* * *
Fleetwell: Mr. Selva, can you summarize the nature of the new contract that Mr. Siskind asked you to present to Mr. Partensky?
Selva: Sure. It pretty much said that he’d be promoted to Principal Research Scientist, with a big raise and so on, and that he’d acknowledge that he invented Slices of Pi on behalf of Mariposa Media Conglomerate.
Fleetwell: Did Mr. Siskind explain why he was asking you to present the contract to Mr. Partensky, rather than, say, presenting it himself?
Selva: Yeah, he said he thought it’d be better coming from a friend. That is, from someone that Mr. Partensky already knew.
Fleetwell: And did you present this contract to Mr. Partensky for his consideration?
Selva: I did.
Fleetwell: Would you describe his reaction to the contract as hostile? Aggressive? Wronged?
Selva: No, not at all–actually he seemed very happy with the idea. In fact he wanted to sign it right there and then–I had to stop him. I told him to take it home over the weekend and talk to his wife, at least.
Fleetwell: Might you have described him as grateful, even?
Selva: Yes, I guess you could say he seemed grateful. I mean, he was definitely grateful for getting the laptop reimbursed. That was what he kept mentioning–he seemed much more interested in that than in the contract.
Fleetwell: Did he express his gratitude to you in any concrete way? Mr. Selva?
Judge Bulger: The witness will please answer the question.
Notes on Napkin (Laurie Partensky’s Handwriting)
(sunday dinner with Igor’s boss!!)
lamb
potatoes (butter at home–cream?)
spinach
wine (look up what is a good one–red with lamb?)
desert?
hair??
new dress???
From Slices of Pi for Total Morons
How Does Slices of Pi Compress Data?
One of the interesting things about Slices of Pi is that it doesn’t work the same way as normal compression algorithms do.
OK, How Do Normal Compression Algorithms Work?
Most compression algorithms look for patterns in data and use those patterns to make the data smaller.
For example, imagine our data was a number like this:
8888888888 8888888888 8888888888 8888888888 8888888888 8888888888 8888888888
That’s a lot of 8s! In fact, if you count, you’ll find that’s 70 of them!
A normal compression algorithm facing this data would say something like “Hmm, I’m noticing a lot of 8s in here,” and compress it to something like:
Write 70 8s
That’s a lot smaller, right? Way to go, normal compression algorithm!
Putting Our Normal Compression Algorithm through Its Paces
Let’s say we tacked another 8 on to the end of our data, so now we had 71, like so:
8888888888 8888888888 8888888888 8888888888 8888888888 8888888888 88888888888
Our normal compression algorithm wouldn’t even break a sweat dealing with the extra 8. Instead of:
Write 70 8s
… it would just compress the data to:
Write 71 8s
So the compressed version of our new data–with 71 8s–would be just as small as the compressed version of our old data–with only 70 8s. Even if we added 100 more 8s, the compressed version of our data wouldn’t grow much:
Write 170 8s
Neat!
But what would happen if we added a 1 instead of an 8, so our data looked like this?
8888888888 8888888888 8888888888 8888888888 8888888888 8888888888 88888888881
Now our normal compression algorithm would have to compress it to something like:
Write 70 8s and 1 1
Wow! The compressed version of our data just got a lot bigger, just from our adding a single 1! What gives?
The Element of Surprise
Let’s think about what just happened in a slightly different way.
Imagine you were playing a game with a friend: your friend picks a number between 1 and 10, and you try to guess what it is. If your friend picked the number “8” 70 times in a row, what would you guess on the 71st time? 8, right? And you wouldn’t be very surprised if 8 turned out to be the right number. But if, after 70 “8s,” your friend suddenly picked a 1, you would be a little surprised. Right?
Well, normal compression algorithms work kind of like you guessing numbers that your friend picks. Remember that normal compression algorithms work by finding patterns in the data–and surprises break patterns, which means our normal compression algorithm has to work harder.
Our normal compression algorithm isn’t surprised when it sees 8 for the 71st time in a row–so it has no problem compressing it into a nice small pattern. But it is surprised when it sees a 1 instead–which means it can’t compress the data quite as much.
In fact, as Claude Shannon, the father of information theory, proved, the more surprising data is, the more information it contains. And the more information there is, the less a normal compression algorithm will be able to compress it!
But, as we’ll see, Slices of Pi works a little differently–in fact, Slices of Pi doesn’t mind little surprises at all–or even big ones!
Text Message from Laurie Partensky to Igor Partensky
remember 2 tell roy we have his hat make sure he knows i gave it 2 u this am but u forgot 2 take it not my fault :-)
Text Message from Roy Selva to Laurie Partensky
Hi Laurie it’s Roy! Igor gave me your number. I’ll be over in your area later this afternoon, I can come pick up my hat if you’ll be around?
From Police Report MA2911034 (Michael Hathaway, Reporting Officer)
Mr. Igor Partensky answered the door. He appeared agitated but gave Officer Harkens and myself permission to enter and was polite and co-operative.
I told Mr. Partensky that we had received reports of a disturbance at this address. Mr. Partensky said that he had come home to pick up a contract he had forgotten that morning, and had found his boss, Mr. Roy Selva, in his bedroom. According to Mr. Partensky, Mr. Selva had only been wearing a hat.
Mr. Partensky told us he had “blown his cap” upon discovering Mr. Selva, and that Mr. Partensky’s wife, Mrs. Laurie Partensky, had heard the shouting and come out of the master bathroom in her robe, which agitated Mr. Partensky further. Mrs. Partensky and Mr. Selva then locked themselves in the master bathroom to get away from Mr. Partensky.
Officer Harkens took Mr. Partensky out on the front lawn while I went upstairs to get Mrs. Partensky and Mr. Selva out of the master bathroom.
[…]
According to Mr. Selva, Mr. Partensky threatened to take the contract and “shove it up [Mr. Selva’s] ass.” Mr. Partensky then attempted to carry out this threat.
Medics on the scene treated Mr. Selva for minor lacerations.
Handwritten Notes of George Baniff, Attorney at Law
Igor Partensky initial meeting
(Referral from Linda White in family law–she thinks he might have a counter-case–had not even retained counsel!)
Doesn’t seem to understand I practice IP, cannot keep his wife from divorcing him
“That bastard” = Ray/Roy? Selva
(Ex) Wife is Laurie
“Slices of Pie”
“Just a toy”
“Just a toy” again (so why is MMC coming after him?)
Extremely theoretical
Extremely naive
Lost copy of new contract (discovery?)
Doesn’t care about $
Doesn’t care about status as inventor
“I can screw bastard who screwed Laurie?”
THIS MOTIVATES
Furious, crying jags, technical mumbo jumbo mixed in
All over the road
Cannot take the stand under any circumstances
From Slices of Pi for Total Morons
Slices of Pi doesn’t mind surprises because it
doesn’t work by looking for patterns in data like other compression algorithms. So how does it work?
The answer is simple: it looks for a copy of the data in the digits of pi.
Blast from the Past
You remember high school geometry class, right? Circles? Pi? All that jazz? Well, just in case you haven’t even balanced your checkbook since high school, here’s a quick refresher:
Pi is the ratio between a circle’s diameter (how big it is across) and its circumference (how big it is around)
The value of pi is 3.14159265358 … etc. etc. etc.
We have to use all those etc.’s because pi is an irrational number, which means that the digits of pi go on forever … and they never fall into a pattern!
Seek and You Shall Find … in Pi!
OK, let’s try out the Slices of Pi algorithm ourselves.
Say we’ve got some data we want to compress. For example, if we had a big ugly number like this:
6939937510582097
… then the Slices of Pi algorithm would search for our data in the digits of pi …
3.1415926535897932384626433832795028841971 6939937510582097 49445 …
There it is!
Once it’s found our data in the digits of pi, the Slices of Pi algorithm can calculate just two numbers …
The number of digits of pi it had to look through before it found the start of the data (40 in this case–count it yourself, but don’t forget to skip the 3 and the decimal point!)
How many digits of data to read from that starting point (16–because that’s how long our original data was!)
… and compress our big ugly number down to something much smaller and simpler:
40 16
Wow! That’s a lot shorter than 6939937510582097!
Where Oh Where Has My Little Data Gone?
But woah, wait! Instead of our big ugly number, now all we’ve got are these two smaller numbers, 40 and 16. Where did our original data go?
Don’t worry! These two numbers contain all the information that Slices of Pi needs to get us our original data back, safe and sound.
This is called decompression, and Slices of Pi does it by just taking the first number (40) and skipping that many digits in pi. Then it reads as many digits as the second number tells it (16) … and voila! We get back our big ugly number, just like we started with:
6939937510582097
Thanks, Slices of Pi!
But How Does Slices of Pi Calculate All Those Digits of Pi?
Great question!
It wasn’t always easy to calculate digits of pi. In fact, for most of human history, the Slices of Pi algorithm wouldn’t have worked very well–it would have taken too long just to calculate the digits of pi and search through them for the data we wanted to compress.
But the advent of quantum computing allowed scientists and mathematicians (like Igor Partensky, the inventor of Slices of Pi!) to find tricky ways to do these calculations and searches much more quickly. Quantum computing is what lets Slices of Pi work so well!
(Fun fact: For a long time, people competed to calculate more and more digits of pi. Before quantum computing, the record was set in 2010 by Shigeru Kondo and Alexander J. Yee, who calculated pi to 5 trillion digits!)
(Fun fact: Some scientists believe that quantum computing allows them to solve certain problems so quickly because it’s actually spreading out all that work over multiple parallel universes. Talk about your collaborative efforts! Pick up a copy of Quantum Computing for Total Morons to learn more!)
Recap: Slices of Pi at Work
Phew, we’ve covered a lot of ground here! Let’s take a moment to recap what we’ve learned.
Unlike normal compression algorithms, which compress by finding patterns in the data, Slices of Pi works by finding a copy of the data in the digits of pi.
Quantum computing is the technology that allows Slices of Pi to calculate and search through the digits of pi so quickly.
Slices of Pi compresses data by turning big ugly numbers into two smaller numbers: the digit of pi where it should start reading to find our data, and how many digits to read.
Well, that wasn’t so hard, was it? Slices of Pi turned out to be pretty simple! But wait: there’s a twist (well what did you expect–there’s always a twist!).
Here’s a hint: what happens if we think smaller?
Memo from Ulysses J. Fleetwell to Stanley Reiner (CEO, Mariposa Media Conglomerate)
While I understand and appreciate your desire to pursue intellectual ownership and a subsequent patent, it is my duty to counsel you towards the legal argument that is most likely to succeed. In this case, that argument is quite clearly not for intellectual ownership of “Slices of Pi,” but rather the pursuit of “shop right” for the algorithm.
My reasoning is as follows:
At the time he invented the “Slices of Pi” algorithm, Mr. Partensky was employed by Mariposa Media Conglomerate as a Field Technician, Class III. Given the standard duties of a Field Technician, Class III, the court is nearly certain to refuse to classify him as an “inventive employee.”
Therefore, Mr. Partensky’s inventions during his employment at Mariposa Media Conglomerate cannot generally be considered works for hire or the intellectual property of Mariposa Media Conglomerate.
Mr. Partensky admits, however, to doing substantial development towards the “Slices of Pi” algorithm on his Toshiba Q100 laptop, for which he requested reimbursement from Mariposa Media Conglomerate–a request which Mariposa Media Conglomerate granted.
Mr. Partensky’s submission for reimbursement suggests that he viewed the laptop as primarily intended for use in the performance of his duties at Mariposa Media Conglomerate.
In which case, Mr. Partensky invented the “Slices of Pi” algorithm largely on Mariposa Media Conglomerate’s equipment and at its expense.
In which case, the requirements for “shop right” are neatly met, and Mariposa Media Conglomerate can claim an implied license, perpetual and irrevocable, to the “Slices of Pi” algorithm.
It is true that “shop right” will not allow Mariposa Media Conglomerate to patent the “Slices of Pi” algorithm, and that this approach leaves the door open for Mariposa Media Conglomerate’s competitors to employ the algorithm as well (perhaps pending separate licensing agreements with Mr. Partensky, if he chooses to pursue a patent himself).
[…]
Another consideration, and one which we should not take lightly, is the sympathetic human element. Even a seasoned intellectual property judge is unlikely to be totally immune to Mr. Partensky’s appeal as the genius underdog, and Mr. Partensky’s counsel, Mr. Baniff, is a deep old file with whom I have done battle before and for whom I have developed a healthy respect. He will not hesitate to exploit the human angle, and will do an excellent job of suggesting that Mariposa Media Conglomerate, by hastily offering Mr. Partensky a promotion contingent on his renouncing all interest in the “Slices of Pi” algorithm, was attempting to perform upon Mr. Partensky’s person the same act which Mr. Selva later performed upon Mrs. Partensky’s.
[…]
If you absolutely insist on pursuing intellectual ownership of the algorithm, rather than “shop right,” we will of course craft the strongest case we can for ownership. We may be able to make the case that “Slices of Pi” is the intellectual property of Mariposa Media Conglomerate by building on the argument that Mr. Partensky developed the “Slices of Pi” algorithm at Mr. Selva’s suggestion, in the hopes of producing work of sufficient value to Mariposa Media Conglomerate to justify financial compensation in the amount of the price of a laptop that he had purchased for personal use. In that case, Mr. Partensky might be viewed as a de facto “inventive employee” working under an implied contract, and the “Slices of Pi” algorithm might be viewed as a “work for hire” and therefore the sole intellectual property of Mariposa Media Conglomerate.
Do consider seriously, however, that any such case will be inherently and markedly in
ferior to the case for “shop right,” and a riskier proposition.
In other words it may be advisable, in this case, to think smaller.
Please advise me regarding your decision at your earliest convenience.
How are Martha and the little ones? We hope to see you all in the Hamptons before Summer bids adieu and Fall comes knocking.
From My 12 Memorable Cases (A Memoir), by Ulysses J. Fleetwell (unpublished)
Despite my counsel, Mr. Reiner and the other decision makers at Mariposa had chosen to swing for the fences–to make the case not for “shop right” but for intellectual ownership of “Slices of Pi.” They had determined, as was their charge and no one else’s, that the financial difference between these two outcomes outweighed the increased risk of the second.
My duty had been to counsel them towards caution, but once they had considered and rejected such counsel, my duty was to deliver them the verdict they had deemed necessary, by any means within my power–and this was the task to which I set myself.
[…]
In following my account of Mariposa Media Conglomerate v. Partensky thus far, gentle reader, you may have found it surprising how many legally irrelevant trivia I sought to know about Igor Partensky–his habits, his foibles, which shows he watched on which nights while dining on which foods–as it may surprise you to know that I came to like him very much.
Other Copenhagens Page 8