Ari Rahikkala
[Recent Entries][Archive][Friends][User Info]
Below are the 10 most recent journal entries recorded in the "Ari Rahikkala" journal:[<< Previous 10 entries]
|
01:46 pm
[Link] | Some people, when confronted with a problem, think "I know, I'll do whatever I practically can given the constraints on my resources and learn to live with the rest".
Now they have zero problems. :(
|
|
01:02 am
[Link] | If I had an infinite amount of computation power available to run just one computation...
... I would run one to search through every program in numerical order until it found one that reproduced the entire content of the latest English-language Wikipedia pages-articles.xml.bz2 (the current state of every article text). Granted, extracting any sort of a usable AI out of the ensuing mess would almost certainly be harder than writing one from scratch in the first place... but at least it'd be interesting to see just how large it would be.
Comedy option: See if a program reproducing pages-meta-history.xml.7z (all article text and discussion, and all history of both) would be significantly bigger. Incidentally, that file (from the last time anyone actually managed to create one without the dumping process dying in horrible pain at some point) is apparently over five terabytes uncompressed, but 7z brings it down to 31 gigabytes. That's sort of quite impressive actually.
|
|
04:09 pm
[Link] | 14:12 <@Dolphin> ari: BASICALLY, RUSSIA IS LIKE YOUR VERSION OF MEXICO (FROM A USA STANDPOINT).
Also, I need to put this somewhere public so I don't have to keep digging it up from IRC logs:
let frl=fst.rootLabel; queue=map(flip Node []).map(length&&&take 1).group.sort; cmb n xs=let (b,e)=splitAt n xs in (Node (sum$map frl b, []) b):e; tree n=head.head.dropWhile(not.null.tail).iterate(cmb n.sortBy(comparing frl)); rpl ss acc (Node (_,a) [])=[(acc,a)]; rpl ss acc (Node _ ns)=concat$zipWith(\s n->rpl ss (s:acc) n) ss ns; hfm ss=rpl ss "".tree (length ss).queue
... OK, so it's not that impressive, but hey, it *is* short enough that I could feed it to lambdabot on Freenode and even have space for arguments :p
|
|
12:03 pm
[Link] | Yeah, not just Turing and Shannon and those other guys. Fuck Andrey Markov too. Fuck him hard.
|
|
02:24 am
[Link] | It somehow took me until now to find out that the whole 3 dB = double intensity difference thing is just an approximation, and that in truth having the "deci-" in the name of the word actually makes sense because the actual rule is that 10 dB = tenfold intensity difference. See: 2 ** (10/3) == 10.079368399158986.
I am somewhat disappointed in myself for not realising this earlier ):
|
|
06:47 pm
[Link] | People keep saying the many-worlds interpretation is untestable. Nothing could be farther from the truth! It's just a little bit expensive and ethically problematic.
See, you start with a bomb big enough to blow up the world.
Then you roll a die. If it comes up anything other than six, detonate the bomb. Keep rolling until everyone is convinced many-worlds is correct.
Incidentally, if overwhelming curiosity is a common trait among intelligent species, I think I just found an explanation for the Fermi paradox. (Multiversal mad science: The real Great Filter!)
|
|
10:26 am
[Link] | Greplin Programming Challenge spoilers below the cut.
( Read more... )
|
|
05:59 pm
[Link] | The next time someone says that randomness is just a property of how values are generated... I'm going to send them a gigabyte of thermal noise. And then I'm going to send them a message saying "replace this with that data I just sent you" and ask how many bytes it took to say that.
(Truth is it's a pretty small, largely philosophical distinction. But it sort of annoys me.)
|
|
02:59 pm
[Link] | Quoth Structure In Fives: Designing Effective Organizations:
This is a book in fives. In this first chapter, we introduce a set of basic mechanisms used to achieve coordination among divided tasks. They number five. Later in this chapter, we develop a visual representation of the organization to help guide us through the book. This has five parts. As we move into the body of the book, we describe the various parameters of structural design. Among the most important of these is decentralization. We shall see that this can take five basic forms. Then, after discussing the situational factors, we introduce our basic configurations of structure and situation. These too number five.
I'm taking a course where one of the requirements is pages 1-23 of this book. Indeed do many things come to pass.
|
|
02:34 pm
[Link] | Anyone who believes that the theory of evolution implies moral darwinism, and who also believes in the theory of gravity, has a moral duty to go jump off a cliff.
|
[<< Previous 10 entries] |
![[info]](http://l-stat.livejournal.com/img/userinfo.gif?v=92.1){kind=small}
ari_rahikkala's journal