From 1993:

I'm distributing this to a few trusted friends in confidence. I just might be serious about it, so please do keep these ideas to yourselves. But do send me comments. If I don¹t try to actually start the following, I might just publish an expanded version in Wired.


by Jaron Lanier

The following three things should be done simultaneously:

Thing one) Make a very cheap set-top box that browses the World Wide Web on TV without needing an operating/window system.

Thing two) Extend HTML (the language of the WWW) so that it can deliver full-fledged applications as pages.

Thing three) Organize an Amway-like business where fiber-wiring kits are distributed by micro-entrepreneurs to neighborhoods. People in a neighborhood sign a boilerplate contract to form a fiber co-op. They give each other rights and responsibilities to route cable through their properties to a neighborhood node, and they collectively purchase the best available high bandwidth connection to the internet at that node. They commit contractually to annual improvement to the neighborhood fiber resource to keep property values up.

* * *

More on Thing One) The first rev of the box plugs into a phone line on one side and a TV on the other. It has some fonts built in that are well chosen for easy reading on TV. It has a remote that can scroll the view on a large mosaic page. The second rev of the box directly plugs into ISDN and third rev into fiber. As the connection gets better, the box get fancier, including larger amounts of RAM, video cameras and microphones, support for higher resolution displays, VR peripherals, etc.. The first rev doesn¹t need much computation power, since it will be bandwidth limited anyway, so it can be very cheap indeed, probably using slightly out of date parts. Note that all it has in its ROM are Internet applications like a Mosaic browser and Eudora;
there is no operating, window, or disk operating system. I believe such a
can be built in moderate quantities for under $50. It is lower-tech than a Nintendo box.

More on Thing Two) A user interface/display kit with a control protocol is also introduced and included in the set top box. So, an application like a word processor or spreadsheet can be running at a remote server while the interface appears in the home. The relevant memory for the application is mirrored (using a bandwidth-minimizing, encrypting protocol) between the server and local memory in the set-top box. The bandwidth requirements between a front end and a server of this kind should be very small, so performance will be comparable to early PCs, even over phone lines. Over higher bandwidth lines, performance should keep up with stand-alone computers. Users who wish to can purchase their own hard disks, but they can also leave their data, safely encrypted, at the server, which might or might not be owned by the
op. This will bring computer access to people who cannot otherwise afford a computer.

Applications can be rented for a very low hourly rate, instead of being bought, and they can come stuffed to order with useful data. Over the term of use, a developer will see as much revenue as if the user had bought boxed software, and there is no piracy. This is a better way to distribute and use software, and it makes the idea of an operating system obsolete. This is the last great Microsoft-sized opportunity for a platform standard.

More on Thing Three) Neither the public nor private sectors have been able to amass enough capital and clout to bring fiber to large numbers of homes rapidly. Gore failed. TCI/Bell Atlantic failed. But what if fiber is installed from the grass roots? Then it becomes imaginable that millions of homes could be wired within a few years. Many lawyers will be put to good use creating enforceable cooperative information easement rights, and court battles might have to be fought against cable interests and the like, but ultimately it is likely that property owners can legally create their own fiber network.

Residents get immediate gratification from the phone-line-based box, which brings them incredible value for very little money. For the cost of a Nintendo, they get hassle-free access to everything on the net, as well as serious computer applications. The widespread use of these cheap boxes will motivate content providers to address the needs of fiber-coop members early on.

Affluent neighborhoods might contract with a specialized firm to install the fiber lines, but most neighborhoods will have ³fiber weekends² where homeowners do the job themselves. This will be possible because of a well designed neighborhood cyberspace kit distributed by micro-entrepreneurs, who also serve to hand-hold residents through the process. Local control makes the process incremental. For instance, fiber can be strung between roofs at first and later dug into the ground. Neighborhoods will be motivated by the micro-entrepreneurs for all the right reasons: to improve the future prospects for their children, to improve their property values, to get access to new business opportunities, to participate in society in expanded ways, and to have fun.

When the fiber is installed, the neighborhood node will probably be connected to the internet by a commercial vendor. At that time, the set-top boxes are upgraded. All the neighborhoods can bid together on goods and services, bringing costs down. The neighborhood node will be improved by a mutual covenant every year. Fiber coops will inevitably become a force that incrementally improves the internet for all. Within a few years residents will be calling up low-resolution but watchable movies on demand, and the quality will improve incrementally with the node hardware and the handling capacity of the net.

This is a case where a cooperative anarchy wins in the free market over business as usual, simply because it costs so much less. Government and corporate thinking on getting America fibered has assumed that a centralized army will be outfitted to do the job. Imagine what a government or a corporation would have spent to get the World Wide Web to the point that it is, as quickly as it has gotten there. The distributed, incremental sweat equity that the community has invested in the WWW can also be applied to hardware infrastructure.

Low income neighborhoods have the most to gain. Right now TVs and Nintendos can be found everywhere. Joining a fiber coop will be cheap compared to those questionable investments. Renters can join a coop and get the low bandwidth boxes in anticipation of a time when they might become owners. They¹ll bring computer and net access to themselves and their kids for a bargain rate in the interim. They¹ll also automatically have a wired organization that might be a political force. If the tide does turn in national politics and there is a movement to support universal access, then the neighborhood coops can only help, in a variety of ways. If not, then at least low income neighborhoods have a fighting chance not to be left out.

If a startup business was to do the above it would make money by selling set top boxes, renting software, and supporting an Amway-like pyramid of micro- entrepreneurs selling neighborhood cyberkits. The startup costs could be as low as, perhaps, $10M to design the set top box, $5M to establish the extended HTML protocol and the technology of the software rental business, and $10M to start up the neighborhood kit business (much of it for legal fees). A second round of capital would be raised to support the marketing and manufacturing ramp-up, but this is fundamentally a very light industry with low capital needs. It seems to me this business could usurp the futures of the telephone, broadcast, and cable industries, the computer hardware and application companies, and most deliciously, Microsoft. But it would do so in a way that wouldn¹t impose style or content on its customers, and that would assure a decentralized, democratic and beautiful information infrastructure, intimately understood and controlled by the citizens who built it and use it.