22 January 2006

The spam problem part 3: Objections to attention bonds

In part 1, I enumerated the approaches to spam eradication I was aware of, and explained my conclusion that the only approach which will work is an economic approach. In part 2 I discussed various options for tackling spam economically, ending with the one I think would actually be acceptable and useful: attention bonds.

Now I’ll run through (and shoot down) a few of the objections commonly brought up when the possibility of involving actual cash in e-mail sending is raised.

  • What’s to stop me subscribing to thousands of e-mail lists and raking in free money? Mailing lists will naturally expect you to whitelist them so they can send you the traffic you’ve asked for, for free. If you don’t, you won’t get any mail from the list, and hence won’t be able to make money from it.

  • This will destroy free mailing lists! No it won’t; see above. Mailing lists will simply refuse to post bond money, and if that means the recipient doesn’t get the e-mail, that’s his problem.

  • Nobody will pay to send e-mail! Nobody who uses e-mail responsibly will need to pay anything to send e-mail. Your expected cost to send e-mail will remain zero.

  • What about people who are assholes and push the spam button to take my bond money, even though what I sent them wasn’t spam? Don’t send them any more e-mail. If they send you e-mail, retaliate by pushing the button on them. So maybe you’re down 10¢ because your brother was a dick, so what?

    But what about people who post really high bond requirements, like $5 a message? Don’t send them e-mail. If they want to price their time so highly that you don’t want to risk having to pay for it, that’s their loss.

  • What’s to stop spammers refusing to pay? The digital cash is sent up front, along with the message. No cash, no e-mail delivery.

  • What about open relays, botnets and 0wn3d machines? If someone breaks into your computer, they may rack up postage charges. They may also steal your personal information and rack up credit card debts in your name too, which is far worse.

    Currently, when someone runs Windows and doesn’t pay any attention to security, everybody else pays for the cleanup. With this system, the owner of the machine would pay. I see that as a feature, not a problem. If users started to get hit by $1000+ bills when their machines were 0wn3d, we’d see Windows users having to buy insurance to cover the cost of their OS choice—which again, would be a good thing. Many businesses already have computer intrusion insurance; and yes, insurance companies charge more if you run Windows.

    Having said that, if you think we should continue to hide the costs of Microsoft Windows, then the solution is simple: make it so that you have to buy the e-postage in advance. So you’d pay your ISP a deposit, and when your e-postage deposit started to run out, you’d get a warning to refill it. This is exactly how my mobile phone plan works with respect to sending SMS messages.

    • I don’t think anyone should have to pay to send e-mail. No problem—simply set your bond fee to zero, and everyone will be able to send you e-mail for free, no matter what.

      No, I mean I don’t think anyone should have to pay to send you e-mail. Tough luck, that’s my decision to make, not yours. If I want to accept e-mail only through a web form that takes credit card payment for a $5 fee, I can do that, and there’s not a damn thing you can do about it. This is just a less objectionable version of the same basic idea, modified to allow most people to continue e-mailing for an expected cost of zero.

    • Won’t everyone need special new software? Eventually that’d be good, but for the interim it could be implemented server side: your e-mail service provider (which may not be your ISP) could let you set up a wallet of postage to be applied to your e-mail. Messages with costs exceeding the configured limit could be returned undelivered, with a note explaining why, and a way to send the message again with a special keyword to accept the higher bond fee.

      This ISP-based interim solution could also allow users to set limits on how much they should be able to spend per day or per week, avoiding the $1000+ 0wn4ge scenario mentioned above.

Finally, let me outline how this scheme might be introduced gradually.

It would start with a number of e-mail service providers (ESPs) beginning to offer the service. They would filter your e-mail into two folders, based on whether the e-mail was bonded or not. You would have a web, e-mail or IMAP interface to request that their server keep the bond; otherwise, a certain time period after your reading the mail, it would be refunded.

For instance, I could have two IMAP folders, INBOX for bonded e-mail, and Unbonded INBOX for the rest. To collect the bonds, I’d drag the e-mail to the Spam folder, and the ESP’s server would do the rest—it would check the folder for messages needing processing, perform the necessary digital cash transactions, and put the collected cash in my wallet.

For sending, similarly the ESP would provide a web interface for me to purchase bond cash via credit card (or have it charged to my regular bill), and set limits for how much bond money to send. Limits could be both per-message and per-day.

Note that ESPs don’t need to be ISPs. I use pobox.com as an e-mail service provider; they could offer me this new bonded e-mail service, and I wouldn’t need to wait for Time Warner to get their act together. (Just as well, since they don’t even offer IMAP yet.)

Brad Templeton outlines a mostly similar system. However, he says he now disavows it, and thinks it’s unworkable. His objections are more thoughtful than the ones above, and need more detailed discussion, so I’m going to save them for later elucidation.

© mathew 2017