It's all for charidee...
June 24, 2003 12:02 PM   Subscribe

Mitch Kapor reckons that by 2029 no computer - or "machine intelligence" - will have passed the Turing Test. If he's right, the EFF wins $20,000 on a bet.

In the well designed and conceptualised Long Bets website , other participants in the Predictions game: Dave Winer, Esther Dyson, Vint Cerf and Ted Danson!
All predictions here; All bets here - discussions so far here.
Any Mefites willing to stake their rep on cherished beliefs? What do you want to publicly predict will - or will not - happen, and by when?
posted by dash_slot- (17 comments total)
I don't have a bet of my own to throw in here at the moment, but I will publicly predict that Kurzweil will win and Kapor will lose. The Turing Test is a brilliant test, but it's based on two factors with a lot of wiggle room, namely, it has to convince a human that it is human. It doesn't need to convince something with an incredible processing/analytical system that it's human, and it doesn't need to convince anybody that it's something more than an irrational, semi-literate, quasi-predictable being. I'd say it should take even less than the fifteen years Kurzweil targets. Maybe ten.
posted by soyjoy at 12:14 PM on June 24, 2003

When I first heard about and saw the Long Bets site, I dismissed it as a bunch of rich fools throwing money to the wind and puffing up their chests as they made their decrees of the future.

But then I saw Stewart Brand speak about it at the commenwealth club in SF. He's one of the driving forces behind it and a great writer and speaker. He explained that this wasn't a rich men's betting club, but that Stewart and friends tried to devise a way to bring back optimism. He said it might have been due to the year 2000 recently passing by, but when he was a child the future was a wonderous and prosperous place people talked positively about, and that it drove a lot of people to do great things like go to college, work for nasa, etc.

He said that he noticed in recent years that people rarely talk about the great wide frontier of what the world will be like in 50 years. That when they do talk of the future, it's often about wars and strife. The long bets chosen for the site are supposed to have some cultural or technical significance, something that points to a major change in our lives.

When I was at the talk, people in the audience were asked to make up long bets they would stand behind, if they had thousands of dollars to blow on it. Mine was that in the next 10 years, I predicted that marijuana would be available as a prescription drug and decriminalized in the US.
posted by mathowie at 12:23 PM on June 24, 2003

Heh, if I had the money to bet, I'd be willing to bet a terrorist will detonate a nuclear weapon in the US within my lifetime. I honestly believe it'll happen, and I honestly don't think there's anything we can do to stop it.
posted by piper28 at 12:44 PM on June 24, 2003

Long Bet: I bet that, dead or alive, I'll still have a student loan balance in 2038.
posted by yesster at 12:47 PM on June 24, 2003

My guess matt, is that the US could be the only Western country that hasn't decriminalised/ licensed for medical use in the next 10 years!

piper28: if not on that side of the pond, we're #2 target over here - and as has been seen at one of the Windsor's palaces this week, security is never foolproof.

What about the monarchy: how long will that last? I'd lay money - £50 - that William, if he ever makes it, will be our last king. How far would I go to make that happen? That's between me & my conscience...
posted by dash_slot- at 12:48 PM on June 24, 2003

Hell, I'm still waiting for just about any of the things promised to us in those AT&T "You Will" QuickTime required adverts from the early 90s to come true and am thinking of adding them to the Long Bets site.
posted by WolfDaddy at 1:01 PM on June 24, 2003

All bets aside, that's one beautiful site.
posted by stevis at 1:19 PM on June 24, 2003

The Turing Test is a brilliant test, but it's based on two factors with a lot of wiggle room, namely, it has to convince a human that it is human. It doesn't need to convince something with an incredible processing/analytical system that it's human

Maybe I don't want to make my bet until I know which particular humans the incipient turing-test hurdling machine will have to fool. Are we talking about computational linguists taking the test, or the folks who can be fooled by AOLiza and talk to inanimate objects?

Though most human beings do have an incredible processing/analytical system.
posted by weston at 1:31 PM on June 24, 2003

Why do you say Though most human beings do have an incredible processing/analytical system, weston?
posted by WolfDaddy at 1:38 PM on June 24, 2003

posted by donth at 1:46 PM on June 24, 2003

I bet that the Long Bets site will be long forgotten by the time most of these bets come due.
posted by dirtylittlemonkey at 1:58 PM on June 24, 2003

donth: I know that now (I even coincidentally used this post's title tag - a UK catch phrase from Harry Enfield's show - in subconscious homage to zool's original comment, which said 'It's all for charity'): a previous 'DP' reference & my comment apologising for same was tidied away by Lord VoldeMatt, presumably 'cos it's worth a second look, ihho.
posted by dash_slot- at 2:06 PM on June 24, 2003

I'd love to see Ted Danson lose his bet, but it will more likely stay open for a good long time.
posted by Zonker at 3:10 PM on June 24, 2003

By 2050 a large number of instantaneous communications/observation/navigation (J.S. Bell theorem quantum) probes will be sent outside the solar system in an increasingly large pattern. They might be used for a myriad of purposes from being an enormous virtual telescope to a sophisticated early warning system.
posted by kablam at 9:10 PM on June 24, 2003

I'm confident the Turing Test will be 'passed' within ten years. Go check out The Loebner Prize for the annual Turing Test contest, including transcripts and scores for entries from years past. This year's contest will be at the University of Surrey in the UK, do go along.
posted by wackybrit at 9:17 PM on June 24, 2003

Hmm, the gamblers better be the ones alive in 2150 if they want to pay out or receive on this bet.
posted by wackybrit at 9:21 PM on June 24, 2003

I'd like to bet that by 2050 there won't be practical, readily available quantum computers.
posted by alumshubby at 10:35 PM on June 24, 2003

« Older We built this (first) city on rock (and roll)   |   Google calls in the 'language police' Newer »

This thread has been archived and is closed to new comments