licensing to any later version of the GPL...
Unless the kind of man who sticks his peen into multiple women, tends to find different partners than those who don't...
What's most noteworthy is that there are two settings for male behavior in the simulator. Each of them is totally unrealistic, and they give completely different results. ... So, what am I supposed to learn here?
Not to mention that there's nothing too new about the results, and somehow people in the 60s understood them without a java applet. Basically what happens is that promiscuous women implicitly quarantine the men who find them. It just so happens, if I understand correctly, that there is a portion of the phase-space, where increasing the rate of promiscuity in women serves to reduce the overall rate of disease by concentrating the disease among the promiscuous while keeping men from sleeping with the unpromiscuous women.
Now, if that sounds like a realistic description of the world to you...
Join our fraternity! We have both kinds of people: scumbags and assholes.
In re your first paragraph: I can only suppose you're a Christian. Would you actually be happier if they specifically invoked your lord Christ for the sole purpose of getting you to spend money on ... petty material frivolities? Seriously?
In re your second paragraph: The situation is not as symmetric as you make it seem. Hint: schools don't have solstice parties, and most students and faculty likely wouldn't even know what it was. That's just a strawman you pulled out of a hat.
Whatever your opinion is of the legitimacy of tax revenue, you can treat them as "gross revenue" just like any corporation and the computations all work out. It's not like economic models are perfect for non-state corporations anyway.
Yeah, you're a troll but whatever.
The S language is a free spec developed at Bell Labs. According to my professor, the statistics community was generally shocked and appalled to find that they had entered an exclusive licensing deal with Insightful Corp., who make the commercial implementation S-Plus. Before this, S was mostly open-source although the term didn't exist at the time.
So in a way, R is just putting things back to the way they were originally, about 30 years ago.
If they don't, it means that use and development of advanced statistical methods in finance, are as much smoke-and-mirrors as the rest of the finance "industry".
It'd make a good test, actually...
No free RPro for linux, and only binary installers for Windows and Mac. Please explain to me how you aren't violating the GPLv2, which is R's license, by doing the latter. (The former is just annoying, not illegal.) Since you refer to your product as an "enhanced distribution of open-source R", I can only assume you didn't recode it from scratch (which would be quite an undertaking).
Since it's open source, I suppose what I need to do to get a linux version, is find someone who has bought the "server" version and get a copy from them... surely you wouldn't have a problem with this, yes?
I remember my intern officemate once talking disrespectfully about some girl at a stripjoint. I don't have anything against this in-and-of-itself, as it is part of the implied service a stripper provides. However, he was in dire need of an Outlook Adjustment for a few other reasons...
So I told him how much the strippers at that establishment make (most of it under the table) for what is basically a part-time job, and he got real quiet. Then I asked him, since I knew he was under the same "patent transfer" arrangement I was, what exactly the difference was between selling your body and selling your mind, apart from the fact that you'll never get your ideas back, ever. He didn't talk to me much after that. Some people can't handle the truth I guess.
Back on-topic (sort of): if Steven Levitt (the Freakonomics guy) is to be trusted, most "foot soldiers" in the drug trade actually do pull a second job at McDonalds. He tells the story that they are similarly undignified jobs, except that one of them incurs a much larger probability of dying: http://www.ted.com/index.php/talks/steven_levitt_analyzes_crack_economics.html
If you plot those points, you'll still see a trend. If you want to ignore your senses and call it an illusion, you can try computing the regression yourself and find that p=0.005, r=-0.81. Even if you take out Matt Smith you still have p=0.03, r=-0.72.
Retchdog again.
The trend is pretty obvious if you just plot the data. I took the opportunity to learn a little bit of google's painful chart API (no criticism please; I see how it could be prettier and I don't care):
Age of The Doctor, per incarnation.
Nonetheless, there is an overall trend to younger doctors. Correlation of age with order of appearance: r=-0.7, with each doctor an average of 2 years younger than the last. (p-value: 0.01)
Without Matt Smith, that goes down to r=-0.6, and 1.7 years younger. (p-value: 0.04)
Thanks for the data!