| Many to Many | 2003-04-09 02:34 4 comments |
 by Flemming FunchHow do we invite democracy to emerge? How does spontaneous cooperation happen? How can humanity self-organize in more useful ways? How can we make our civilization more synergetic? I am here particularly interested in how we can use our existing or almost existing technologies to arrange our information and our means of communicating in such a manner as to invite these things to happen. My instinct is that our society could be organized in a drastically different way, from the bottom up, in a way that will allow just about all of us to do productive work that we are happy with, and in a way that is vastly more productive as a whole. If we just knew how. I believe our information networks could provide some core leverage in getting us there, but they aren't yet. It is kind of like we're almost ready to operate collectively at a much higher level, but we're still communicating with tin cans connected with string, and we all have paper bags over our heads, so we can't do anything very complex together, except for rather clumsily and haphazardly. There are pretty good tools for communicating person-to-person. I can be in continous communication with individuals all over the world when it comes to exchanging words in a fairly linear manner. There are also pretty good tools for allowing small groups to work together. If we know who we are and what we want to do, we can use mailing lists, bulletin boards, WIKIs, chat rooms, project managers and other nice software for working together. We also have tools which allow us easily to start speaking to the world, such as web logs. We can keep our thoughts organized, and easy to access for those who seek them. And we have access to search engines and directories of various kinds that allow us to look for information within practically all information that has been posted publically. And if we're happy with either the most popular choices, or the choices that some of our friends know about, or even with random choices, we'll do fairly well, and there's lots to look at. But that's where the problems come in. There's a disconnect. For a single person or a small group, everything can be connected. You can keep track of all issues or ideas. You can coordinate your activities directly, and you can have a consensus on things. You can even do that amongst a small network of people who have separate weblogs. But when it goes beyond a certain scale, maybe 20 people, maybe 100, the coherence falls apart, and you inevitably get to deal with more of a crap shoot. You speak, but you're not sure who'll hear you. You seek and you'll find things, but you'll never know if you get the complete picture. People who really ought to speak with each other might not find each other. Pieces of information that really ought to be correlated with each other might never be. It is hard to know what is good, better and best. Hard to know what is true and what is not. Hard to see the context of people and information and activities. Mostly it is fuzzy, haphazard guesswork. Or it is just the old fashioned approaches of asking around, doing due dilligence, following your instincts. Not that that is all that bad. But we need something of a higher order for 100s of thousands of people to self-organize in complex ways, resulting in emergent synergies we couldn't have imagined. And, yeah, maybe it just sort of happens by itself. Suddenly large groups of people start manifesting unusual collective intelligence. Then again, maybe there's a key catalytic ingredient we need to invent first, which we haven't. We don't have any very good way of collaboratively organizing information in a way that scales from one person to six billion. Our currently most useful technology for dealing with ALL our (public, electronic) information is the search engine, ranking sources based on their aggregate popularity for linking. It allows most anybody to find *something* of relevance on a given subject, very quickly. That has made it much faster to find answers to certain types of questions. But it doesn't help much in providing structure and context and completeness. Should all information in existence be placed in its logical position in a giant ontological tree? Some people probably has that vision of what is needed. But it all moves too fast, and there are too many disagreements on what goes where, and it isn't all hierarchical anyway. Is it the Semantic Web we need? Well, that will greatly help. But it isn't quite clear how the whole web becomes semantic. And even if every item on the web has markup that explains what it is, and there is widespread standardization of how various things are represented, the biggest job would still be remaining - of how actually to add up all the meaning and make it accessible. There are many helpful in-between kinds of facilities. Directories, ranking schemes of various kinds. Shared topics, trackback, etc. But most are either variations of the ranked search engine approach, or of the facilities for small groups. Nothing that scales continously in between. A list of topics is useful if there are a few dozen, but maybe not if there is 50,000 I postulate that we need nothing less than a technology that allows any piece of information to relate to all other information. Not necessarily all the time, or at the same time, but approximately, or potentially. Like, say I have something to say, or something to sell, or something I'm looking for. I'd like my communication to be there for EVERYBODY. Yeah, yeah, so does every spammer. My point is that we need an approach where that is actually a good thing, and where it is feasible. My communication touches everybody. But everybody has different receptors. So it will only stick in some places. Same thing the other way. I want to be informed about everything that everybody is talking about. Except for that I only want to retain the stuff that actually fits for me. I have my own receptors, which will only allow specific items to stick. Sounds a little naive maybe, and lots of people can come up with lots of reasons why it is silly, and nobody can quite say how to do it. But it isn't silly. If you have a car you want to sell, of course you want EVERYBODY to know about it who might want it and who practically might be able to buy it. And of course you don't need to force it down the throat of all the people who don't want it or who can't possibly buy it. Enormous amounts of resources are spent on delivering communications in the shape of carpet bombing, and looking for information as if it were needles hiding in rows of haystacks. Spammers send out 50 million messages just to find 500 people who might want what they offer. A search engine search might give you 500,000 matches, and you'll grope around in the dark for one that you like, leaving out all the rest. Maybe we can take the consequence of this, and use a comparable amount of resources to actually offer our information to everybody, and actually look through ALL possible matches when we're looking for something. That requires a different technology. A distributed multi-dimensional database, with any number of dimensions in it. A normalized relational database stores an item in just one place. You don't store an address in 1000 places, just because 1000 people need it. That's the kind of thing I'm talking about. But a relational database has just a couple of dimensions, and can only practically simulate just a few more. I want to say what I'm saying just once. But I want it to be in everybody's database. I want anybody who has something to offer to say it only once. But I want it in my database so I can find it, no matter where or how they said it. I don't really want to go search for things, to maybe find them. I want to have them right away. It is a question of dimensions. Things only have to be sent around, and duplicated in redundant or wasteful ways, because there are too few dimensions to work with. If all I know about 50 million people is their address, I'll have to write to all of them to ask if they want my product. Or I just put up a webpage and hope that they find me. But if their needs and desires were in my database, in a useful format, I'd just attach my offer directly to that. In a multi-dimensional space of a sufficient number of dimensions, there is one place for everything. An item might have 10,000 degrees of coordinates to pin-point it, but it is only in one place. Our information world seems so terribly fuzzy only because we try to describe it in 1, 2, 3, 4 dimensions, and since that isn't enough, things just don't seem to fit together. It is a geometrical problem. Our world has acquired more dimensions and we still try to represent it in flatland. There are lots of other things to solve. But one of the main things we need is a way of representing data with arbitrary numbers of dimensions. I'm aware of no existing technology that does that well. Maybe it isn't possible before quantum computing is for real. But I hope it is, and that it is just a matter of algorithms. If we each can succeed at being in a two-way relationship with 100s of thousands of people simultaneously, being able to meaningfully let them know what we want, and being able to meaningfully perceive what they want - that's when something truly new is likely to emerge in terms of democracy and spontaneous collaboration. One reason our information management tools are inadequate in helping us with that is that they're mostly modelled on our physical world tools, which already are inadequate in representing our lives. Desktops, filing cabinets with folders, ledgers, spreadsheets, documents, bulletin boards, letters in the mail. We need something more like a brain, with billions of neurons, each having thousands of synaptic connections to other neurons, and able to make more connections when needed. Many to many. |