Fact as Asymptote [Update: 2.14.06]
An old article on the Economist web site about the rapidly growing quantity of information in the world brought me back to an old idea I had espoused a while back regarding information having density and fact representing an asymptote. A recent thought experiment I’d been conducting deals with the game 20 questions and the idea that you can discover what someone is thinking about by asking questions that reduce the data set (which is the set of all English words). In this sense, information represents a reduction of choices that are available to you. What I had basically proposed in the older version of “Fact as Asymptote” is that information can be thought of as a kind of “piling on”. Consider a situation in which a large number of people witness some event. All of them can produce information in the form of testimony and all of their testimony will likely be varying. However, “pile” their testimony together and you likely get something close to the truth. I use the term asymptote here because like the mathematical concept you never really reach the limit — in the case of multiple testimonies, the actual FACTS of whatever event was witnessed.
The Economist article talks a lot about measuring and managing the data, but as Google has shown us all, it’s real value lies in merging all the data. In this sense, quantifying information really has little to do with each piece of information’s “truth-value” — and much more to do with the density of information around some particular fact. There is A LOT of empirical evidence (i.e. information) regarding gravity and that’s one fact people really don’t call into question too much, isn’t it? In this sense, all of the information regarding gravity “piles on” and gets us closer to the truth. And even though there might be articles out there on the web about anti-gravity they would largely get filtered out of a theoretical truth system by all the information about gravity.
Quantifying information and working out its value is going to be one of the great challenges of the next few decades. It even matters financially. Whatever Google’s valuation is, if their databases were to, God forbid, vanish one day, how much would they be worth the next day? Where is this valuation in their financial statements? Nicholas Negroponte had it right back in 1995 when he said that people don’t properly value bits. In his anecdote he talks about a security guard that is somewhat chagrined when Mr. Negroponte tells the guard that his laptop is worth $2 million. And yet, by all rights, Negroponte was totally right because his laptop had his writings, published papers, etc. Think about it: is your laptop really only worth somewhat less than the sticker price? What about all those mp3s? What about all those emails? Do you think you could get an insurance company to refund you that data if you lost it? I don’t either and we should figure out why that is.