The modern pauper commands more wealth than an ancient emperor: he wears clothing made in some other part of the
world from fibres imported from yet other regions of the globe; he eats food shipped across thousands of miles and delivered
fresh; he listens to music each minute of which has had hours of labour spent to maximise its value. That's all due to a
growing economy. Sure, he's a pauper in relation to me--but he's a very rich man indeed in relation to the pauper of a
dozen years ago, or a century, or in the time of King Edmund.
If I may be so bold, how many concubines do you have in your harem?
Unfortunately people do not perceive reality this way. Think about the snipers' case. The snipers killed less than 15 people in three states with millions of people.
Probability of dying of a stroke or choking on a fish bone is far higher, even if you live there.
However, people were afraid to go shopping and leave their homes, schools were closed.
It seems that people (with generous help from the media) tend to exaggerate probabilities of cataclysmic events, while underestimating the mundane risks. I guess it is called imagination.
That does not seem to make much sense - all weights in the neural networks big or small will compress roughly to the same amount. Surely you are not implying that if the network is trained, the weights will have some sort of magical easily compressible patter.
You could have some sort of Bayesian framework with a prior on your weights to try to see when your network is trained, but the the prior should probably not come from gzip.
Think about this, if every person could write their own prescriptions (I'm NOT advocating this, just a thought experiment)
they could probably treat their known illnesses well (assuming they are responsible, and a bit intelligent).
I agree. Why do you have to make a disclaimer that you are not advocating it though?
Well, the point to some degree is that there are no answers. In fact one of the lines (coming from the book, I believe) is: "there are no
answers, there are only choices".
Unfortunately Soderbergh replaced the beautiful final scene of Tarkovsky's "Solaris" with some quasi religious drivel - "am i alive? does not matter -
everything is forgiven". Oh, my.
Overall, it was a good effort, but not on par with
the original film.
Re: Tarkovsky's Solaris vs. Soderberg's Solaris
on
Review: Solaris
·
· Score: 1
That was I had thought too - the drive symbolizes the transition from Earth to the outer space, from the human reality to Solaris.
I think in the film the station is in orbit, while in the book it is hovering.
Here is a quote I pirated from the Oxford English Dictionary. Note the date:
1668 J. HANCOCK Brooks' String of Pearls (Notice at end), Some dishonest Booksellers, called Land-Pirats, who
make it their practise to steal Impressions of other mens Copies.
You are mistaken - I am at a fairly large private midwestern university. The highest paid people here are all doctors on the medical faculty.
Quite a few of them make more than the president of the university.
Business school professors also are also very highly paid (even non-tenured entry level positions pay >130k). The average for tenured faculty in physical sciences is well over 100k.
Only a few top administrators make much more than that.
I don't have the precise breakdown for operating expenses but paying professors and other research stuff dominates the list.
Well, it seems that the big problem here is that
there is no learning involved. People create a very specialized system which does one thing (such as playing chess) very well. So while you can consider it to be a result in AI, it is very limited as the system exhibits no flexibility whatsoever.
A typical example is speech recognition. Under carefully controlled conditions it works reasonably well. However try to use the same system over the phone and the recognition rate drops like a rock.
One key attribute of natural intelligence is its ability to adjust and for all the power of modern computers we are as far from it as we were 50 years ago.
It has a huge electricity surplus from hydro and geothermal plants built after it got it's independence
when European colonialization collapsed in the 40's.
Colonization is a strange word to use. As if its population did not consist almost exclusively of the descendants of the vikings who settled Island about a thousand years ago.
Would you say that Norway was a colony of Sweden as well?
Thanks for explaining. I am familiar with the tools but mostly from the theoretical point of view so it is very interesting to me how they are used in "real life". It is amusing that you are using GA (i guess any relevant feature selection mechanism could work) to remove the
temperature features which are (presumably) useless but for some reason look good on paper.
So you are using a neural net for identification?
I am quite curious what kind of neural net you use and what preprocessing you do. I would you appreciate if you could describe it briefly.
I have played with sregorg. He beat me 5:1. His hardware has probably improved since I played with him. By the way, we are tied 1-1 but you are no doubt a far better zh player.
Is it really true that some programs are competitive at crazyhouse? I used to be a decent player but nowhere close to the best and I could beat even the best crazyhouse programs on the free chess server once in a while. I think a top human player would beat them every time. But then again it was a while ago.
It used to drive me mad how you work hard to get an advantage, finally get the computer in a corner, one wrong move and you are dead...
For
a site like Gamespot or Slashdot, it would be very hard to maintain objectivity and credibility in their journalism, since
they'd basically be representing a product.
Objectivity and credibility of Slashdot's journalism?
Investigators in protective gear pick through a pile of smoking, twisted metal for clues to the
crash. One reaches down, pries back some steel and pulls the black box from the wreckage.
Slashdot Mirror
If I may be so bold, how many concubines do you have in your harem?
It seems that people (with generous help from the media) tend to exaggerate probabilities of cataclysmic events, while underestimating the mundane risks. I guess it is called imagination.
That does not seem to make much sense - all weights in the neural networks big or small will compress roughly to the same amount. Surely you are not implying that if the network is trained, the weights will have some sort of magical easily compressible patter.
You could have some sort of Bayesian framework with a prior on your weights to try to see when your network is trained, but the the prior should probably not come from gzip.
42 is one byte.
I agree. Why do you have to make a disclaimer that you are not advocating it though?
Unfortunately Soderbergh replaced the beautiful final scene of Tarkovsky's "Solaris" with some quasi religious drivel - "am i alive? does not matter - everything is forgiven". Oh, my.
Overall, it was a good effort, but not on par with the original film.
I think in the film the station is in orbit, while in the book it is hovering.
Here is a quote I pirated from the Oxford English Dictionary. Note the date:
1668 J. HANCOCK Brooks' String of Pearls (Notice at end), Some dishonest Booksellers, called Land-Pirats, who make it their practise to steal Impressions of other mens Copies.
Some people are good at it. It is called drawing.
Business school professors also are also very highly paid (even non-tenured entry level positions pay >130k). The average for tenured faculty in physical sciences is well over 100k. Only a few top administrators make much more than that.
I don't have the precise breakdown for operating expenses but paying professors and other research stuff dominates the list.
Salaries of professors, lab assistants, grad students, etc are by far the biggest expense.
A typical example is speech recognition. Under carefully controlled conditions it works reasonably well. However try to use the same system over the phone and the recognition rate drops like a rock.
One key attribute of natural intelligence is its ability to adjust and for all the power of modern computers we are as far from it as we were 50 years ago.
Any more than, say, the US might wish to harm "rouge" nations?
Those evil communists. They are everywhere.
Colonization is a strange word to use. As if its population did not consist almost exclusively of the descendants of the vikings who settled Island about a thousand years ago.
Would you say that Norway was a colony of Sweden as well?
For example: http://www.energy.ca/OIL.html
They did not prove anything. Galileo did.
Galileo proved it himself by dropping balls of different weights from the tower in Piza.
Nice demonstration but nothing new or unexpected there, it had been known for more than 200 years.
Thanks for explaining. I am familiar with the tools but mostly from the theoretical point of view so it is very interesting to me how they are used in "real life". It is amusing that you are using GA (i guess any relevant feature selection mechanism could work) to remove the temperature features which are (presumably) useless but for some reason look good on paper.
So you are using a neural net for identification? I am quite curious what kind of neural net you use and what preprocessing you do. I would you appreciate if you could describe it briefly.
I have played with sregorg. He beat me 5:1. His hardware has probably improved since I played with him. By the way, we are tied 1-1 but you are no doubt a far better zh player.
It used to drive me mad how you work hard to get an advantage, finally get the computer in a corner, one wrong move and you are dead...
Objectivity and credibility of Slashdot's journalism?
I hope you were joking.
These boxes are in fact orange!
That is pretty funny. Increases in computing power tend to have logarithmic rather than exponential effect on many problems.
Yeah, sure. Keep thinking that.