Sunday Papers: No Prisoners

Darkside, like so much, begins with a scream.

A runaway train barrels down the tracks, bound for the bottom of a lake. In swoops Ethics Man. He pulls a lever, switches the train onto another track, saving a carriage full of people. The train runs over a boy standing on the other track, killing him. Did Ethics Man make the moral choice?

“What is The Good?” asks Emily, more than once.

Emily McCoy befriends The Boy caught on the wrong side of the tracks. Emily is a philosophy student shocked by the inhumanity shown to the victims of thought experiments and sometimes she hears voices. She and The Boy set off on an Oz-like quest to find The Wise One, who “tells the secret of life to anyone prepared to make the journey.”

This is Tom Stoppard’s paean to madness, celebrating the 40th anniversary of Pink Floyd’s Dark Side of the Moon – an album that stayed in the UK charts for the average lifespan of a miniature dachshund. The radio play meanders and nightmares through a history of moral philosophy and its consequences: a thought experiment about thought experiments.

If you haven’t already, pour a large whisky and lie back, cup the glass to your chest, close your eyes, and listen.

After a spellbinding speech set to The Great Gig in the Sky, Emily and The Boy are arrested, accused of witchcraft, of bewitching the audience. Facing death by burning at the stake, they await trial, imprisoned.

“It’s classic,” says Mr Baggett, Emily’s philosophy teacher. “Two prisoners – they can’t confer – they’re thinking if they give evidence against each other, their lives will be spared and they’ll go to prison. If they both keep silent, they’ll get off on the main charge and get a shorter sentence. But: if one keeps silent and the other turns state’s evidence… spot the dilemma.”

I am not frightened of dying. Any time will do

The Prisoner’s Dilemma is a game, which doesn’t mean it is fun. Games, in this sense, are mathematical models of strategies and decision-making. Rational players weigh up their options – the benefits and the costs of acting – and strive for their own self-interest.

When confronting the Dilemma once and only once – where you will never experience a repeat encounter with your co-conspirator – your best (that is, rational) bet is to screw them over, giving evidence against them. In the parlance of game theory, this is called ‘defecting’. Unfortunately, your partner-cum-opponent’s best strategy is to screw you right back, and you both lose out and lose big.

Emily and The Boy overturn the cold, inevitable predictions of mathematics by “making false confessions to sacrifice themselves, each for the other.” Mr Baggett continues: “I’ve never come across a case like this in the game of Prisoners’ Dilemma. It’s competitive altruism.”

But two biologists from the University of Pennsylvania might not have been so surprised. The pair recently discovered that “competitive altruism” may be deeply entrenched in the game of life.

In their paper published this week, Drs Alexander Stewart and Joshua Plotkin show how generosity is the only long-term, viable way out of the Prisoner’s Dilemma.

Emily’s despair comes from being stuck in a one-off dilemma. Game theorists, on the other hand, have long known that repeating the Dilemma over and over among the same players can lead to cooperation. This is the Iterated Prisoner’s Dilemma – a game causing far less existential angst due to its happy – or, less unhappy – outcomes.

Until last year, that is.

Last year, William Press and Freeman Dyson, two physicists, discovered a whole new way of playing the Prisoner’s Dilemma. The new way of playing – a set of strategies called, for mathematical reasons, zero-determinant (ZD) – stirred game theorists. Not least because researchers have studied the game since early on in The Cold War and thought the Dilemma long ago gave up its secrets. No: because the ZD strategies suggested it is selfish behaviour, after all, that rules the jailhouse.

In a world where players play ZD strategies, one person can control the outcome for their opponent, forcing them to suck up a prison sentence or even the death penalty. In the long-term, extortion wins.

there is no dark side in the moon, really. as a matter of fact it’s all dark

The Witch Finder, retorting Mr Baggett’s almost sentimental awe at Emily and The Boy’s ‘competitive altruism’ says:

“Altruism is a relic of 19th Century moral thinking, now understood to be its opposite, that is: selfishness in disguise, genetically programmed for long term benefits. In other words, there is no such thing as altruism.”

Drs Stewart and Plotkin may beg to differ. They show how extortion is not the only ZD strategy: at the opposite end of the spectrum lies generosity.

If we imagine a prison yard rather than the two original prisoners cowering in solitary confinement, players interact in less of a head-to-head, one-on-one fashion – similar to how people, animals and other organisms face each other in real life.

In such large, freely mixing populations where players adopt the most successful ways of playing, extortion eventually out-foxes itself and falls apart. The generous strategies – the ones which cooperate with opponents and forgive those who trespass against them – win out. In the researchers’ computer simulations, populations evolved to become generous.

“You might think being generous would be a stupid thing to do, and it is if there are only two players in the game,” said Dr Stewart, “but, if there are many players and they all play generously, they all benefit from each other’s generosity.”

I never said I was frightened of dying

Like the original Dilemma and its depressingly rational outcome that Emily rues, this new paper is, too, a mathematical abstraction of the world. To torture the meaning of Ezra Pound’s advice to poets: go in fear of abstractions. Models are not life, not even the models acting how we ourselves desire to act.

Emily fears the mathematically rational actor, lamenting: “If kindness is only selfishness in disguise, like the Witch Finder said, the question, ‘What is The Good?’ wouldn’t be about anything except what’s best for you and what’s moral about that?”

Maybe, finally, Dr Plotkin can comfort the distressed and disturbed Emily: “Our paper shows that no selfish strategies will succeed in evolution.”

“What is The Good?” asked Emily, more than once. “It is nothing but a contest of kindness.”


References:
Stewart and Plotkin – 2013. From extortion to generosity, evolution in the Iterated Prisoner’s Dilemma. PNAS: http://dx.doi.org/10.1073/pnas.1306246110

Darkside, by Tom Stoppard: http://www.bbc.co.uk/programmes/b038xr3n