The main problem with data is that it's easy to copy. In fact, sending it from one place to another is essentially an act of replication. The mv command is alien to most people's experience of the Internet. They may use it every day (hardly realizing it) within their own filesystems, but between separate systems on the Net, their experience is that of replication, not relocation.

This alone makes control of one's data problematic, especially when the first-person possessive voice isn't quite right. “My” data often isn't. For example, take profile or activity data kept by a service provider. It's from you and about you, but you don't own it, much less control it. Transaction data is created by a buyer and a seller together, but the canonical form of the data is what's kept by the seller and provided (by copying) in the form of a bill or displayed on an encrypted personal Web connection. Sellers don't go much further than that. The idea of sharing that information in its raw form, either during a transaction or later on request by the buyer, is alien at best to most sellers' IT and legal departments. As John Perry Barlow put it in “Death From Above” (way back in 1995, w2.eff.org/Misc/Publications/John_Perry_Barlow/HTML/death_from_above.html), “America remains a place where companies produce and consumers consume in an economic relationship which is still as asymmetrical as that of bomber to bombee.” In fact, this is still true of the whole business world.

Yet, internetworking of that world brings a great deal of symmetricality to it, imposed by the architecture of the Internet and its growing suite of protocols. The bank that used to occupy the most serious building on Main Street—or a skyscraper in a big city—is now but one location among a trillion on the Web. Yours is another. The word “domain” applies to both of you, even if your bank's “brand” is bigger than yours. Of your own sense of place and power on the Net, the words of William Cowper apply (www.bartelby.com/41/317.html): “I AM monarch of all I survey; / My right there is none to dispute...”

Yet, as William Gibson famously said, “the future is here but not evenly distributed.” Bomber/bombee power asymmetries persist in the B2C (business-to-consumer) world of everyday retailing. When you buy something, the transaction data in most cases comes to you only in the form of a receipt from the seller and a bill from the credit-card company. Neither is offered in formats that allow you to gather data on the spot or later over a secure Net connection—not easily, anyway.

If we could collect that data easily, our self-knowledge and future purchases would be far better informed. In fact, collected data could go far beyond transaction alone. Time, date, location, duration, sequence—those are obvious ones. How about other bits of data, such as those involved in dealings with airlines? For example, your “fare basis code” (HL7LNR, or some other collection of letters and numbers) contains piles of information that might be useful to you as well as the airline, especially as you begin to add up the variables over time.

A marketplace is no better than the knowledge and practices that buyers and sellers both bring to it. But, while the Net opens many paths for increasing knowledge on both sides, most of the knowledge-gathering innovation has gone into helping sellers. Not buyers.

Today, that's changing. More and more buyers (especially the geeks among them) are getting around to helping themselves. In particular, two new development categories are starting to stand out—at least for me. One is self-tracking, and the other is personal informatics.

Compared to its alternative (basically, guessing), self-tracking is “know thyself” taken to an extreme. Alexandra Carmichael, for example, tracks 40 things about herself, every day. These include mood, chronic pain levels, sexual activity, food intake and so on. She's a star in the Quantified Self community (www.kk.org/quantifiedself), which is led by Gary Wolf and Kevin Kelly. Among topics at QS meetups are chemical body load, personal genome sequencing, lifelogging, self-experimentation, behavior monitoring, location tracking, non-invasive probes, digitizing body info, sharing health records, psychological self-assessments and medical self-diagnostics, to name a few.

Now, would any of these be extreme if they were easy and routine? Well, that's the idea. ListenLog (cyber.law.harvard.edu/projectvrm/ListenLog), one of the projects I'm involved with, doesn't make sense unless it's easy, and unless the data it yields is plainly valuable.

This brings us to personal informatics, which is a general category that includes self-tracking and extends to actions. All this data needs to live somewhere, and stuff needs to be done with it.

In the commercial realm, I see two broad but different approaches. One is based on a personal data store that might be self-hosted by the customer or in a cloud operated by what we call a fourth-party service (serving the buyer rather than the seller—to differentiate it from third parties, which primarily serve sellers). As Iain Henderson (who leads this approach) puts it, what matters here “is what the individual brings to the party via their personal data store/user-driven and volunteered personal information. They bring the context for all subsequent components of the buying process (and high-grade fuel for the selling process if it can be trained to listen rather than shout).” The other approach is based on complete user autonomy, whereby self-tracking and personal relationships are entirely the responsibility of the individual. This is exemplified by The Mine! Project (themineproject.org/about), led by Adriana Lukas. As she puts it, the difference between the two approaches is providing vs. enabling (www.mediainfluencer.net/2009/04/enabling-vs-providing).

Either way, the individual is the primary actor. As distribution of the future evens out, the individual has the most to gain.