Sbagliando s'impera

Barabási, Albert-László (2010). Bursts: The Hidden Pattern Behind Everything We Do. London: Dutton. 2010. ISBN 9781101428429. Pagine 323. 19,34 $

images-amazon.com

Letto subito dopo la sua uscita, nella primavera-estate del 2010, ma poi non recensito.

Barabási è un fisico di origine ungherese, nato in Transilvania nella comunità Székely. Questo ne spiega, anche se non ne giustifica, l’acceso nazionalismo: gli ungheresi ritengono, forse a ragione, di essere stati penalizzati nel trattato di Trianon, al termine della 1ª guerra mondiale, a vantaggio della Romania. Il suo nome è legato soprattutto alla teoria delle reti, e in particolare delle scale-free networks, di cui trattava il suo precedente volume destinato al pubblico non specialistico, Linked: The New Science Of Networks (Link. La scienza delle reti). Quando l’ho letto, una decina d’anni fa, nella torrida estate del 2003, ne sono stato fortemente e favorevolmente colpito. Mi era anche capitato, per motivi professionale, di leggere qualche cosa di più tecnico e più accademico di Barabási, che ormai guardavo con ammirazione e rispetto.

Bursts ha in parte tradito le mie aspettative e vedo, dalle recensioni che si trovano in rete, che non sono il solo. Il problema è che, in questo libro, Barabási divaga in continuazione. Non che questo mi spiaccia: anzi, mi diverte molto e mi ci riconosco anche. Però, in tutto questo divagare, rischia di perdersi il messaggio principale: è questo che – nonostante i molti riferimenti al libro che costellano questo blog (a proposito della trilogia di John Twelve Hawks, ma anche di un suo articolo sulla “rete dei sapori” pubblicato su Nature) – mi aveva finora dissuaso dalla recensione.

La tesi centrale di Bursts è che i comportamenti umani siano caratterizzati da bursts, esplosioni di attività all’interno di lunghi periodi di (relativa) quiete (il titolo del libro è tradotto in italiano Lampi, ma mi sembra renda meno di scoppi o esplosioni; allora forse meglio sprazzi). Pensate a come rispondete ai vostri messaggi di posta elettronica: benché i messaggi vi arrivino pressoché di continuo, soprattutto se ne ricevete molti dall’estero e dunque la loro cadenza non è influenzato dai ritmi circadiani del posto in cui vivete, è probabile che tendiate a limitare alcuni momenti della vostra giornata a rispondere. Il perché è molto semplice: le cose da fare sono tante, il tempo è poco. Perciò, istintivamente, date alle cose da fare delle priorità; e quelle che non ricadono tra quelle prioritarie, spesso giacciono inevase: per giorni, per mesi, per sempre.

Paradossalmente, proprio perché bursts e priorità sono ineluttabili, prevedere il comportamento umano è più facile, non più difficile.

Our tasks and responsibilities are poised to queue thanks to a shortage of time. If we could simultaneously work on an arbitrary number of tasks, no one would need a priority list. Time is our most valuable nonrenewable resource, and if we want to treat it with respect, we need to set priorities. Once we do that, power laws and burstiness become unavoidable. [posizione Kindle 1822]

Peccato che questi due importanti messaggi, esposti con chiarezza e rigore, si disperdano in un libro che fa del detour la sua cifra.

Tanto per cominciare, oltre metà del libro (14 capitoli su 28, cui vanno aggiunte 15 immagini originali dell’artista transilvano Botond Reszegh: «There is a theorem in publishing that each graph halves a book’s audience. Its corollary for e-books: Each image halves the number of devices that can properly display it.», pos. 4468) è dedicata alla storia di un eroe Székely, Dózsa György alias György Székely che, se non ricordo male, non ha moltissimo a che fare con il suo tema principale. Fino allo spaventoso supplizio del trono incandescente che vedete raffigurato qui sotto.

upload.wikimedia.org/wikipedia

Condivido largamente, a questo punto, la recensione di Clive Thompson sul Wall Street Journal del 30 aprile 2010:

Mr. Barabási worries that burstiness makes us trackable online by corporations and government, particularly as digital tools like mobile phones produce records of our goings and doings.
This is genuinely fascinating stuff, and when he focuses on the science, Mr. Barabási is a superbly clear writer. But science constitutes a surprisingly small fraction of “Bursts.” Mr. Barabási spends much of the book delivering real-life stories that are supposed to illustrate his principles. Some, like an account of Albert Einstein’s correspondence in 1919 with a little-known scientist, neatly illustrate how bursts govern our lives. But other stories aren’t so successful— particularly Mr. Barabási’s elaborate account of how a Crusade in 16th-century Hungary turned into a gore-splattered civil war. On its own, the Hungarian conflict makes a riveting story, but Mr. Barabási devotes more than a quarter of the book to its telling—yet never convincingly connects the tale to his theme. It became, for me, a maddening distraction. In the end, Mr. Barabási has written a thought-provoking book. But the most rewarding passages appear only, as it were, in bursts.

* * *

Naturalmente, data la classe di Barabási, il libro è colmo di riflessioni e spunti interessanti  (consueti riferimenti alla posizione Kindle):

[…] exploding prevalence […] [552]

We live in a data-rich world.
Putting these data to lucrative use propels the development of further technologies that aim to discover even more about each of us. [555]

Once again, the truth failed to cooperate […] [951]

The problem wasn’t his method but his data. [981]

In 2007, eleven years after its initial discovery, the Lévy character of animal foraging was no longer considered a hypothesis but a well-established scientific fact and had inspired hundreds of publications by ecologists, animal researchers, mathematicians, and physicists. Thus the entire scientific community was shocked when they read that year’s October 25 issue of Nature: a paper coauthored by Edwards, Sergey, and many others concluded that any likeness the path of the wandering albatrosses had to a Lévy trajectory was an artifact of the measurements. [2541]

Science itself often follows a Lévy pattern—a huge jump ahead is trailed by many small, localized steps that appear to take us nowhere, or perhaps even backward in some instances. These are not wasted moves, however, but necessary to testing the boundaries of the new paradigm. [2608]

Predictably Unpredictable
Since the publication in 2005 of The Traveler, a New Age “high-tech paranoid-schizophrenic thriller” with an Orwellian twist, a peculiar debate has absorbed cyberspace. The book takes us into a world where life is free of crises and surprises, a world of ennui-inducing normality. This peace and apparent security is maintained by a worldwide system of computers called the Vast Machine, fed by millions of surveillance cameras, sensors, and detectors. Only the members of a once-powerful ancient society and their sword-carrying protectors, the Harlequins, are aware of the Vast Machine’s reach and are willing to stand up to it.
The ongoing debate this book continues to inspire on blogs and bulletin boards alike might easily focus on the eerie parallels between our own post-9/11 society and the tightly monitored world described in it. But it does not. It might also center on the book’s literary merits, except that, as one critic put it, the writing “is pitched to perhaps a seventh-grade reading level,” an assessment few would challenge. The debate is instead about John Twelve Hawks, its author.
The blockbuster sales and movie rights ought to have elevated Hawks to national celebrity, putting him among the likes of Stephen King and Dan Brown. Yet they did not. And it isn’t because the media shuns him either. The real reason that you never hear about Hawks is that nobody seems to know him. He does not sign books and does not participate in promotional tours. In fact, he has never been seen in public and supposedly communicates even with his editor only through an untraceable satellite phone. Just like the Harlequins on perpetual run from the Vast Machine, John Twelve Hawks lives off the grid, a paranoid seclusion that fuels ongoing speculation regarding his true identity.
The book’s central character is a Harlequin who preserves her off-the-grid anonymity by never using credit cards, opening bank accounts, or staying at permanent addresses. Aware that “any habitual action that showed a Harlequin taking a daily, predictable route to some location” will allow the Vast Machine to predict her whereabouts, she “cultivates randomness.” That is, she relies on a random-number generator, or RNG, to guide her decisions. “An odd number might mean Yes, an even number No. Push a button, and the RNG will tell you which door to enter,” freeing her actions from predictable patterns.
The book is a tale of a battle between good and evil that takes us briefly into something like that fifth dimension Theodor Kaluza proposed to Einstein, throwing into the mix Japanese sword fights and quantum computing. It also again begs the question, could one build a Vast Machine that foresees our actions?
We find it perfectly acceptable that particle physicists can predict within a picometer of accuracy the trajectory of a proton or that rocket scientists can launch a satellite that nine months later drops a robot on Mars. Unlike protons or satellites, however, humans tend to seek new experiences in a continually changing world, making it impossible to foresee their long-term actions. Indeed, given my hectic travel schedule, until recently I found any attempt to predict my whereabouts a few weeks in advance to be a hopeless exercise, fueling my hope that the Vast Machine will always stay where it belongs—in the realm of science fiction. Lately, however, I have begun to have my doubts. [2761: la mia recensione alla trilogia è qui]

Given how impenetrable our past has become, perhaps it’s no small wonder that our future is uncertain. [2976]

Today each person doing research on human dynamics increasingly faces a similar dilemma: How do we avoid contributing to the creation of a surveillance state or conglomerate, a back-to-the-future ticket to Orwell’s 1984?
Hasan has a refreshing answer to this question.
“Intelligence agencies, regardless of who they are, all operate in an industry where their commodity is information,” he observes. “The reason their information has value,” he adds, “is because no one has access to it.”
His solution? Give it up, and it becomes worthless. “It is the secrecy applied to the information that makes it valuable,” he says. And with that, he joins the Szeklers and hides in plain sight, pouring his life out onto his Web site. [3215]

As we have seen, predicting an individual’s behavior is getting steadily easier. And the future is far more valuable than the past, as our travel and purchasing plans are possibly the most potent commodity in our economy. And while secure firewalls and privacy laws protect our pasts, our futures, predicted by sophisticated algorithms, are up for grabs. With that we arrive at a new paradigm I call prospective privacy. It boils down to this: Who owns the information about our future actions and behavior? Who should profit from it? [3244]

But no physicist has ever successfully predicted the trajectory of 10^23 molecules in a gas, either, and that hasn’t stopped us from predicting the gas’s pressure and temperature—arguably far more important than the trajectory of each individual molecule. The same is true for human dynamics. Our deep-rooted unpredictability does not need to bubble up at the level of the society. If we carefully distinguish the random from the predictable, we might be able to foresee many features of the social fabric. [3711]

Any discussion about privacy is a discussion about trade-offs. Giving up the privacy of our medical records may allow the insurance companies to refuse coverage, but not sharing the data could limit the quality of the medical care we receive and thwart research toward the development of better cures. Giving up information about our shopping habits may be perceived by some as an uncomfortable loss of privacy, but others are more than willing to part with it for free or discounted services. Giving up information about our employment history and communication patterns may expose us to potential criminal investigation, but may also reward us with higher security and decrease our chances of being caught in a criminal activity or terrorist attack. [4348]