Oleg Zabluda's blog

Saturday, May 28, 2011

California vs New York via Google Correlate
Using Google Correlate to find search queries which correlate with being in CA and anti-correlate with being in NY and vice versa:

California vs New York:
http://correlate.googlelabs.com/search?e=id:Xd-l7q5bsR3&t=all#default,100

New York vs California
http://correlate.googlelabs.com/search?e=id:7sB5jnBsYkP&t=all#default,100

Tuesday, December 15, 2009

Biological energy production systems
Anaerobic (homolactic acid fermentation)

type: non-oxidative
name: glycolysis (homolactic acid fermentation)
fuel: 1 molecule of glucose
pH: acid generating
waste: 2 molecules of lactic acid
yield: 2 ATP

Anaerobic (heterolactic acid fermentation)

type: non-oxidative
name: glycolysis (heterolocatic acid fermentation)
fuel: 1 molecule of glucose
pH: acid generating
waste: 1 lactic acid + 1 ethanol
yield: 2 ATP

Aerobic

type: oxidative
name: oxidative phosphorylation
fuel: glucose or fatty acids
pH: acid generating
waste: carbon dioxide
yield: up to 38 ATP

References:

Monday, May 08, 2006

Corvids

Corvidae
- crow -- n зоол. ворона
- ravens -- n зоол. ворон (Corvus corax)
- jay -- n зоол. сойка
- rook -- n зоол. грач (Corvus frugilegus)
- jackdaw -- n зоол. галка (Corvus monedula)
- magpies -- n зоол. сорока

Friday, April 28, 2006

The On-Line Encyclopedia of Integer Sequences

Wednesday, October 19, 2005

Silicon Artwork

This miniature choo-choo train was discovered rolling down the tracks on a LeCroy MVV 200 analog shift register integrated circuit. The "tracks" upon which the train is apparently riding are the high speed shift register.
"No purchase necessary", "Keep away from fire", and "not for resale" are clearly visible in the magnified portion shown as an inset within the photomicrograph. The pad containing this warranty is 450 microns tall by 1850 microns wide and sports 25 lines of text, with each character being between six and eight microns high. This disclaimer--probably the smallest ever written--was found on a Hewlett-Packard "Aspen" (Acquisition Signal Processing ENgine) chip used in digital oscilloscopes in the late 1980s and early 1990s.

Tuesday, October 18, 2005

"Ideas for Startups" by Paul Graham
In his essay Ideas for Startups, Paul Graham writes:

They think creating a startup is just a matter of implementing some fabulous initial idea. And since a successful startup is worth millions of dollars, a good idea is therefore a million dollar idea.
If coming up with an idea for a startup equals coming up with a million dollar idea, then of course it's going to seem hard. Too hard to bother trying. Our instincts tell us something so valuable would not be just lying around for anyone to discover.
Actually, startup ideas are not million dollar ideas, and here's an experiment you can try to prove it: just try to sell one. Nothing evolves faster than markets. The fact that there's no market for startup ideas suggests there's no demand. Which means, in the narrow sense of the word, that startup ideas are worthless."

Wednesday, September 28, 2005

Homo Habilis

Current view on human evolution says that modern human evolved from Homo Habilis -- Handy Man (Человек умелый). How simplistic! Apparently it takes a degree in paleoanthropology not to see the obvious. Just looking at some of the people around is enough to conclude beyound all doubts that some H. Sapience specimens evolved from a currently undiscovered in the fossil record, Homo ahabilis -- Unskillful Man (Человек Нeумелый). Ditto with Homo ergaster -- Working Man (Человек работающий).

- posted by Oleg Zabluda @ 4:24 PM | permalink | | | technorati | bloglines | feedster

Monday, September 19, 2005

Electric Eel Fishing

The Electric eel (Electrophorus electricus) is capable of generating powerful electrical shocks, of up to 600 volts and 1 ampere of for the total of .6Kw for approximately 10 microseconds.
Juveniles produce smaller voltages -- about 110V (220V in Europe).
An electric eel can grow up to 2.5 m in length and 20 kg in weight.
When agitated, it is capable of producing these intermittent electrical shocks over a period of at least an hour without signs of tiring.
Incidentally, 600 volts is exactly the voltage required to cause the initial ionization of a fluorescent lamp, so you can use an instant-start ballast transformer to practice handling eels.
High Voltage

How to fish an electric eel:

Bait: a battery. Depending on the eel size, AAA, AA, C or D dry cell. A full-grown eel can even go for a led-acid car battery. As always with fishing, it is best to use the same bait as fish hunts in its native habitat.
Electric eels bite best during a lightning, when electromagnetic field lures their food source out of their burrows to recharge.
After an eel is extracted from the water, bring it's tail in contact with the head to cause short circuit.
For short-term storage, insulate the eel with electric tape.
For long-term storage, provide grounding compliant with a local Electrical Code. Note that since eel handling is happening in wet environment, US National Electrical Code (NEC) requires, since 1965, that you use Class A Ground Fault Circuit Interrupter (GFCI).

Troubleshooting and repairing an electric eel:

If an electric eel appears to be broken, the first step is to check, and if necessary, replace, the fuse. WARNING: DO NOT EXCEED the fuse rating!
CAUTION: DO NOT DISASSEMBLE the eel. Take it to a qualified technician when service or repair work is required.

Electric eel is excellent baked, stewed, grilled or smoked. Though considered a fatty fish, the eel is high in vitamins A and D, as well as being a good source of protein, e^- and Na⁺.
Electric eel recipes:

Wednesday, August 24, 2005

Nukular stuff

As is well-known, atom consist of electrons in orbital(s) around a nukulus, consisting of a subset of nukular particles called nukulons. Nukulons in atomic nukuli are bound by the nukular force (or inter-nukulon potential), which is the residual strong nukular force, which binds up and down quarks and gluons inside nukulons.

[Thermo]nukular bombs (nukes) and [thermo]nukular reactors release nukular energy from the nukuli employing nukular reactions such as nukular fission or nukular fusion.

Another type of nukular reaction is beta decay, caused by weak nukular force, between leptons (such as electron and neutrino) and quarks. Grand unification theory unifies strong nukular, weak nukular and electromagnetic forces into a single electronukular force.

The branch of physics studying these exciting phenomena is called nukular physics and those who study it are called nukular physisists.

Another fascinating discipline is cell and molecular biology. Every eukaryotic cell contains a nukulus. Cell nukulus is where DNA (Deoxyribonukulic acid) is. DNA consists of a sequence of nukulotides. Cell nukulus is filled with nukuloplasm, and, among other things, contains nukuloli. Nukuloli are made of protein and ribonukulic acid (RNA). After nukuloli are synthesized from nukulotides in the nukuloplasm they leave cell nukulus throgh nukular pores.

More on this here:

WHAT'S WRONG WITH "NUCULAR"
Hey, what do you know. The correct pronunciation of "iron" is (ī'ərn)
MORE ON NUCULAR
"Physics for Future Presidents" on "nukular"

Monday, August 22, 2005

Robot can get back on its feet

Robot can get back on its feet (video).

Uh-uh.

When They come for us, knocking Them off their feet will not be sufficient anymore. Let's just hope that the "off" button is not too hard to find.

It is clear from this story, that if such a robot has a built-in bluetooth adapter, he will be stealing laptops from cars. What do we do? What dooo we do?!

Here is what: we make sure the laptops smell like food, then just watch Them duking it out with the bears.

But here is a problem: what do we do when robot bears come after us?

Turns out we are all covered. We have the right to bear arms. And as you saw in the video, without arms They can't get back on their feet.

P.S. We've seen a robot get back on its feet about a year ago during a RoboExpo in San Francisco. But that one was small and looked non-threatening. I could have easily smashed him with a sledgehammer. I should've. First they came for the show. I did not speak out because I was not a show....

Anupam comments:

back here(india), we don't [have the right to bear arms]. oh mygawd, are we doomed ? possibly. unless of course the researchers in their infinite wisdom chose to use tcl. in which case we are fine. as the bots would tickle themselves to death after the beer-binge.

It probably does use tcl via Expect.

send "standup"
expect "human azimuth=0 elevation=0 distance=1"
send "SMACK"
send "liedown"

Somehow antropomorphic robots look scarier. I wonder how arctomorphic robots will be perceived.

"Aw look this little cute mechnical grizzly can get back o...." SMACK

Monday, August 01, 2005

Teddy Bear

In 1902 U.S. President Theodore Roosevelt had an incident on a bear-hunting trip in Mississippi, when he refused to shoot a bear cub. "Teddy's Bear" was immediately publicized by political cartoonists.

On February 15, 1903, Morris Michtom and his wife Rose made and displayed two stuffed bears in the window of their Brooklyn candy store shortly thereafter, and said they had received President Roosevelt's written permission to call them "Teddy's bears". The Bears sold like wildfire, and within a year, Michtom closed his candy store, and founded the Ideal Novelty and Toy Co. - still one of the biggest toy firms in the world over ninety years later.

German toy maker Margarete Steiff had started to produce stuffed toy animals in 1880; the first one was a little elephant. Her nephew Richard Steiff convinced her to produce a toy bear cub in 1902. They were completely unaware of what was going on in New York. The bear first appeared in public at the 1903 Spring Toy Fair at Leipzig, but - to Richard's disappointment, nobody seemed interested. Legend has it that it was only as Richard was packing away the stand at the end of the fair, that an American toy buyer came up to him, seized the bear, and ordered 3000 on the spot.

Michtom's bear had a more endearing, baby-faced appearance, while Steiff's more closely resembled a real bear cub. Within a few years of their invention, Teddy Bear-mania had swept the world. Roosevelt adopted the bear cub as his mascot for a successful re-election campaign, and Steiff redesigned their bears to create a more appealing face.

A.A. Milne created a fictional bear called Winnie-the-Pooh, named after a stuffed bear owned by Milne's son, Christopher Robin Milne. He first appeared appears in the book published October 14, 1926.

Wednesday, July 20, 2005

Rose Family
Flowering plants

Rosales
- Rosaceae (Rose family)
  - Rosoideae
    - Rose -- Роза
    - Rubus
      - Rasberry
      - Blackberry
      - Strawberry
      - Wild Strawberry -- земляника
      - Cloudberry -- морошка
      - Loganberry -- логанова ягода
      - Boysenberry -- бойзенова ягода
  - Maloideae
    - Apple
    - Pear
    - Quince -- айва
    - Rowan, ashberry -- рябина
    - Hawthorn -- боярышник
  - Prunoideae
    - Cherry
    - Plum
    - Peach
    - Nectarine
    - Apricot
    - Almond -- миндаль
    - Sakura
    - Blackthorn -- терновник
    - Bird Cherry -- черёмуха
- - Mulberry -- шелковица
  - Fig

Tuesday, June 28, 2005

Cave bears (Ursus spelaeus)
Cave Bears were 30% larger then Brown/Grizzly Bears. About 15000 years ago Cromagnons kicked last of them out of his cave. We needed the caves for ourselves -- there was the first recorded real estate bubble going. And that was that. Don't mess with H. Sapience. Neanderthals attempted to do ths same earlier, but couldn't. Wussies.

Monday, June 20, 2005

Vespids & flies

Vespids
- bee -- пчела
- wasp -- оса
- bumblebee -- шмель
- drone -- трутень
- hornet -- шершень
flies
- horsefly -- слепень
- gadfly, botfly -- овод

Thursday, May 19, 2005

Cache-timing attacks on AES.
[via: Schneier on Security]

AES

Daniel J. Bernstein published a profound paper on AES Timing attack. There is very little new in the article from cryptology or cyptography point of view. The side-channel timing attacks of this nature are well known for decades. Moreover, the attacks are not against AES algorithm as such, but against a specific software implementation (OpenSSL on Pentium III).

However it's clear that it is applicable to any other modern CPU, and to any software implementation of any cryptographic algorithm which uses S-boxes. This and other things that I will point out later, led me to conclude that the article and its implications are truly profound. But first of all a quote from the article:

Was there some careless mistake in the OpenSSL implementation of AES? No. The problem lies in AES itself: it is extremely difficult to write constant-time high-speed AES software for common general-purpose CPUs. The underlying problem is that it is extremely difficult to load an array entry in time that does not depend on the entry's index. I chose to focus on OpenSSL because OpenSSL is one of today's most popular cryptographic toolkits. (Constant time low-speed AES software is fairly easy to write but is also unacceptable for many applications. Rijndael would obviously not have survived the fist round of AES competition if it had been implemented in this way.) Is AES the only cryptographic function vulnerable to this type of attack? No. The relevant feature of AES software, namely its heavy reliance upon S-boxes, is shared by cryptographic software for many other functions.

Emphasis is mine. If you don't know the particulars of clock-resolution performance of modern CPUs, just take his word for it. If you do, most likely you don't know half of it. Read the article for all the gory details. The bad news is that an attacker doesn't even need to understand any of the detils. He just need to know that time depends on index, and simply measure it.

Any software algorithm on any modern CPU which involves a table lookup, like S-boxes, for example, is subject to timing attacks. Since many algorithms, including AES, can be implemented without table lookups, it wouldn't be a terribly big deal, if it weren't for one fact: apparently cryptographers participating in AES standardization were completely unaware of the fact that table lookups are a vulnerability. Again quoting from the article: (note that [19] refers to "Report on the Development of the Advanced Encryption Standard (AES)" and [9] refers to "Resistance Against Implementation Attacks: A Comparative Study of the AES Proposals")

7 Errors in the AES standardization process

RAM cache hits can produce timing characteristics," Kocher wrote in 1996 in [16, Section 11], after explaining in detail how to extract secret keys from timing leaks in another operation. A decade later, we still have cryptographic software, in particular, AES libraries, whose timing obviously varies with, among other things, input-dependent RAM cache hits. How did this happen?
Was the National Institute of Standards and Technology unaware of timing attacks during the development of AES? No. In its Report on the development of the Advanced Encryption Standard," NIST spent several pages discussing side-channel attacks, specifically timing attacks and power attacks. It explicitly considered the difficulty of defending various operations against these attacks. For example, NIST stated in [19, Section 5.1.5] that MARS was difficult to defend" against these attacks.
Did NIST decide, after evaluating timing attacks, that those attacks were unimportant? No. Exactly the opposite occurred, as discussed below.
So what went wrong? Answer: NIST failed to recognize that table lookups do not take constant time. Table lookup: not vulnerable to timing attacks," NIST stated in [19, Section 3.6.2]. NIST's statement was, and is, incorrect.
NIST went on to consider the slowness of AES implementations designed to protect against side-channel attacks. For example, NIST stated that providing some defense" for MARS meant severe performance degradation." NIST stated in [19, Section 5.3.5] that Rijndael gained a major speed advantage over its competitors when such protections are considered." This statement was based directly on the incorrect notion that table lookups take constant time. NIST made the same comment in its summary assessments of the finalists," and again in its concluding paragraph explaining the selection of Rijndael as AES. See [19, Section 6.5] and [19, Section 7].
NIST was not the first to make these comments. In a paper Resistance against implementation attacks: a comparative study of the AES proposals," the Rijndael designers commented that a comparative study of the efficiency of the different algorithms on platforms that allow the described attacks should take into account the measures to be taken to thwart these attacks"; incorrectly stated that table lookup is not susceptible to a timing attack"; and concluded that Rijndael was favorable," i.e., relatively easy to secure." See [9, Section 5], [9, Section 3.3], and [9, Section 4.12].
In response to this paper, one of the Rijndael designers characterized cache-timing attacks as irrelevant for cryptographic design." This characterization is out of whack with what actually occurred in the design of the Advanced Encryption Standard. Cryptographic designers should and the AES designers did, at great length consider the difficulty of protecting against timing attacks. Where the AES designers erred was in asserting that table lookups take constant time.
It is hard to guess what might have happened if this error had been pointed out during the AES standardization process. None of the fifteen AES candidates provide high performance on all popular general-purpose CPUs|except by using instructions with input-dependent timings, notably S-box lookups.

As often happens, bugs and vulnerabilities often sneak in in the interface between systems and disciplines. Here it happened on the boundary of three disciplines: cryptography, software, and hardware. Out of 3 pair-wise boundaries, I can speak authoritatively about one: software developers don't understand performance of modern CPUs. It's just too expensive to learn. And by the time you do, the knowledge is obsolete. And I doubt that the other two boundaries are any better.

Monday, May 09, 2005

Stupid translator
Automatic computer translators don't work very well because they don't understand the meaning of what they are translating. Arttu Ylärakkola created a funny illustration of this fact. You type a phrase in English, the translator asks AltaVista Babelfish to translate it to some other language, then back to English. Currently it supports French, German, Italian, Portugese, and Spanish. Here is one example:

Original sentence: I am crazy
French: Could not relocates!
German: I am moved
Italian: They are crazy
Portugese: I am wild
Spanish: I am crazy

Here are more examples:

XML, XHTML, CSS, JavaScript books

Tuesday, April 05, 2005

Satellite images of planes from Google Maps
Planes on Moffett Field, sfo, and a Military bone yard, SR-71 on the tarmac, Travis Aitforce Base courtesy of Google maps. More Google Sightseeing
- posted by Oleg Zabluda @ 10:42 AM | permalink | | | technorati | bloglines | feedster

Tuesday, March 29, 2005

Bay Area birds

Shorebirds and Waders:
- stilt -- n зоол. ходулочник, зуёк долгоногий (Himantopus himantopus)
- avocet -- n зоол. шилоноска, шилоклювка (Recurvirostra spp.)
- sandpiper n зоол. кулик (Actitis hypoleucos)
- rail -- n зоол. погоныш (Porzana)
- plover n зоол. ржанка (Charadriidae)
- black-bellied plover -- зоол. тулес (Squatarola squatarola )
- snipe -- n зоол. бекас (Capella gen.)
- curlew -- n зоол. кроншнеп (Numenius arquatus)
auk n зоол. чистик, гагарка бескрылая (Alcidae)
coot -- n зоол. лысуха (Fulica gen)
merlin -- n зоол. кречет (Falco gyrfalco)
hawk -- n зоол. сокол (Falco)
falcon n сокол
eagle -- n зоол. орёл (Aquila)
eagle-owl -- n зоол. филин (Bubo bubo)
golden eagle -- n зоол. беркут (Aquila chrysaëtos)
sea-eagle -- n зоол. орлан (Haliaëtus gen.)
harrier -- n зоол. лунь (Circus)
osprey -- n зоол. скопа (Pandion haliaetus)
kestrel -- n зоол. пустельга обыкновенная (Falco tinnunculus)
kite -- n зоол. коршун
loon -- n зоол. гагара (Gavia)
wood duck -- зоол. каролинская утка (Aix sponsa)
widgeon -- n зоол. свиязь (Anas penelope)
pintail -- n зоол. шилохвость (Anas acuta)
night heron -- зоол. кваква (Nycticorax nycticorax)
bittern -- n зоол. выпь (Botaurus stellaris)
(black) brant goose -- n зоол. казарка чёрная (Branta bernicla)
teal -- n зоол. чирок (Anas gen.)
scoter -- n зоол. турпан (Oidemia)
shoveler -- n зоол. утка-широконоска (Spatula clypeata)
double-crested cormorant -- n зоол. баклан хохлатый (Phalacrocorax auritus)
wren -- n зоол. вьюрок, королёк и др. птицы из семейства воробьиных
wren -- n крапивник (Troglodytidae)
house wren -- n зоол. домовый крапивник (Troglodytes aedon)
phoebe = pewit -- n зоол. чибис (Vanellus gen.)
ibis -- n зоол. ибис (Threskiornis)
blackbird -- n зоол. дрозд чёрный (Turdus merula)
thrush -- n зоол. дрозд (Turdus)
hermit thrush -- n зоол. пёстрый американский дрозд (Hylocichla guttata)
missel thrush -- n зоол. дрозд-деряба (Turdus viscivorus)
song thrush -- n зоол. певчий дрозд (Turdus musicus)
scaup -- n зоол. -- чернеть морская (Nyroca marila; тж. scaup duck)
grebe -- n зоол. поганка (Colymbus spp.)
bufflehead -- n зоол. гоголь американский (Bucephala albeola)
kingfisher -- n зоол. зимородок (Alcedinidae fam.)
dunlin -- n зоол. чернозобик (Calidris alpina)
knot -- n зоол. исландский песочник (Calidris canutus)
bufflehead -- n зоол. гоголь американский (Bucephala albeola)
titmouse -- n (pl -mice) зоол. синица (Parus)
chickadee -- n амер. синица
siskin -- n зоол. чиж (Fringillidae fam.)
pine siskin -- n зоол. сосновый чиж (Spinus pinus)
pine warbler - n зоол. сосновая древесная славка (Dendroica pinus)
willow warbler -- n зоол. пеночка-весничка (Phylloscopus trochillus)
wood warbler -- n зоол. пеночка-трещотка (Phylloscopus sibilatrix)
goldfinch -- n зоол. щегол (Carduelis carduelis)
indigo bunting -- n зоол. американский зяблик (Passerina cyanea)
finch -- n зяблик
sparrow -- n зоол. воробей (Passer)
pewee -- n амер. зоол. оливковый тиранн (Sayornis)
junco -- n зоол. юнко (Junco gen.)
nuthatch -- n зоол. поползень (Sittidae)
wren -- n зоол. вьюрок, королёк и др. птицы из семейства воробьиных
waxwing -- n зоол. свиристель (Bombycilla gen.)
flycatcher -- n зоол. мухоловка (Muscicapidae)
oriole -- n зоол. иволга (Oriolidae)
starling -- n зоол. скворец (Sturnus vulgaris)
lark -- n зоол. жаворонок (Alauda gen.)
robin -- n зоол. малиновка, зарянка (Erithacus rubecola)
mockingbird -- n зоол. пересмешник (Mimus polyglottos)
flicker -- n амер. зоол. золотистый дятел (Colaptus auratus)
swift -- n зоол. стриж (Apus gen.)
pigeon -- n зоол. голубь (Columba livia)
jay -- n зоол. сойка (Garrulus glandarius)
quail -- n зоол. перепел, перепёлка (Coturnix communis)

Friday, March 18, 2005

Animal Classification

Prokaryots
- Bacteria
- Archaea
Eukaryots
- Fungi
- Protista
- Plants
- Animals
  - Chordata
    - Vertebrates
      - Amphibians
      - Reptiles
      - Birds
      - Mammals
        
        Monotremes
        
        Platypus
        
        Echidnas
        
        Marsupials
        
        Oppossums
        
        Kangaroo
        
        Placental
        
        Rodents
        
        Insectivors
        
        Carnivors
        
        Felidae
        
        Felis
        
        Felis chaus: swamp cat, reed cat, swamp lynx -- камышовая кошка
        
        Mustelidae
        
        Mustela (Weasel) -- горностай, ласка и др. животные семейства куньих
        
        Stoat -- горностай
        
        European Mink -- норка
        
        Black-footed Ferret -- хорек
        
        Domestic Ferret -- хорек
        
        Polecat -- лесной хорек
        
        Otter -- выдра
        
        Sea Otter -- калан, морская выдра
        
        Badger -- барсук
        
        Martes
        
        Marten -- куница
        
        Sable -- соболь
        
        Gulo
        
        Wolverine -- росомаха
        
        Herpestidae
        
        Mongoose -- мангуст
        
        Mephitidae
        
        Skunk -- скунс
        
        Primates
  - Arthropods
    - Insects
    - Millipedes
    - Centipedes
    - Trilobites
    - Crustaceans
    - Arachnids
      - Spiders
      - Scorpions
      - Mites, Ticks
  - Echinoderms -- иглокожие

Sunday, March 13, 2005

Amazon.com: Books: Refusenik: Trapped in the Soviet Union

Just finished reading two books:

"Refusenik: Trapped in the Soviet Union" by Mark Azbel

"Fear No Evil: The Classic Memoir of One Man's Triumph over a Police State" by Anatoly (Natan) Sharansky

Both are very good

Thursday, March 10, 2005

Computer-assisted image colorization

Colorization Using Optimization

Colorization is a computer-assisted process of adding color to a monochrome image or movie. The process typically involves segmenting images into regions and tracking these regions across image sequences. Neither of these tasks can be performed reliably in practice; consequently, colorization requires considerable user intervention and remains a tedious, time-consuming, and expensive task.
In this paper we present a simple colorization method that requires neither precise image segmentation, nor accurate region tracking.[...] In our approach an artist only needs to annotate the image with a few color scribbles, and the indicated colors are automatically propagated in both space and time to produce a fully colorized image or sequence. We demonstrate that high quality colorizations of stills and movie clips may be obtained from a relatively modest amount of user input.

Marked B/W image

Result

A pdf paper is also available.

Sunday, March 06, 2005

[via Google Blogoscoped]

Thursday, March 03, 2005

Side channel attacks: Physical security

[via: Bruce Schneier's Crypto-gram]

Flaw in Winkhaus Blue Chip Lock

The Winkhaus Blue Chip Lock is a very popular, and expensive, 128-bit encrypted door lock. When you insert a key, there is a 128-bit challenge/response exchange between the key and the lock, and when the key is authorized it will pull a small pin down through some sort of solenoid switch. This allows you to turn the lock.
Unfortunately, it has a major security flaw. If you put a strong magnet near the lock, you can also pull this pin down, without authorization -- without damage or any evidence.

A great video and whitepaper on Bumping Locks from CCC Physical Security Workshop. There two more good videos there.

About Me

7 Errors in the AES standardization process