That title is a phrase that cannot be constructed in American politics. We somehow seem to have forgotten the meaning of “religious freedom”. Romney has a press conference to explain why a member of the fourth largest christian sect in the United States is qualified to be president in light of his unorthodox religion - this is proof positive that the United States has become a theocracy in which religious tolerance has vanished from public discourse.
As a lifelong atheist I have mostly tried to keep away from religion, because religion is evil. This is mostly possible in my life, provided I am willing to ignore the constant reminders that America is a mostly christian country. I really don’t care that much, but it has started to bug me that our foreign policy is mostly guided by religious intolerance in response to religious intolerance.
At what point can humanity come to respect all individual opinions about the origin of life and the role of faith?
I recently submitted a paper with the title “income Inequality in the Attention Economy’. This is my first paper in welfare economics, and the results in the paper came as a surprise to me. Among other things it shows that an increasing amount of attention is concentrated on a tiny number of websites. Among over 200 million websites (suitably defined), fully 50% of the traffic falls in only 1000 web sites. If you look only at the top one million websites (ignoring the bottom 199 million web sites), then the same thing is true - only 0.1% of the websites in this group get 50% of the traffic. In other words, in the attention economy of the web, the rich are getting richer.
It’s interesting to speculate about why this is true, and what the implications are for “heavy tail markets” that economists have spoken about lately. Suffice it to say that I was surprised by the results!
A recent posting by Bruce Schneier about some beautiful Japanese manhole covers reminded me of something that occurred a few weeks ago. While taking a walk around the Google complex I noticed an unusual manhole cover. As my colleague and I continued walking I realized that this offers an entirely new form of advertising opportunity, where advertising could be embedded into manhole covers. We could even make it so that when someone stepped on the surrounding surface, it would register as a “click”, resulting in a payment by the advertiser.
Now you’re probably thinking that this is the stupidest idea you have ever heard of. In reality I’d be surprised if someone had not already done this. There are definitely limits of how far we should be willing to tolerate advertising, and lots of people complain about the pervasiveness of advertising around the world. In recent months the authorities who run the Golden Gate Bridge were pondering whether to put advertising on the bridge. I think that goes way over the line of what is reasonable, but when society keeps insisting on lower taxes, our public facilities will eventually be starved to the point where they look for advertising as a business model.
I just have to admire a society that will have such beautiful manhole covers without being obnoxious about it.
Over the years we’ve all seen technological trends come and go (I’m old so I’ve seen more come and go than some others). I regard Facebook as one of these fads that will fade quickly.
As a mathematician I have been interested in the study of random graphs that arise from social phenomena. Examples include the link graph of the web, the email graph connecting email addresses together, the sexual contact graph, the coauthorship graph, the coworker’s graph, etc. In trying to understand how these graphs evolve, I’ve noticed a lot of buzz surrounding websites that try to build explicitly upon this phenomonen, namely social network sites like LinkedIn and Orkut. In the past I even joined a couple of these to see what the hype was about. In each case, after screwing around with them for about ten minutes I’ve been thorougly underwhelmed with the technology and the privacy compromises that they involve. I predict that these will go the way of geocities (or at least they should).
First of all, they offer a walled garden model, where only people who surrender their privacy are allowed to participate. In order for any of these sites to have any value to you, you have to put some energy into surrendering your information to the control of the closed network. If your friends choose to put their effort into the same network, then you can benefit from it. On the other hand, if your friends put their effort into another network (e.g., myspace or orkut or 360 or linkedin), then you end up having your friends walled off from each other. I have friends all over the world, and I don’t see any value in having them walled off from each other. If the web taught us anything, it is that walled gardens are inferior to gardens without walls.
The privacy issue is a confusing one to many people. Some people are attracted to Facebook because they perceive that it offers some control for them over the information that they share with friends. What they overlook is that in order to gain this control, they have to surrender other forms of their privacy. In particular, in order to join the Google group at Facebook, I have to give my work email address to Facebook. In exchange for surrendering this information to a faceless corporation, I would get to exercise some control over the sharing of information that I put into facebook. Specifically you can limit your data to those who you declare to be in your list of friends. It even offers some fine-grain control over which of your friends can see certain pieces of information. Whoopdy doo.
By contrast, when you create a web site like this one, you have no control over who can read it or what use can be made of it. That drives some control freaks and fearful people crazy. The advantage of giving away control is that you don’t limit your ability to communicate. As a social animal, I like to share information with my friends, but I’m not particularly interested in using a crude web tool to exercise limited control over who gets to see what. I adopt the notion that if you don’t want to say it in public, then you probably shouldn’t type it at all.
In my opinion there is far too much paranoia in this world about privacy, and sites like facebook prey upon this paranoia in a cynical way to exploit the data of others. There is a lot of data about us as individuals that circulates in this world, and sometimes that data gets used against you in ways you may not like. I see sites like Facebook as a placebo against this trend.
Anyway now I’ve joined facebook. Just don’t expect me to put any time into building my network - I like my gardens without walls.
At work the other day, one of the cafes had a theme that exploited the homophones of peas and peace. In particular the menu had a peace sign on it, and they had dishes that used peas in them. One of the dishes was “Visualize Whirled Peas”. Not only do the chefs at Google cook great food, but they have a great sense of humor.
Neal Koblitz recently wrote an article in the Notices of the American Mathematical Sociey titled The Uneasy Relationship Between Mathematics and Cryptography. This article is actually the third by Koblitz, following two previous articles with Alfred Menezes:
There’s something weird about the division in these communities, because it appears to be the mathematicians rather than the computer scientists who are concerned about the mismatch between reality and models of security. I have personally found the work on security in the theoretical computer science community to be both interesting and compelling, but I am often dismayed by the tendency to exaggerate its importance for practical systems.
In particular the terminology of “probably secure” is both misleading and coercive. [see note below] Such cleverly chosen phrases sound appealing at first, but they invite improper conclusions and sound grander than they are. The problem is that it sounds like security is being proved, which may lead non-theoreticians to conclude that they can put more faith in practical systems based on the theoretical protocols than they should. The real nature of “provably secure” systems is that they prove asymptotic reductions between idealized but largely impractical models of cryptographic systems and computational problems that we hypothesize to be difficult. In reality we have almost no evidence for supposing that any computational problems being hard other than our own ignorance or relgious conviction. Even if we could prove that P!=NP, an asymptotic statement of complexity would require considerable work in order to make any inference for practical systems.
A more appropriate term than “provably secure” would be “hypothetically secure” or “theoretically secure”. The term “hypothetically secure” would more accurately reflect the fact that it is only secure if an unproved hypothesis is true (and then only asymptotically or for some unpspecified parameters against a hypothetical adversary). The term “theoretically secure” would emphasize that the term “secure” can only be interpreted in a theoretical model, and it is both dangerous (and often impractical) to make any inference from the theoretical statement of security to any practical statement of security. The concept of “proof” is one that exists only within theoretical models, and we should be cautious about exaggerating the inferences that can be made from theoretical models. Theoretical models are useful, and practioners can wisely make inferences from the theoretical results, but these inferences are something that takes place in the real world, not within a theoretical model.
The articles by Koblitz and Menezes have inspired responses from several members of the complexity community. Recently two have appeared by Jonathan Katz and Michael Mitzenmacher. I also call attention to the response by Oded Goldreich and his most recent response to the Koblitz article. I was pleased to see that Oded voiced a very similar point of view to mine regarding the terminology “provably secure”. The field would be better off if this terminology was stricken from the vocabulary.
I found the first article by Koblitz and Menezes to be quite compelling and interesting, but the followon article was less interesting and the most recent discussion by Koblitz and Goldreich seems to reflect an escalating resentment between groups who have different points of view. I think we would all be wise to return to the pursuit of knowledge and forego the personal attacks.
There are multiple forces that have contributed to the breakdown of respect between groups. Part of the problem is the exaggerated language that pervades the field, and part of the reason is the mismatch in culture between computer science and mathematics. Mathematicians have always regarded conference publications as little more than ephemeral conversation rather than archival publication, whereas computer scientists regard conference proceedings as their primary means of publishing. The influence of money on the field is also a problem, both for the potential of people to make vast sums on commercial products, and the need to compete for meager resources for fundamental research.
Much of science is incremental and sometimes tedious. The need of scientists to continue to publish at regular intervals in order to show progress results in a lot of uninteresting publications. Like most people, I recognized very early that the concept of zero knowledge was fundamental to the understanding of cryptography. Unfortunately, this basic idea was carried to a number of extremes, increasingly straining my patience by the study of minor nuances on the basic concept. After a few years the connection to cryptography was completely lost, but people continued to mumble things containing the term “zero knowledge” while claiming to do cryptography. At some point I remember being on a program committee in which half of the papers contained the acronym ZKIP, but I coined the term ZCIP to describe their contribution to cryptography. ZCIP = Zero Content Indistinguishable Papers.
Unfortunately I think the discourse regarding the role of complexity in cryptography has degenerated to a point where it may take some time to recover. I’m am as guilty as anyone else in this with my quip about ZCIP, but I was motivated by what I saw as exaggeration of importance. The contributions to our understanding that have resulted from the complexity-theoretic point of view are unquestioned. At the same time, we should be careful to understand and respect the relationship between the pursuit of basic science and the development of practical and useful information security systems. Practitioners of both activities need to step back and gain some appreciation for people who are on the other side of this divide.
Update: An an anonymous writer has pointed out that this was a peculiar typo. I meant to type “provable” but instead it came out as “probable”. It goes to show you that the choice of language used to describe a concept is important.
I’ve recently been reading a lot of economics in order to understand what the big problems are in the field. I could have saved myself a lot of work by watching this.
I swear I’ve sat through presentations like this. The graphs looks suspiciously like a power law distribution, which makes it more realistic. Be sure to watch it to the end…
If you’re interested there is also the chicken chicken paper.
I recently took a trip in which I was amused to see two operating systems crash in dedicated applications.
The first instance was on Delta airlines in-flight video system, which about an hour into the trip decided that it really didn’t want to finish playing the video, so it prompty gave a kernel panic and started rebooting. The hilarious part was that it was a flight out of silicon valley, and when the crash occurred there were over a hundred screens showing the kernel panic, and about ten guys on the plane started trying to read the log messages on the screen in front of them to try and figure out what happened. Needless to say when the thing came back up it didn’t remember where it was in the movie, so naturally it started at the beginning.
The second instance was in the national air and space museum near Dulles Airport, where one of the information kiosks was displaying the helpful (and familiar) blue screen of death of windows, with the message:
“A fatal exception 0E has occurred at 0167:BFF9DFFF. The current application will be terminated… Press any key to continue.”
Of course, the kiosk didn’t have any keys, so the instruction wasn’t of much use. Also, since it was the ONLY application that is supposed to run in the kiosk, the computer was functioning as little more than an inefficient space heater.
It’s amusing to note that both of these computer applications had something to do with aviation. Thank goodness these weren’t being used in the autopilot of the airplane for my trip. Maybe someday we can all have the oxygen masks drop down because some idiot used strcpy in the code for the airline software. I can at least console myself with the fact that nuclear weapons don’t yet run windows.
OK I’m not unique in saying that an airline sucks. Most of the world thinks that the airlines suck. The thing that distinguishes the airline industry is that they are so accomplished in finding a way to behave like crap.
I went to have my wife “check in” for a flight today. They say they offer online checkin - the reality is that it’s just a come-on to try and get more money for specific seats. When we went to select a seat, the options were:
1. pay $140 more for a first class seat. No thanks…
2. pay $29 more for “economy plus”. That’s $6 per inch. Not much, but since my wife is all of 5′2″ it’s not really an issue for her to have more leg room. So we decided to pass.
Then you get to find out that they won’t actually release any seats for her to check in on. So it turns out that the entire online checkin thing was just a marketing opportunity. This is the moment when you realize that the airline is not using information technology to streamline their business - they are using it to piss of their customers and build negative loyalty. Good luck with your next bankruptcy guys.
