The last unbreakable code?

I noticed a brief article in The Guardian with the captivating headline “Can Google be taught poetry?”.

By feeding poems to the robots, the researchers want to “teach the database the metaphors” that humans associate with pictures, “and see what happens,” explains Corey Pressman from Neologic Labs, who are behind the project, along with Webvisions and Arizona State University….

The hope is that, with a big enough dataset, “we’ll be delighted to see we can teach the robots metaphors, that computers can be more like us, rather than the other way around,” says Pressman. “I’d like them to meet us more halfway.”

That sounds utopian, magnificent, turning away from the harsh and narrow-minded informaticism to grand humane concerns. And yet, it reminded me of a recent article in the New Yorker “Why Jihadists Write Poetry”:

Analysts have generally ignored these texts, as if poetry were a colorful but ultimately distracting by-product of jihad. But this is a mistake. It is impossible to understand jihadism—its objectives, its appeal for new recruits, and its durability—without examining its culture. This culture finds expression in a number of forms, including anthems and documentary videos, but poetry is its heart. And, unlike the videos of beheadings and burnings, which are made primarily for foreign consumption, poetry provides a window onto the movement talking to itself. It is in verse that militants most clearly articulate the fantasy life of jihad.

Whatever the motives of Neologic Labs — and I’m guessing they have a pitch to investors that doesn’t rely upon the self-actualisation of smartphones, nor on the profits to be turned from improving the quality of poetry — can we doubt that sooner or later this technology is going to be applied to improving the quality of government surveillance, escaping the literal to follow human prey down into the warrens of metaphor and allusion. It will start with terrorists, but that’s not where it will stop.

Imagine, just to begin with, China equipping its internet with a cybernetic real-time censor that can’t be fooled by symbolic language or references to obscure rock lyrics, which the software will be more familiar with than any fan. Protest movements will be extinguished before people are even aware that they were ever part of a movement.

Hannah’s sweets

The following problem appeared on one of yesterday’s GCSE maths exams, leading to considerable frustration and media attention:

Hannah has 6 orange sweets and some yellow sweets.

Overall, she has n sweets.

The probability of her taking 2 orange sweets is 1/3.

Prove that: n^2-n-90=0

^ is “to the power of”

Now, I am a professional probabilist, and I wasn’t immediately sure how to do it. Why not? Well, there’s something missing: The problem doesn’t tell us what Hannah’s options are. Did she pick sweets at random from the bag? How many? Are we asked the probability that she took 2 orange sweets rather than 3 yellow, given that she actually prefers the orange?  Did she choose between taking sweets out of the bag and putting it away until after dinner?

There should have been a line that said, “She picks two sweets from the bag, at random, without replacement, with each sweet in the bag equally likely to be taken.”

According to the news reports

Hannah’s was just one of the many supposed “real life” problems that the students were required to tackle.

This is just an example of the ridiculous approach to mathematical “applications” induced by our testing culture. It’s not a “real life” maths problem. It’s a very elementary book problem, decked out with a little story that serves only to confuse the matter. You are supposed to know a standard rule for decoding the chatter. If you try to make use of any actual understanding of the situation being described you will only be misled. (Richard Feynman described this problem, when he was on a commission to examine junior high school maths textbooks in California in the 1960s. His entertaining account is the chapter “Judging Books by their Covers” in Surely You’re Joking, Mr Feynman.)

The annuity puzzle

Stanford biodemographer Shripad Tuljapurkar has written a very thoughtful post about the “annuity puzzle”: Why do people generally not choose to purchase annuities that would seem to protect them from a major risk: Being feeble and impoverished 30 or 40 years after retirement? His explanation, which is surely right as far as it goes, is that the shunning of annuities is a rational response to the compensating default risk from the insurance company. You have to live quite a long time to make your nut on an annuity. The “risk” — the probability of living that long — is low, and (he argues, persuasively) one could reasonably conclude that it is outweighed by the likelihood of a financial crash in the interim.

From a behavioural economics perspective, this matches closely one of the standard explanations for discounting: Future returns are drastically uncertain, so we develop the habit of preferring immediate gratification. So this falls in the category of attempts to explain seemingly irrational economic behaviour by showing that it is in fact rational when you take into account limited information or costs of acquiring or analysing information. Of course, any economic theory inevitably struggles to deal with questions of insurance and annuities, where the risk involves the life of the economic agent. The celebrated analysis of this problem by Jack Benny is still relevant.

But while this is a cogent argument for why people shouldn’t buy annuities, I’m skeptical of it as an explanation for why they don’t buy annuities. First, the annuity puzzle is a phenomenon of average people, not savvy investors. I doubt that most people think much about the risk of established financial companies defaulting. One prominent study (based on surveys conducted in 2004) found that 59% of Americans would trade half of their Social Security annuity for an actuarially fair lump sum payment. I’m pretty sure that they are not thinking that they can find a safer investment, with less risk of default, than Social Security. Continue reading “The annuity puzzle”

“Soon enough”

This picture of a British army tank having crushed an automobile that strayed into its path in a small German town has gotten quite a lot of attention.

Tank vs car

Here is the comment of the British military spokesperson:

“Our tank crews go through a very rigorous training process,” he said, reportedly adding that three members of personnel inside monitor the road “which is why they were able to stop soon enough”.

Looking at the photograph, I wonder what would have counted as not stopping “soon enough”. One can imagine similar applications of this Zieglerism. “The Titanic had very rigorous iceberg detection procedures, which is why it was able to stop soon enough.” “The Bush administration had very rigorous antiterrorism procedures, which is why they were able to defend the country adequately against Al Qaeda attacks.” (All but one! 7 1/2 years without an attack!)

More self-deconstructing clichés: Europe edition

I have commented before (here and here) on the weird linguistic phenomenon of clichés being modified to eliminate their actual meaning. Here is an example from yesterday’s BBC report on David Cameron’s attempts to convince other European leaders to support his efforts to rescue his leadership of a fractured Conservative Party reform the European Union:

This was a chance to try to repair burned bridges.

The whole idea of the expression “burn your bridges” is that THERE’S NO MORE BRIDGE! You can’t repair it! Sure, in reality a burned bridge might not have burned completely, so repairs could still be undertaken. But why invoke a metaphorical burned bridge if you actually mean to play down the burn?

What is the writer thinking? “Many people complain that David Cameron has burned his bridges to fellow European leaders. While this is true, those bridges are constructed largely of metaphorical stone, so the damage from burning is not nearly as great as if they had been constructed of metaphorical wood, and repairs are still eminently possible.

“Some in the Conservative Party argue for dynamiting the main pylons of the metaphorical bridges. Metaphorically.”

One might similarly tell of how Alexander the Great, on arriving in Persia, ordered that the ships be burned. But only on the edges, of course, because otherwise they would no longer be seaworthy.

A tricky judgement call for the advertising agency

So you’re the head of the Adobe account, seeking to convince customers that Photoshop is a professional level software tool accessible to the masses. It’s used for important work by experts! It makes news! So now, the question is, do you seek an endorsement based on this news report (from Der Spiegel)?

Russland macht noch immer Kiew für den Abschuss von Flug MH17 verantwortlich. Doch die Fotos, die ukrainische Luftabwehrsysteme in dem Absturzgebiet zeigen sollen, sind offenbar gefälscht. Laut Experten hat der Kreml mit Photoshop manipuliert.

[Russia still claims that Kiev is responsible for shooting down flight MH17. But the photos that supposedly show Ukrainian air defence systems in the area of the crash are blatantly fake. Experts say the Kremlin manipulated them with Photoshop.]

Plaque assay

I was just in Paris for a few days. Walking past the Lycée Simone Weil, in the 3rd arrondissement, I noticed a plaque, such as one sees quite commonly on public institutions:

À la mémoire des jeunes filles, élèves de cet établissement, autrefois école de couture die la ville de Paris, déportées et assassinées de 1942 à 1944 parce qu’elles étaient nées juives, victimes innocentes de la barbarie nazie avec la complicité active du gouvernement de Vichy.

Plus de 11400 enfants furent déportés de France dont plus de 500 vivaient dans le 3ème art de Paris.

Ils furent exterminés dans les camps de la mort.

Les élèves du Lycée Simone Weil ne les oublieront jamais.

[To the memory of the girls, pupils of this establishment, which was then the Paris School of Dressmaking, deported and murdered from 1942 to 1944 because they were born Jewish, innocent victims of the Nazi barbarism with the active complicity of the Vichy government.

More than 11400 children were deported from France, of whom more than 500 lived in the 3rd arrondissement of Paris.

They were exterminated in the death camps.

The pupils of the Lycée Simone Weil will never forget them.]

As I read it, the formulation seemed to me strikingly perfect. The text avoids all the pitfalls that similar texts have been criticised for, whereby they seemed to either be minimising the horror, or pushing away blame, or somehow alienating the victims. The victims were “jeunes filles”, “innocent victims”, “murdered because they were born Jewish” (thus emphasising that it was a purely racist crime. They were “exterminated”, they lived right here, and then this somewhat wishful phrase at the end, usually attached to heroic martyrs, “The pupils will never forget them.” Most striking was the attribution of responsibility to “Nazi barbarism with the active complicity of the Vichy government.” They clearly were concerned to make absolutely unambiguous that they were not minimising French responsibility. Not just “complicity”, but “active complicity”. (Though it wasn’t the “French government”, but only the “Vichy government”.)

I was impressed first, then irritated. Precisely because they managed to tick every box and engrave such a perfect text on the plaque, it made it clear what a formulaic activity it is. (Perhaps the final sentence, unassailably high-minded just as it is clearly not true in any meaningful sense, also drove that point home.) It’s not that they did anything wrong, and I’m glad that they put all these plaques up. There’s just a limit to what you can achieve with a plaque, and perfecting the art of the memorial plaque in some ways undermines the spirit that it is meant to express.

Last refuge

Northern Ireland under pressure after Irish gay marriage referendum win

I’m looking forward to seeing the arguments they will use to resist the pressure. Perhaps Northern Ireland can present itself as a last refuge in Europe for the non-gay-marrying, Christian bakers and florists, antigay clerics. It will help if they can get a high profile asylum case. Maybe the straight son of two North London radical feminists fleeing an arranged marriage to another man.

In all seriousness, I doubt there is anyone who is not astonished at the rapid international progress on same-sex marriage. That includes, most especially, those of us who came of age when acceptance of same-sex relationships was the norm in our student environments and/or recognised the logical force of the argument decades ago. Logical coherence doesn’t usually carry the day in politics nor, I have to admit, should it necessarily. But here we see the power of ideology in political affairs, against those who suppose that politics is merely about balancing competing interests. It’s the ideology of marriage, which had changed enormously in the past two centuries. People like Andrew Sullivan recognised in the 1980s that people’s intuitive understanding of marriage, in Europe and its cultural confrères, had evolved to where it was actually quite hospitable to same-sex marriage. Those of us who felt little emotional attachment to marriage immediately recognised the coherence of this position, but assumed that it would take approximately forever to get over that emotional hurdle, at least everywhere but Holland.

And because I personally attached little weight to marriage, I was definitely among those who thought this an unpromising, because unnecessarily charged, ground to fight on for gay rights. I didn’t see what Sullivan saw, that marriage equality could be the linchpin of a coherent struggle that could overthrow the entire framework of homophobia.

How will we know how many beans we have?

… if we get rid of the bean counters?

I can understand why you’d be annoyed by “bean counters” if, say, you are running a gourmet coffee shop, and your employees keep stopping the production of espresso in order to count the beans. But if, say, you’re running the Congressional Budget Office, I’d say that “bean counters” are exactly what you want. Not Jeb Bush, though. Speaking out in favour of “dynamical scoring”, a procedure (a generous designation) for making budget deficits disappear by counting the positive mojo of conservative principles in the budget on the credit side,

Bush first said he was “all in” for eliminating the “bean counters” who use the traditional “static scoring” method.

The problem with getting rid of the bean counters is, it doesn’t actually get you more beans. There’s no problem disappearing the deficit from your budget with creative accounting, but it doesn’t affect the debt. The creditors aren’t going to take “dynamic scoring” in lieu of payment.

Beans are stubborn things.