Irrational Predictions

The MIT club of SoCal managed to snag Dan Ariely on his way out of town, and had him speak to a gathering of MIT and Duke alums and friends.  I took the coast road down to Santa Monica in less time than it took another attendee to get there from Westwood, which is nearby.  Crazy. For various reasons, I got there early for the reception, getting to talk to some nice people, but got into the lecture room late.  It was a hotel conference room set up with chairs.  I decided that front row was better than back row, and actually wound up talking to him a little before he went up when he sat down right there to put on his microphone.  Cool. 


You can find his talks on TED and other sites and he's a good speaker.  Inspiring too.  I've read his stuff but still had ideas while he was talking tonight.  His topic was cheating.  He and his cohorts have found that pretty much everyone/every population cheats at about the same rate for non-cultural tasks.  That rate is about 10%.  Given a chance to lie with impunity and make money at it, most people won't be obvious - instead of taking the whole $20 possible when they "earned" $4 at the task, they're likely to as for $6. What they've found is that it's a combination of cost-benefit analysis, self assessment, and social acceptance.  If everyone is doing it, you can do it too and still feel that you're honest.

Then you can mess with this baseline in a few ways.  Cheat at something long enough, and suddenly you hit the "What the Hell?!" point (as they've named it), and you start to cheat a lot more.  There are things that can knock you back into baseline, though, like confessing, even if no one ever knows what you confessed.  Or not cheating long enough to get there.  Culture plays a role - some things are cheating in one culture that aren't in others.  Being reminded of morality or given a chance to feel superior to a rival will get people to not cheat as well.

There are cumulative effect to cheating though.  The research was started in response to Enron.  It turns out that there are very, very few people who cheat because they like to and for the sport of it.  But since nearly every cheats just a little - and about as much as their peer group, groups that cheat a little can spiral into cheating a lot.  One group cheats a little for a while and either hits the WTH threshold, or another group cheats just a leetl bit more than them, then a third group cheats just a little bit more than them, and the next thing you know, Lindsey Lohan is making bail.  Everyone's doing it, she's just doing enough more than the next person to stand out.  Enron was similar (my words, not Dan's).

This says to me that we really do need regulators and watchdogs.  Not because we're evil, but because people inherently cheat a bit. (Like speeding.)  We cheat less if someone's catching us, or when we're reminded to be responsible.  So we need mechanisms for that reminder, to avoid blundering into situations that spiral out of control, like subprime mortgages. 


 While he was talking, he also brought up how people feel pain when they pay for something as they do it.  It seems to me that this is what prevents internet paywalls from working well. If we just funded an account with something like $50 and had, say, green links to paywall sites and normal blue links to free sites, then paywall sites grabbed a penny or two every time you went to their site (once a day) or a new article (not for each page of an article you greedy guys) or every hour you spent there, they could get money with your consent, but without you having to stop, enter all your personal data, and interrupt your thought process.  Every so often, the site would remind you when you're at your last 10% and you can refill.  Or have auto refill from paypal up to a certain amount per month.  This way you get more content, less advertising, less clicking, and they make money to bring you more content.  the account could maybe even work in fractions of cents.  Say every youtube video is 1/10 of a cent.  For protection, the account would be such that it would only allow fractions of cents to go out, and only if you clicked something (they can't just come and grab it), but with low enough $ amounts that it wouldn't be life ending if you had to cancel a compromised account and make another.

Seriously, if youtube is getting 10 million hits a day and each hit paid (let's go lower!) 1/100 of a cent, they can basically print money.  The New York Times could actually hire editors that would allow them to use the word "torture" when reporting torture and keep sending reporters to the scene.  Things like that.  Because it's not that I'm unwilling to pay for content.  I'm just not willing to pay as much as most sites want, upfront, for unknown benefit, after tons of interruption.  And I don't want to be using my high limit credit card for a $2 charge. 

I think that brings me to 3 types of credit - your basic credit card for medium/large purchases from reputable places, your mini-credit card for the $2-5 purchases in places likely to steal your info, and your micro account to bleed off hundreths of a cent at a time.  It's like paypal, but easier and more invisible.  So I can surf to a hundred sites and it only costs a dollar.  But I don't have to see that dollar leave.

I had a couple of other genius ideas, but I'm too overtired to remember them.  I did figure out what to get my brother for christmas that could potentially rival 12 pounds of birthday bacon for awesomeness.