You're looking at the constraints wrong. Most games can fit the assumptions quite easily. For example, if you think of the rounds not during a game, but as
repeated iterations of the same game.
Take blackjack. Every time you play the game with a fresh set of cards,
the ultimate probability of winning is the same, provided you use the same
strategy each time. That counts as one round, with a winning probability p.
Also, you're wrong about p >0.5, there are strategies for p 0.5 .
Mathematics results are usually phrased in very general / simple terms that aren't necessarily intended to fit a particular situation. You're expected to look around and recognize when a situation you come across fits.
No, Occam's Razor suggests that the obvious enemies of Iran are the
obvious culprits, namely US/Israel.
Inventing fairytales about Russian double indirection to damage
America is way too complicated, and believing an American intelligence
analyst about the fairytale existence of a double indirection by Russia just to
attack America's reputation (ie not even a real attack) is even more complicated.
With 34 subjects the error in the confidence interval for the proportion is
roughly +/- 0.17. They had about 2/3 of their subjects use this vocal
pattern. Seems like they can claim that the lower bound is 49% which may be
all they needed to make their point.
There is a highly unwarranted assumption here that the proportion sought is constant across the whole US population.
So this argument is really saying: if the trend occurs in the exact same proportion everywhere among all females in the US, then it's enough to look at 34 females at a single university to estimate the global proportion.
Of course this is nonsense. The bolded assumption is highly suspect. There are pockets of culture in the US all over, and in each such pocket, the proportion would be very different (eg among the Amish, expect it to be zero, etc).
So all that their test really shows is that: if the proportion across the one university campus is constant (still suspect), then 34 subjects are enough. But to test the constancy assumption, there's only one way: test close to all the
females in that campus anyway, and confirm the proportion across all of them.
Repeat for all females in the US.
Nonsense, it doesn't matter what his degrees are in, or what he claims
his specialty is in. A bachelors degree is a guarantee of (at least)
basic numeracy, which means if he fails a high-school numeracy test,
then he's not fit to be given a bachelors. Nothing else needs to be said.
Actually, a calculator is a hindrance. One of the virtues of mental
arithmetic is that one gets a "feel" for numbers and magnitudes, and
how they behave. People who use calculators exclusively never learn
that skill.
It's like putting people in a motorized wheelchair so they
never learn to walk. In theory it's not a bad idea - a wheelchair with a powerful motor would give us the ability to drive around faster than we can walk or run, and carry lots of luggage around etc. In practice it's a stupid idea, obviously.
What you should have done in that one problem was not used a
calculator, but looked at the sizes of the numbers given in the
multiple choices, and then picked the choice where the magnitude was
in the correct ballpark.
"I won't beat around the bush," he wrote in an email. "The math section had 60 questions. I knew the answers to none of them, but managed to guess ten out of the 60 correctly. On the reading test, I got 62% . In our system, that's a "D", and would get me a mandatory assignment to a double block of reading instruction.
He continued, "It seems to me something is seriously wrong. I have a bachelor of science degree, two masters degrees, and 15 credit hours toward a doctorate.
The guy's quite right. He shouldn't have a bachelor, let alone two masters and 15 credit hours towards a doctorate.
Unfortunately, too many students are in a similar position. Universities have been turned into for pay degree mills, and
the qualifications the higher education industry produces are generally not worth the paper they are printed on.
What users don't get is that the more people use adblock, the more
marketers will have to extract every last penny they can out of the
users they can. That means dirtier, high ROI ads, pop-ups, etc. [...]
I don't think that's right. There's no Nash equilibrium here. Advertisers
compete with other advertisers, so they each have an incentive to extract
every last penny or else it will just go to another advertiser.
The only thing that differentiates ad providers is the reach and operating cost of their network.
The equilibrium with users occurs when there's an ad provider monopoly. If
(say) there's a single advertising platform, then all ads go through it and so
it can stay "benign", because, where else could the ads be displayed? There's no point in improving the returns as long as the returns are positive.
According to this theory, as long as (eg) Google buys up all
advertising platforms that grow a certain size, they can keep the ads
unobtrusive. But once there are two large providers, say Facebook and
Google, if they compete in the same space then the ads would get
nastier unless they collude.
But you're claiming that the C family is superior *because* programmers have better control over memory layout and access to registers etc. If you remove that, what do you get? Just another Java/C#/$insert_favourite_procedural_language. That's where C/C++ is headed with current hardware trends.
I don't necessarily believe functional languages are the answer, but C (the language, not the standard library) will have to evolve a lot. We're seeing tiny steps with OpenMP but it's nowhere near enough.
That's outdated thinking, I'm afraid. It's very difficult to know just
what exactly a modern CPU does during a bunch of cycles - there's
parallel pipelines with speculative execution, reordering of
instructions for efficiency, cache access contention with multiple processors etc. The old C family of languages is slowly drifting away from the underlying hardware architecture.
Along the same lines, do some types of jobs lead to stable
equilibrium configurations of some sort (which cannot be easily escaped)? For
example, does learning to take orders and being a good employee
reconfigure the brain in different ways than being an entrepreneur and
making up your own decisions?
Is it possible to become the latter if you've already spent 20 years being the former?
No, individual people don't have a say in energy policies, countries do. That's why countries are meaningful in this case. To be more exact, regions with common industrial energy policies and closely related plant designs would be what matters, but countries are a good approximation. You'll note I mentioned China, the US and the EU, not the individual countries of Europe. I suppose I should have combined the US and Canada probably due to the close economic dependency between the two countries.
The per-capita stats you link to are physically meaningless (they're politically motivated statistics). What counts in terms of
environmental impact is the total output, so you should be linking to
this instead.
For those who don't want to click the link, China, the US, and EU are the top 3 polluters, unsurprisingly.
You're right of course that search engines make a reasonable effort to
not show/distribute suspect content, but as someone who's done machine learning, I have a healthy skepticism of the possible success rate for automated methods (non-automated methods just don't scale, and arguably haven't scaled since the days long ago when Yahoo was king).
A great historical example for this are web server logs, which people use to mine for all sorts of information including passwords/etc. Early web spiders would simply follow all the links in them indiscriminately, because they're great sources of URLs.
It's hard to make a good job of filtering those URLs with a script or AI program. There are so many possibilities that even a tiny failure rate of 0.1% becomes gigantic when you multiply this by the scale of the web. (the current web is at least 1 trillion urls...)
Well, yes and no. I suspect search engines for web pages would do better to
argue that their behaviour constitutes fair use because, among other factors,
the original material is readily available on-line from the original source
anyway.
That's only a statistical majority of content though. Various types of websites (commercial or otherwise) often try to hide some of their information behind an access barrier, and this has always been the case since the early days of the net.
The technologies that have been used to control access weren't always well thought out, and even then misconfigurations were/are common. Also, it's always been common for people to mirror data (anonymously or not) without having been given permission. The upshot is that spiders collect a lot of web pages and media content that aren't at all intended to be readily available online from the original source. That makes the fair use argument on its own too simple IMHO.
And Google also provides these snippits for free.
They do not sell them.
So do pirates. Are you arguing that libraries, Google, and pirates should *all* be able to copy anything they like, and make it available as snippets? Pirates use bittorrent. The nature of the bittorrent protocol is to cut up a file into tiny snippets, and distribute each of the snippets separately, not necessarily to the same destination.
They do so only within the exemptions for fair use as
provided in the copyright law.
There's no fair use
when the whole of the book (or say 80% of its contents) is
distributed. Google's use is not fair use.
And a
word to each person can hardly be considered distribution of a book.
Each person walks away with essentially Nothing.
What
each person walks away with is irrelevant. There's no distinction
between distributing to a single individual or distributing to a
group. In both cases, it's distribution.(*)
But let's do a quick
back of the envelope calculation. The way Google books actually
works is: people get to see a full page, not a single word.
Now, let's say the average book contains about 300 pages. That means
if 300 people look at a different section of the book, then the full
book has been distributed once on average. In actual fact the number
of pages a person gets to see is more than one, let's say people
look at about 10 pages on average. In that case, when 300 people
have looked at the same book, then Google has distributed the full
book about 10 times already on average.
Of course, anybody who's used
Google books knows that there are some pages that are deliberately
missing each time you come back to look at the same book. Let's say about
20% of all pages are deliberately missing this way. Now when 300
people look at the book, they each look at about 10 pages from the
80% that's available, so Google hasn't technically distributed a
single full book, but it still distributed 80% of the full book,
about 10 times altogether.
That's not fair use. You can't
distribute 80% of a book under fair use. You *might* be able to get
away with distributing less than 10% of a book under fair use. If
Google books did that, then the number of deliberately unavailable
pages would have to be more than 90%, and that would make the
service pretty much unusable.
Amazon actually do this right. If
you use their "look inside" service, you can't see most of the book,
it's well under 10%: just the table of contents and a few more
pages.
That's what Google have to do if they want to argue "fair
use", but that would cripple the service. As it is, they're massive "fair use" infringers so they had better
use another legal argument.
(*) At this point, you might
also like to think about bittorrent distribution, which uses the
same principle of cutting up a file into tiny independent pieces
distributed separately - but that's a digression.
Say I have a book with 1 million words, and I have 1 million customers, and I show each customer only one word, but it's always a different one. Have I distributed the whole book or not?
(I've simplified the issue obviously, but the principle remains valid with the actual Google books practices).
Slashdot Mirror
Cue Yakety Sax.
If he's turning into a penguin, someone please tell him to stop it!
Take blackjack. Every time you play the game with a fresh set of cards, the ultimate probability of winning is the same, provided you use the same strategy each time. That counts as one round, with a winning probability p.
Also, you're wrong about p >0.5, there are strategies for p 0.5 .
Mathematics results are usually phrased in very general / simple terms that aren't necessarily intended to fit a particular situation. You're expected to look around and recognize when a situation you come across fits.
Inventing fairytales about Russian double indirection to damage America is way too complicated, and believing an American intelligence analyst about the fairytale existence of a double indirection by Russia just to attack America's reputation (ie not even a real attack) is even more complicated.
KISS.
There is a highly unwarranted assumption here that the proportion sought is constant across the whole US population.
So this argument is really saying: if the trend occurs in the exact same proportion everywhere among all females in the US, then it's enough to look at 34 females at a single university to estimate the global proportion.
Of course this is nonsense. The bolded assumption is highly suspect. There are pockets of culture in the US all over, and in each such pocket, the proportion would be very different (eg among the Amish, expect it to be zero, etc).
So all that their test really shows is that: if the proportion across the one university campus is constant (still suspect), then 34 subjects are enough. But to test the constancy assumption, there's only one way: test close to all the females in that campus anyway, and confirm the proportion across all of them. Repeat for all females in the US.
Nonsense, it doesn't matter what his degrees are in, or what he claims his specialty is in. A bachelors degree is a guarantee of (at least) basic numeracy, which means if he fails a high-school numeracy test, then he's not fit to be given a bachelors. Nothing else needs to be said.
It's like putting people in a motorized wheelchair so they never learn to walk. In theory it's not a bad idea - a wheelchair with a powerful motor would give us the ability to drive around faster than we can walk or run, and carry lots of luggage around etc. In practice it's a stupid idea, obviously.
What you should have done in that one problem was not used a calculator, but looked at the sizes of the numbers given in the multiple choices, and then picked the choice where the magnitude was in the correct ballpark.
The guy's quite right. He shouldn't have a bachelor, let alone two masters and 15 credit hours towards a doctorate.
Unfortunately, too many students are in a similar position. Universities have been turned into for pay degree mills, and the qualifications the higher education industry produces are generally not worth the paper they are printed on.
I don't think that's right. There's no Nash equilibrium here. Advertisers compete with other advertisers, so they each have an incentive to extract every last penny or else it will just go to another advertiser. The only thing that differentiates ad providers is the reach and operating cost of their network.
The equilibrium with users occurs when there's an ad provider monopoly. If (say) there's a single advertising platform, then all ads go through it and so it can stay "benign", because, where else could the ads be displayed? There's no point in improving the returns as long as the returns are positive.
According to this theory, as long as (eg) Google buys up all advertising platforms that grow a certain size, they can keep the ads unobtrusive. But once there are two large providers, say Facebook and Google, if they compete in the same space then the ads would get nastier unless they collude.
Oh. That's one of the Taxi Openings! Are we keeping track of the fare in old or new money?
Ok, I'll start. Camden Town!
I don't necessarily believe functional languages are the answer, but C (the language, not the standard library) will have to evolve a lot. We're seeing tiny steps with OpenMP but it's nowhere near enough.
That's outdated thinking, I'm afraid. It's very difficult to know just what exactly a modern CPU does during a bunch of cycles - there's parallel pipelines with speculative execution, reordering of instructions for efficiency, cache access contention with multiple processors etc. The old C family of languages is slowly drifting away from the underlying hardware architecture.
Along the same lines, do some types of jobs lead to stable equilibrium configurations of some sort (which cannot be easily escaped)? For example, does learning to take orders and being a good employee reconfigure the brain in different ways than being an entrepreneur and making up your own decisions? Is it possible to become the latter if you've already spent 20 years being the former?
No, individual people don't have a say in energy policies, countries do. That's why countries are meaningful in this case. To be more exact, regions with common industrial energy policies and closely related plant designs would be what matters, but countries are a good approximation. You'll note I mentioned China, the US and the EU, not the individual countries of Europe. I suppose I should have combined the US and Canada probably due to the close economic dependency between the two countries.
For those who don't want to click the link, China, the US, and EU are the top 3 polluters, unsurprisingly.
Iran Blocks American Virtual Embassy Forthwith!
White House Cries Foul! Why Are Iran Such Meanies? !!1
Readers Vote, We Report:
[1] Iran Are Meanies
[2] This Response Was Obvious, And Iran Are Still Meanies
A great historical example for this are web server logs, which people use to mine for all sorts of information including passwords/etc. Early web spiders would simply follow all the links in them indiscriminately, because they're great sources of URLs.
It's hard to make a good job of filtering those URLs with a script or AI program. There are so many possibilities that even a tiny failure rate of 0.1% becomes gigantic when you multiply this by the scale of the web. (the current web is at least 1 trillion urls...)
That's only a statistical majority of content though. Various types of websites (commercial or otherwise) often try to hide some of their information behind an access barrier, and this has always been the case since the early days of the net.
The technologies that have been used to control access weren't always well thought out, and even then misconfigurations were/are common. Also, it's always been common for people to mirror data (anonymously or not) without having been given permission. The upshot is that spiders collect a lot of web pages and media content that aren't at all intended to be readily available online from the original source. That makes the fair use argument on its own too simple IMHO.
And that's Rule 43 I believe. Wow, you sure know a lot of rules!
That's Rule 27, right? Never admit defeat, especially when it's obvious you've lost ;-)
So do pirates. Are you arguing that libraries, Google, and pirates should *all* be able to copy anything they like, and make it available as snippets? Pirates use bittorrent. The nature of the bittorrent protocol is to cut up a file into tiny snippets, and distribute each of the snippets separately, not necessarily to the same destination.
There's no fair use when the whole of the book (or say 80% of its contents) is distributed. Google's use is not fair use.
What each person walks away with is irrelevant. There's no distinction between distributing to a single individual or distributing to a group. In both cases, it's distribution.(*)
But let's do a quick back of the envelope calculation. The way Google books actually works is: people get to see a full page, not a single word.
Now, let's say the average book contains about 300 pages. That means if 300 people look at a different section of the book, then the full book has been distributed once on average. In actual fact the number of pages a person gets to see is more than one, let's say people look at about 10 pages on average. In that case, when 300 people have looked at the same book, then Google has distributed the full book about 10 times already on average.
Of course, anybody who's used Google books knows that there are some pages that are deliberately missing each time you come back to look at the same book. Let's say about 20% of all pages are deliberately missing this way. Now when 300 people look at the book, they each look at about 10 pages from the 80% that's available, so Google hasn't technically distributed a single full book, but it still distributed 80% of the full book, about 10 times altogether.
That's not fair use. You can't distribute 80% of a book under fair use. You *might* be able to get away with distributing less than 10% of a book under fair use. If Google books did that, then the number of deliberately unavailable pages would have to be more than 90%, and that would make the service pretty much unusable.
Amazon actually do this right. If you use their "look inside" service, you can't see most of the book, it's well under 10%: just the table of contents and a few more pages.
That's what Google have to do if they want to argue "fair use", but that would cripple the service. As it is, they're massive "fair use" infringers so they had better use another legal argument.
(*) At this point, you might also like to think about bittorrent distribution, which uses the same principle of cutting up a file into tiny independent pieces distributed separately - but that's a digression.
Say I have a book with 1 million words, and I have 1 million customers, and I show each customer only one word, but it's always a different one. Have I distributed the whole book or not?
(I've simplified the issue obviously, but the principle remains valid with the actual Google books practices).