Proposal: Augment computer-to-computer Internet phone with local telephone-to-Internet-phone gateways to create a new telephone-to-telephone long distance network via the Internet. Result 1: No computer is needed for cheap long distance or international voice phone calls. Result 2: ecash gets its "Killer Ap", a service that (a) people want to buy and (b) requires micropayments for economic feasibility. (Details below.) Result 3: Creates a large, decentralized market that is difficult for anyone to regulate and that provides many opportunities to make money. Obvious Nonresult: Achieving privacy in long distance or international voice requires further development. Hardware and software beyond that in an ordinary telephone is needed at each end. Who Can Do It: ISPs (Internet Service Providers) are in the best position to operate nodes of this decentralized network. They have the Internet connection, the phone banks, the technical expertise, and some business experience. Telephone <---> ISP <---------> ISP <---> Telephone Scenario (Result 1): Granny Smith in Paducah wants to chat with her grandson Adam Smith, who is studying economics in Hong Kong. From her ordinary touch-tone phone she makes a local _voice_ call to Cheap Speech, Inc. and at the prompt enters the phone number for Adam. Cheap Speech finds that TalkToMe Ltd is a network node in Hong Kong within a local phone call of Adam and, through the Internet, establishes a connection. Then it patches in the voice call from Granny as TalkToMe dials out and reaches Adam. Granny Smith <--> Cheap Speech <-----> TalkToMe <--> Adam Smith Payments (Result 2): More than half the cost of running a traditional long distance voice business is for handling the billing. The cost of the technology to provide the service is thus less than the cost of the billing! Cheap Speech and TalkToMe cannot provide cheap voice service, as in the example above, unless they use a much more efficient payment system. That is why the low transaction cost of ecash is so important and that is why this message is being sent to cypherpunks rather than a telecom list or news group. So who uses the ecash? Surely Granny and Adam Smith can't shove it through ordinary telephones. They just want to get ordinary voice connections with ordinary telephones as they have for many years, only cheaper. Here is how I think ecash fits in: Granny Smith --> Cheap Speech Granny pays Cheap Speech through ordinary means, probably by monthly credit card charge, which is common for ISPs. The overhead for this probably is only around 5% (plus staff for account setup and support) because the ISP already is set up for that kind of payment system and Granny Smith is a local person providing repeat business, not a bad apple on a spending spree with a stolen credit card number. Cheap Speech --> TalkToMe Cheap Speech and TalkToMe do not have any previous business arrangement with each other and probably never have done business with each other before. They also are on opposite sides of the world. But Cheap Speech consults a Rating Service that says TalkToMe is reliable, so Cheap Speech sends a small amount of ecash to TalkToMe to open a connection and dial out to Adam. Since ecash clears instantly, TalkToMe does not need to know anything about Cheap Speech. No international billing network is needed. Little overhead is incurred. TalkToMe --> Adam Smith TalkToMe provides this service in exchange for the ecash from Cheap Speech. The main problem I see with this scheme is that Cheap Speech may have a cash flow problem. Payment to TalkToMe must be made immediately whereas payment from Granny Smith may take a couple of months to arrive. If, however, the volume of incoming calls (ecash coming in) matches the volume of outgoing calls (ecash going out), then the cash flow will balance out. Also, Cheap Speech could offer Granny a discount for prepayment. Regulation and Profit Opportunities (Result 3): Big, centralized organizations are big targets for lawsuits and regulation. They are the "deep pockets" lawyers love to pick. They are the leverage points for power-hungry politicians. But the Internet was designed to survive nuclear war. It was designed to route around outages, no matter their cause. A decentralized, Internet-based, international network of small, independent voice service providers could enjoy similar advantages. Many companies are creating _computer-to-computer_ Internet phone software. See the NetWatch Top Ten - Voice / Video On The Net at URL: http://www.pulver.com/netwatch/topten/tt24.htm We can depend on that technology to improve. I do not know of any companies who also are providing interfaces to ordinary telephones and designing a decentralized network with an ecash-powered payment system. That is our advantage. But the network described above has many limitations. These are opportunities to make money by selling your solutions to customers. Examples: It needs a directory and rating service for voice providers such as Cheap Speech and TalkToMe. (Perhaps Raph Levien will expand his rating services to a new domain?) The network will not be useful until it has many nodes. That is partly why this proposal is being sent to a large audience. Cypherpunks who cut their teeth building a remailer network perhaps can build a voice network, too? Think of it as a challenge. People who travel a lot will want cheap long distance telephone service while on the road. How do they arrange that if their only contract is with a service provider near their home? We need a more flexible payment protocol than the one described above. Enhanced services - fax, teleconferencing, time-delay and retry, phone mail, and collect calls. Multiple brands of ecash will require conversion services. Currently Mark Twain Bank's offering looks best, but when ecash succeeds, expect many more. Eventually, the network may work best with ecash denominated in currencies other than government-sponsored fiat currencies. Conclusion: If this idea is so good, then why am I telling you about it? (A) I cannot do it all myself. (B) Feedback on the idea from knowledgeable people is valuable. (C) If it succeeds, it will create a market with opportunity for plenty of people to make money, including me. (D) We have a limited window of opportunity to get this done. My guess is that we have roughly one year before other people embed their solutions into the Internet and financial system so deeply that this network will be locked out from commercial success. According to the theory of increasing returns, whoever gets to market first usually gets the market. (E) Even though the decentralized structure should help reduce exposure of the network-as-a-whole to harmful lawsuits and regulations, individual local providers may face sanctions from regulators of the FCC, state PUCs, national Telekom monopolies, etc. once they realize what's happening. (Operators of cypherpunk remailers endure similar risks.) The faster the system can be developed and deployed, with a large number of satisfied customers, the more widespread support it will have, the harder it will be to stamp out, and the safer life will be for everyone involved. It would be nice to be able to say "The Genie is out of the bottle." CW