I give credit to Dave Winer of Userland Software for inventing web logging, and I think the idea then was to publish, to share your thoughts with everyone else. But most people’s thoughts aren’t really worth sharing. Most web logs are little more than lists of annotated bookmarks and the value of those bookmarks can probably be best derived through a web aggregator, in which case people would be writing not to be read but to be counted, which isn’t nearly as much fun.

A lot of this comes down to production values, which is a subject those in the web log world tend to ignore because it is to their advantage to do so. There is a lot of bad television, but its packaging is such that we still seem to sit through the shows. Network TV spends perhaps $500,000 on an hour. How much do you spend on each web log entry? No wonder most web logs are so boring.

But Joe Reger wants us to not think so much about the web log publishing model and instead use the technology — preferably HIS technology — as a personal freeform database with analytical tools to take the measure of our own lives. Here we’ve been thinking about web logs as a way of reaching out to the world when they may be as much or even more useful reaching into ourselves.

I think he is onto something. Personal data mining means that I’d be mining my own data, learning about my own little world. If the FBI wanted to do that (they probably do) then I’d be opposed, but personal data mining offers personal payoffs. Imagine if your web log chirped up one day suggesting out of the blue that maybe, just maybe certain trends in the entries were suggesting that you need a vacation or your business is in peril or your kid is abusing drugs or that you probably have cancer. If such knowledge was hidden in your web log data, wouldn’t you rather know than not?

Read the rest in i, cringely’s column.