24 March 2006

Dream or Nightmare? You decide

So last night, I woke up in the middle of the night from a dream where I went back to work at Pixar. The movie we were working on a was a new Star Wars movie. The Empire is trying to build a big weapon, which was a giant cigar shape that sits underneath a Star Destroyer (I guess they hadn't come up with the Death Star thing yet). At the end of the movie, the star, Bruce Willis, hotwires the device so that when the Empire fires it, it blows up.

18 March 2006

The Pay-to-send Email World [was: RE: nytimes: goodmail]

I've thought through schemes like this any number of times (it's not a new idea, contrary to what Dyson implies in her article), and many of the things I value about the internet become unworkable if most internet email shifts to a pay-to-send basis. Unfortunately, you actually have to think about it to realize that.

Running a mailling list like friends@ourshareddomain.com [note to blog readers -- this discussion thread was moved over from a thread on a mailing list I'm on] is one of the things that will go away in a pay-to-send internet -- current membership totals about 30, so sending a message to ourshareddomain.com means somebody has to fork over $0.30. Well, who's gonna be the person to fork out that $0.30? We have no way to make you pay at the moment you hit send. This discussion thread is just about to cross over the $1 point. Would you pay $1 to be part of every email thread on the hiking alias?

Advocates of pay-to-send schemes claim, "Oh, no, you can still send free email, it's simply that free email won't have the little icon next to it." Yet, Goodmail and other pay-to-send scheme advocates claim they're going to resolve the spam problem. Well, they only are solving the spam problem to the extent that covered recipients ignore all email without the little icon. So, they're lying about their side effects: for them to achieve the benefit they promise (fixing the spam problem), free email must be ignored. So much for friends@ourshareddomain.

Don't think that will happen? One of the geniuses of this idea as Goodmail is advocating it is that they're going to share their revenue with AOL. Once that happens, your ISP will have a vested financial interest in making it harder and harder for you to read free email (after all, they'll be "solving the SPAM problem" for you). Their motivation will be to start quarantining (i.e., not sending to your mail client by default) free email.

The scotchnight problem recurs over and over when you look at a pay-to-send internet: evite; topical mailing lists e.g. sourceforge.net; alumni networks at your college; the mailing list for a condo homeowner's association; the expixar mailing list; the neighborhood association mailing list; etc., etc. What happens in that form of the internet is that all players in the internet will be highly disincented to relay email in any form whatsoever.

A second problem has to do with open standards. Goodmail -- and all of its' proprietary competitors -- have strong motivation to make their network closed, requiring their own software in order for you to share in their revenues, because it represents real money. That is, they are motivated to make it so that non-free email cannot be relayed except by their own software. If Dyson is right and their are multiple competitors, the nightmare will unfold: to be a viable ISP, you need to have five or six different proprietary email relays installed, and somehow sort out which mail should be processed by which relay, and have revenue-sharing agreements in place with those five or six companies. That speaks to what the pay-to-send proposal in this form is really about: providing a new revenue stream for large ISPs.

The pay-to-send schemes allows ISPs to collect money for sending spam, rather than no one being paid for it. The large ISPs are willing to see this experiment run because if it's successful, they and the new companies will share in taxing the senders of Spam.

Dyson asserts without proof that "I also think it [Goodmail] and its competitors will eventually transform into services that more directly serve the interests of mail recipients." Of course, she has to state that without proof because there is no evidence any such thing will ever happen. She's asserting the equivalent of, "Give some more money to the beleaguered phone company, and they'll probably decide to share some with you." Both by logical deduction and historical precedent, she's wrong: they will keep that money and be thankful they're being paid for the spam they deliver.

Lastly, Dyson makes the counter-economic assumption that because the ISPs are making money, they'll be more vigilant in guarding against Spam. Unfortunately, the pay-to-send scheme that we have long experience with -- the Post Office -- demonstrates nicely why this is false. In every country, the post office is highly motivated to promote the sending of junk mail, because it is their primary source of revenue. The fact that it costs $0.20 or so to send me a piece of junk mail has clearly not stopped a deluge of companies from deciding to assault my mailbox. The post office makes sure there's no effective way to screen it, because effective screening of junk mail would put them out of business (you can register for "no junk mail" at the post office: all that does is stop mail that isn't specifically addressed to you. So by the Post Office's definition, the AOL disks I periodically receive are not, repeat not, junk mail).

Some people might still prefer the pay-to-send internet, because one thing you can't deny is that the Nigerian and Chinese spammers will largely be gone. Like postal mail, your overall volume of email will be lower than the current internet. Like postal mail, most of email will still be junk, but like postal mail, the delivery service will have been paid to bring you the junk. And like postal mail, you won't trivially send emails to 30 friends, you'll have to find some cheaper, easier way to get a hold of them, like email used to be.


On Mar 18, 2006, at 9:35 AM, david fix wrote:

I don't know about the merits of her position but this argument strikes me as sophistry.

"What shocks me most about the opposition to Goodmail is that people who claim to believe in the free and open Internet, with its welcome attitude to innovation, want to shut down an idea. That's wrong."

From: Bill Polson
Reply-To: ...
To: ...
Subject: nytimes: goodmail
Date: Fri, 17 Mar 2006 21:33:43 -0800

This seems reasonable to me, but on a recent hike I advocated something like this and got an earful from a fellow hiker. No names, but he hikes too fast and knows way too much about cooking and obscure technologies.



March 17, 2006
Op-Ed Contributor

You've Got Goodmail

A COMPANY called Goodmail Systems thinks it has come up with a potential (and partial) solution to the problem of spam and fraud on the Internet. According to Goodmail, market forces are the answer, rather than the kinds of ineffective regulations that have so far failed to solve the problems.

What Goodmail is proposing is a sort of FedEx for e-mail. For a penny or less per message, the sender gets guaranteed delivery for mail and the promise that it will stand out in the user's mailbox. The recipient pays nothing. (Goodmail, of which I am not an investor, has tested its system with the participation of a few companies, including this newspaper.)

Internet service providers like America Online, which receive and process mail in bulk, can share in Goodmail's revenue if they want, as long as they promise to pass the mail to their customers without filtering it for spam. The payment encourages AOL to adopt the service and to display a "certified e-mail" icon to users on each "stamped" message, indicating that the message is wanted and safe.

Goodmail's customers have to prove that recipients want their mail, and Goodmail checks the sender's mailing behavior and manages the quality of the mail through a system that makes it easy for recipients to complain about unwanted messages. Too many complaints and the senders lose their accounts.

This all sounds perfectly sensible to me, but Goodmail has been met with a barrage of criticism and calls for a de facto boycott from several nonprofit and public interest groups. These organizations seem to think that all Internet mail must always be free, just because it was free before. Yet they pay for computers and Internet access and office supplies, just like everyone else.

Goodmail, in my eyes, does not raise moral issues. It simply wants to make the Internet a better place — and yes, make a little money along the way.

Of course, the critics say, this is the first step. Pretty soon all mail will cost money, and then the free, open world of the Internet will be closed to poor people, nonprofits and other good guys, while multinational conglomerates fill their ever-growing pockets.

I agree that pretty soon sending most e-mail will cost money, but I think that's only right. It costs money to guarantee quality and safety. Moreover, I think the market will work, and that it will not shut out deserving senders, if we only let it work freely.

In fact, I hope Goodmail succeeds, and that it has lots of competition. I also think it and its competitors will eventually transform into services that more directly serve the interests of mail recipients. Instead of the fees going to Goodmail and AOL, they will also be shared with the individual recipients.

There's no question that we need to try some new approaches to e-mail. The current situation — where the Internet is ridden with spam, most mail is unwanted and some is objectively dangerous — illustrates a market failure. When the Internet began, it was a small community, one that needed neither markets nor regulations. The people on the Web mostly behaved well, and things worked.

But the Internet has become mainstream, which means we have abuse and crime and freeloaders. We need a way to deal with the abuse. Spam filters are imperfect, partly because the senders of "bad" mail are getting better and better at defeating them. And those who take advantage of the Internet aren't paying for the costs of their abuse.

Instead, the recipients' Internet service providers are paying the cost of spam filtering, and recipients are paying when legitimate messages get caught in those spam filters.

Senders should bear the costs of sending mail, and it should be the senders' duty to figure out whether each piece of mail is wanted. Ultimately, I believe, Goodmail or its successors will develop a mechanism to rebate some of the fees to the senders whose mail is wanted. That's why I don't worry about individuals and nonprofits being squeezed out.

In the short run, AOL and others will serve as the recipients' proxies. If they don't do a good job of ensuring that customers get the mail they want, even from nonpaying senders, they will lose their customers. And in the long run, recipients will be able to use services like Goodmail to set their own prices for receiving mail.

In my case, I'd have a list. I'd charge nothing for people I know, 50 cents for anyone new (though if I add the sender to my list after reading the mail, I'll cancel the 50 cents) and $3 for random advertisers. Ex-boyfriends pay $10.

What shocks me most about the opposition to Goodmail is that people who claim to believe in the free and open Internet, with its welcome attitude to innovation, want to shut down an idea. That's wrong.

If people like those little stamps that mark their mail as safe and wanted or as commercial transactions, then let the customers have them. And let other companies compete with Goodmail to offer better and less expensive service.

Goodmail isn't good because it's new, but neither is it bad because it's new. If it's a good model, it will succeed and improve over time. If it's a bad model, it will fail. Why not let the customers decide?

Esther Dyson, an investor in technology start-ups, is the editor of Release 1.0.

QotD: [GameDev] Design Docs

"At the onset of development, completed design docs are not always very useful to us, as we're not yet sure what will be fun."

-- from the Post-Mortem of Harmonix's "Guitar Hero", the best game of 2005 IMHO, published in Game Developer magazine

17 March 2006

Snow on Mt. Tam

Last month we went hiking up Mt. Tam into the snow. Dan Lyke has some photos here, it was amazing.

15 March 2006

Proud to be a Member

I'm only halfway through it, but already the March 2006 issue of the Commonwealth Club newsletter has knocked me up side of my head twice. Daniel Pipes, a Middle Eastern scholar who was living in Egypt in 1979 when the Oslo accords were signed, has a scathing article called "The Palestinian-Israeli War." Robert Fisk is a "mouse journalist" who has been sneaking around Iraq (and Bosnia and Afghanistan and...) in decrepit rental cars and getting shot at for 20 years.

While I don't rush to embrace either's worldview, they both have been up close talking to the participants more often than I have and have a logic to their point of view.


I couldn't find either printed article anywhere on the web; your local library may be able to oblige (or you can borrow my copy when I'm through!). However, the Commonwealth Club has the audio archives of the lectures online.
Pipes' lecture is at: http://www.commonwealthclub.org/archive/05/05-11pipes-audio.html
Fisk's is here: http://www.commonwealthclub.org/archive/05/05-11fisk-audio.html

02 March 2006

Brain Training and Seniority as a determinant of Ranking

So this evening at Scott & Amy's house we checked out Brain Training, a "game" for the Nintendo DS that alleges to try and keep your brain fit into old age. It's a series of simple puzzle and math games that really is kind of akin to doing crosswords puzzles to "keep the brain sharp".

The UI was a little difficult for us only because I can't read Japanese fluently, but in general the UI of Brain Training is super supportive. It very clearly tells you where to click next and when you should get ready because the "test" is coming up soon. It reminded of a persistent difference between Japanese and North American games: Japanese games are often easy, with an emphasis on putting in your time rather than insight and/or skill.

Brain Training has some drills that require thought, or at least attention, since one of the raison d'etres for it is to keep your mind sharp. But in general it very gently guides you through a series of exercises which, if followed, promise to decrease your "brain age". The most obvious analogy is to Japanese RPGs, which typically reward time and effort rather than any particular skill.

Today it finally occurred to me to congruous this is with the postwar Japanese culture: in the traditional sarariman structure, your place within the company is mostly determined by seniority, with individual merit and/or achievement having only a minor effect. These sort of games that kind you very strongly through them are an entertainment equivalent, where the 60 hours of gameplay will surely be rewarded by saving the universe, whether or not you applied any special insight along the way.

Of course, the sarariman culture is collapsing in the post-buble Japanese economy and currently Korean and Chinese fashions are all the rage there. Who knows what systems Japanese games will reflect in ten years?