Slashdot Mirror
Ron Rivest Suggests Probability-Based Micropayments
Karl J. Smith writes "Rivest has solved the micropayments problem with encryption and statistics. You throw away some transactions so that you don't have to pay bank fees, and process the rest. Hiawatha Bray has written an article and Rivest's new company is PepperCoin."
Yes, that is the way to make micropayments take off: patent them.
Ok, randomization has its uses, but what advantage does it have over just waiting till the micropayments sum up to $10 and sending them then?
Same old, same old?
"To any truly impartial person, it would be obvious that I am right."
this looks like hidden advertising to me but i won't argue that point....
and it's based on 'patent pending technology' that is somehow acceptable by slashdotters (see here for more info)
this sounds like a lot of marketing hype. why not just have a company that processes micropayments in mass -- if i buy 10 songs for $1.00 each from 10 record labels during 3 months i should be charged $10 as soon as it is profitable to charge me, possibly at the end of the three months, possibly after my tab is at $5.00. i think this is basically what happens with peppercoin but in a more complex, mathematically obtuse way.
finally, what's up with all the hot women on the peppercoin page? it's like i'm supposed to be able to buy them with peppercoins.
fear is the mind killer
Okay.. from the merchant's side.. he does not wanna mess with trying to account for a 5 cent sale.. so lets calculate the a 0.005 probability ( thats 5 cents out of 10 dollars ) and assign that probability to a ten dollar token, that the token is any good. So, in effect, the merchant is gambling he is going to get paid - in this case, for the sum of 5 cents, he accepts a 0.005 probability he gets $10. Basically, its just like gambling, where PepperCoin is the "house". But over millions of transactions, statistics would approximate the same return to the merchant as if he tallied all the micropayments.. but the merchant does not have to worry with millions of tiny payments, he works with thousands of larger consistent payments. And is willing to accept the accounting simplicity as tradeoff against any probability error, as well as the overhead of the "house cut". This technique allows the processing of billions of payments without keeping detailed records on each... the only thing going through is the statistical averages of who gets paid what.
Well anyway, thats my *understanding* of how this thing works...
One neat thing is that it appears any identifying information to the purchaser would be lost in the "noise". comments invited.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
I have only read it quickly, but there seems to be no mention of the way PepperCoin will charge the customers. Since the PepperCoins' value is transferred from PepperCoin to the merchant and this transaction is "optimized", the other transaction (PepperCoin <=> customer) is important. It seems to me that this would only(?) work with a pre-paid amount (otherwise the customer would have to purchase frequently enough to be charged for several transactions at once), so the claim from the article: Letting consumers buy hit music recordings for a buck or less, without charging $10 a month in subscription fees, could be just the thing to ignite the micropayment market. is questionable.
"I love my job, but I hate talking to people like you" (Freddie Mercury)
I think if you randomize you will get a chance to fudge some data; I mean, if in the end your average price of item turns out to be like 49.68 cents averaged over long term, you will have a very unlikely chance of noticing this discrepency. especially most (ALL?) financial software rounds to the cent.
At the same time, the above is assuming that EVERYTHING is 50 cents. Now, imaging there are things costing different amounts of money, and calculating if papercoin is ripping you off that 0.3% becomes difficult if not impossible.
Now, of course, I can't quite figure out how does papercoin charges the consumer. That's really weird because THEY can't be hit with the 25c charge everytime either or they will go under; so they will either have to
1) act like a bank / paypal and have you keep a balance.
2) wait until your "sum" is large enough and charge it all at once.
both have serious problem.
Of course - this entire thing is really a credit card system problem, that can really only be solved by the credit card companies - but they seem to have no incentive to do so, so... we might be stuck here for a while.
My life in the land of the rising sun.
For the user, sign up for a PepperCoin account, providing your credit card number, and when you want to make a purchase:
The token is a digitally signed token with the merchants "name," the consumer's "name," the amount of the transaction, and a value of either $0 or $10 (to the merchant.) Your PepperCoin account is charged $0.50.
The merchant, upon receiving a token, sends you the product, and if the token is worth $10, keeps it for later.
At the end of the [day / week / month / quarter] send all the $10 tokens to PepperCoin. PepperCoin sends back the money for the total value of the tokens. What you'll find is that (money received) / (total number of tokens collected) is $0.50. The merchant will be charged a fee for the service, so you might see something like $0.45 per purchase (10% fee.)
Back to the consumer ... over time you'll accumulate $10 or more in purchases at which point your credit card will be charged. If, let's say, 6 months elapse, and you still haven't accumulated $10, you'll be charged your current balance.
See ... PepperCoin makes about 10% of all the purchases minus the cost of credit card transactions to the consumers (about 5%), the merchant gets $0.45 instead of $0.20 on a $0.50 purchase, and the consumer is charged dollar-for-dollar what they spent.
--- Jason Olshefsky
Karma: Poser (mostly affected by adding this line long after everyone else did)
... it's the credit card company charging so much per transaction. Why work around that problem?
The "market" for credit cards is skewed because the transaction charge is applied to the merchant rather than the purchaser. If the charge did come direct from the purchaser, the purchaser would choose a credit card that offered the lowest charge. As it is, the merchant has no choice (other than saying "I don't accept Amex), so competitive pressures don't apply.
Peppercoin-type operations will further mask the skewed market - we will all end up worse off; except of course for the Visas and MasterCards of this world.
Rob.
My guess is this system was likely not designed for use by run-of-the-mill merchants with transaction volume below the millions (and conceivably billions). Like many have pointed out, your typical store merchant would laugh at the prospect of roulette-based revenue.
This system was designed to solve the problem of handling billing and payment collection for A LOT of transactions per unit time. Think NASDAQ. Think VisaNet. Think McDonald's-years. Think pay per wireless packet, a concept routinely floated by Rivest's MIT colleagues including Dr. David Clark.
Coupled with a computationally efficient token verification scheme, I could see how this system could turn standard billing practice/procedures on its head, provided the big corporations have enough smart people in their stables to say, "Rivest is right." For instance, if my statistics memory serves, this system should effectively enable stepless billing (without increments or round-off issues) - in other words, finest-grain discrete-time pro-rating for services provided, tunable per application to some arbitrary epsilon.
I think music downloads are a red herring. It's entirely possible that PepperCoin will never see the light of day as a consumer payment service. But I'm very curious to see what the world's largest accounts receivable departments have to say about it.
BUT, not anytime soon, and you've identified the exact reason why: peppercoin patent monopoly. No reasonable merchant nor consumer should bet on a scheme that locks you into one vendor, especially for something as vital as your very revenue source. We like money because it is 100% transferable -- I can get it from anyone willing to trade with me. Credit cards are also competitive -- if I don't like Visa, I can try AmEx or Discover or MasterCard, and most vendor's have a single machine that can take any of the above. If I don't like peppercoin, there's no alternative I can switch out for -- the system is closed, patented, and sealed. Sure, there are other micropayment schemes that have lived and died, but if I wanted to start a peppercoin-compatible service, tough luck; it'll be at least 17 years before we get a legal shot at that.
Explain to me, o banks, why it costs you $2 to give me money from my own accout? Why it costs you $10 to wire transfer some money from one account to the other? Why it costs $1 to give me a balance statement? Why it's 75c to use your ATM card at anywhere but a supermarket? These are just the costs for consumer-visible transactions; the costs of using a credit card or ATM to the business owner must be similarly padded.
These are database transactions. They happen almost instantly and they consume resources at a tiny fraction of the cost we're being charged. It's electricity being sent over a wire; the marginal cost is so close to zero you need calculus to describe it. This is why micropayments don't work yet, and elaborate schemes like this randomization are even necessary at all. PayPal and similar systems have eliminated these costs, but "real" banks refuse to, because they make an assload of money off of charging for the movement of electrons.
It's rare that you're presented with a knob whose only two positions are Make History and Flee Your Glorious Destiny.