DJALCHEMI

Long bets

Reading Martin Rees's book Our Final Century — an analysis of a range of natural and man-made risks that may threaten our existence in the next hundred years — led me to discover the Long Bets web site.

The foundation that runs this site aims to use long bets to encourage long-term thinking (see this earlier posting on long-term thinking and this one). The idea is to encourage people to be accountable for making long-term predictions about our future, and ideally to put their money where their mouth is by making a bet that it will come true. If they win the bet, they don't get the money — it goes to a charity they nominate — but they get kudos. Here's a Wired News feature on long bets.

Martin Rees's bet is that "By 2020, bioterror or bioerror will lead to one million casualties in a single event" — that's eight times more than the known death-toll from the Indian Ocean tragedy at the time of writing. As I write, the 232 people who have voted on this prediction are split precisely down the middle on whether they agree or disagree with it. In his book, Rees explains why he is confident of winning.

The full list of predictions ranges from two to two hundred years into the future, and covers topics from energy, artificial intelligence and space travel to religious identity and the Middle East.

I've been wondering what kind of prediction I might make. Perhaps, in the spirit of Jacob Walker's proposal that "Music CDs compatible with current (2002) CD players will still be sold regularly in 2015", I might predict that in 2015 more that 50% of non-sports and news radio programmes will not be listened to live off-air but at some later time, and that 50% of this listening will be done on a pocket-size media player.

But that is rather prosaic topic, mainly concerned just with convenience, and only partly complying with the rule that "the subject of the Prediction or Bet must be societally or scientifically important". A slightly more interesting area is how the store of the world's music and culture is becoming progressively more accessible. To pick one example, I remember how difficult it was, less than twenty years ago, to find Woody Guthrie records. It took repeated visits to specialist shops and the Oxford Street Virgin Megastore (then one of a kind), to get a collection of four or five albums. Now a search for Woodie Guthrie on Amazon gets you a list of ninety albums, half of which are available for delivery to your door within two weeks (including the previously very rare Ballads of Sacco and Vanzetti, which was one of my treasures). Continued sales of these old recordings show the 'long tail' in action.

So I might predict that "the catalogue of music recordings readily available in the northern hemisphere will continue to increase by 50% every five years until 02025 when it may start to plateau or 'saturate'". But to make this into a bet that I could win or lose, I need some index or metric of 'ready availability' (since everything that will be available is available now if you have sufficient money and time). Has anyone got any suggestions for this?

There's a nice irony that Our Final Century is published in the USA as Our Final Hour, which seems to imply that long-term thinking is so alien in that market that consumers wouldn't bother about risks a whole 95 years away, even though a significant number people alive now should still be alive in 02100 given current life expectancy trends.