Thursday, February 14, 2019

The iPhonification of Everything

So you've got an iPhone XS in Space Grey. Congrats, so do 20 million other people. Maybe you have different cases but otherwise the hardware in all these phones are virtually identical. Yet you can tell with a glance that this is your phone. You can personalize apps and other elements of the home screen. It's your calendar and email and music.

What? You've dropped your phone over Niagara falls. Luckily you've backed up your data. So you go back to Apple and buy another Space Grey iPhone XS and restore your data. Physically it's a completely different phone but for all practical purposes it's though you still had the original phone. Your phone is not defined by the device but the data that resides on it.

It's not just phones. I can log into Google on anyone's Chrome browser and it will feel like my machine.

Now we've all heard about a future world where nobody owns cars and we get driven around in self-driving Ubers, Lyfts and Waymos. One argument against this world is that people feel connected to their cars and unwilling to commute in some generic vehicle. But one can also imagine the car knows who you are, knows how you like your music, your lighting, how you adjust your seats even how your car drives. It becomes your car. Maybe even has electronic bumper stickers that change to support your political party.

You can imagine the same for hotel rooms, your office, maybe even your apartment. It won't replicate your dog (or will it?) but as we get define more by our data than our things, do our things matter at all?

Monday, February 11, 2019

I think ze was confused -- in favor of genderless pronouns

You've probably heard the following:


               At first I didn't want to get an X but now that I have it, I can't imagine life without one.

X could be telegraph, radio, TV, color TV, VCR, CD player, streaming, Netflix, Amazon prime, an uber account, Washer and Dryer, Car Phones (remember those), Cell Phones. If you go back in history  wrist watches or sun dials (or wrist-sun-dials!).

This has happened to me recently though not with an object. I read an article someplace saying that ze can be used instead of he or she. It was referring to nonbinaries (using `they' never quite sounded right) but actually it would be great if this was a general genderless pronoun. I am not making a political statement here (although I doubt I have any readers who are against genderless pronouns).

Once I came across the term ze I found places to use it and now I can't imagine not using it.

In a recent article I wrote I needed to say that someone was probably confused, but I did not know their gender. I used

                                                         Ze was probably confused

which is much better than

                                                         S/he was probably confused

                                                         He or she was probably confused

                                                         The student was probably confused

                                                         They were probably confused.

Note that the first two leave out nonbinaries.

0) In the article I put in a footnote saying what ze meant. In the future I may not have to.

1) Will ze catch on? This blog post is an attempt to hasten the practice.

2) Is there a term for his/her that is non-gendered? If not then maybe zer.

3) Will there be political pushback on this usage? If its phrased as a way to include nonbinaries than unfortunately yes. If its phrased as above as when you don't know the gender, what do you do, then no.

4) Is  nonbinary the correct term? If not then please politely correct me in the comments.

5) Has Ms replaced Miss and Mrs?

I have used the term ze several times since then- often when I get email from a student such that I can't tell from the first name what their gender is, and I need to forward the email, such as

                   Ze wants to take honors discrete math but does not have the prerequisite, but
                   since ze placed in the top five in the math olympiad, we'll let zer take it.




Thursday, February 07, 2019

An Immerman-Szelepcsényi Story

As a grad student in the late 80's I had the opportunity to witness many great and often surprising theorems in computational complexity. Let me tell you about one of them, the Immerman-Szelepcsényi result that nondeterministic space is closed under complement. I wish I had the original emails for this story but instead I'm working from memory and apologies if I get some of the details wrong. I'm expanding from a short version from the early days of this blog.

I started my graduate work at UC Berkeley in 1985 and then moved to MIT in the summer of '86, following my advisor Michael Sipser. In the summer of 1987, Neil Immerman, then at Yale, proved his famous result building on his work in descriptive complexity In those days you didn't email papers, he made copies and sent them by US postal mail to several major researchers in complexity including Sipser. But Sipser was away for the summer, I believe in Russia, and the paper sat in his office.

Immerman also sent the paper to a Berkeley professor, probably Manuel Blum, who gave it to one of his students who decided to speak about the result in a student-led seminar. I forgot who was the student, maybe Moni Naor. I was still on the Berkeley email list so I got the talk announcement and went into complexity ecstasy over the news. I asked Moni (or whomever was giving the talk) if he could tell me details and he sent me a nice write-up of the proof. Given the importance of the result, I sent the proof write-up out to the MIT theory email list.

Guess who was on the MIT theory list? Neil Immerman. Neil wrote back with his own explanation of the proof. Neil explained how it came out of descriptive complexity but as a pure write-up of a proof of the theorem, Moni did an excellent job.

We found out about Robert Szelepcsényi when his paper showed up a few months later in the Bulletin of the European Association for Theoretical Computer Science. Szelepcsényi came to the problem from formal languages, whether context-sensitive languages (nondeterministic linear space) was closed under complement. Szelepcsényi, an undergrad in Slovakia at the time, heard about the problem in a class he took. Szelepcsényi's proof was very similar to Immerman. Szelepcsényi's paper took longer to get to US researchers but likely was proven and written about the same time as Immerman.

Even though both papers were published separately we refer to the result as Immerman-Szelepcsényi and is now just some old important theorem you see in introductory theory classes.

Sunday, February 03, 2019

Don't know Football but still want bet on the Superb Owl?

(Superb Owl is not a typo. I've heard (and it could be wrong) that the  NFL guards their copyright so you can't even say `Buy Beer here for the YOU KNOW WHATl' but instead `Buy Beer here for the big game''. Stephen Colbert a long time ago go around this by calling the game Superb Owl.)


If I knew more about football  I might place a bet related to the Superb Owl. What kind of bets can I place?

1) Bet the point spread: Last time I looked the Patriots were a 2.5 point favorite. So either bet that Patriots will win by more than  2.5 or the Rams will lose by less than 2.5 or just win.

2) Over-Under: bet that either the total score will be over 56.5 or under it.

 There are prop-bets-- bets that are ABOUT the game but not related to the final score.

I've seen the following

1) Tom Brady will retire after the game. I wonder if Tom Brady (or a friend of his) could bet on this one knowing some inside information. Not i any state in America, but off-shore...

2) Jamie White will score the first touchdown.

3) Will Gladys Knight's   National Anthem go longer than 1 minute, 50 seconds (it was 1:47 seconds a few days ago but it shifted to 1:50).

Amazingly, this last one is what Josh Hermsmeyer (on Nate Silver's Webpage)  chose to focus on: here. Note that:

1) The people who picked 1 minute 50 seconds as the over-under probably didn't do much research. They might have set it to get the same number of people on both sides, which may explain the shift; however, I can't imagine this bet got that much action. Then again, I'm not that imaginative.

2) Josh DID. He did an  analysis of what is likely (he thinks it will go longer)

3) So- can Josh bet on this an clean up? Can you bet on this and clean up?

4) There is an issue: Some kinds of bets are legal in some places (betting who will WIN or beat a point-spread is legal in Las Vegas-- the Supreme court struck down a federal anti-betting rule). Some prop bets are legal. The Gladys Knight one is not.  Why not? Someone could have inside information! Gladys Knight would!

So you CAN bet  Rams+2.5 beats the Patriots LEGALLY

but to bet Gladys Knight's National Anthem will take more than 1 minute 50 seconds you might need to use  BITCOIN, and go to some offshore account. Too much sugar for a satoshi.

5) There is another issue- there is no such thing as a sure thing (I blogged on that here). People who bet on sports for a living (I know one such person and will blog about that later) play THE LONG GAME. So to say

           I will withdraw X dollars (for large X)  from my investments and bet it on 
          Gladys Knight's  Star Spangled Banner to go more than 1 minute 50 seconds
          because its a sure thing

Would be... a very bad idea.

The above was all written the day before Superb Owl. Now its the next day and Gladys Knight has sung the National Anthem. So who won the Gladys Knight Bowl? The answer is not as straightforward as it could be, see here.