Building the Internet on a Flimsy Foundation

About 10 years ago when the Internet was young and people were desperate to navigate the new "information superhighway," we took a wrong turn and ran off the road. We started building the Internet as if we knew what people think, and thus where they wanted to go. We built it on a foundation of "intent."

Don’t get me wrong -- it wasn’t only the early adopter geeks and new age freaks that made this mistake. Society in general skidded off into the dirt.

Intent is difficult to predict before it becomes an action. Here's an example of what I mean. Last night we walked into town for a late night dish of ice cream. Was my intent a result of the advertising I viewed earlier in the local paper? Or a psychological desire to gain weight, and slowly commit diabetic suicide because of repressed oedipal rage? Or was it the result of a political gesture of solidarity stimulated by my support of organic, local farming? Or was I just was hungry?

Damned if I know.

If I can’t figure out why I chose a particular action, how could a stranger, or a computer running an algorithm, hope to do better? Even if they had a complete video diary of my every word and movement, they wouldn’t be able to predict my intentions.

The shear hubris of believing we have the capacity to reliably predict intent, and thus action, has misinformed the last decade of political and economic decisions. Apparently, we "knew" Iraqi President Saddam Hussein intended to attack the U.S. with weapons of mass destruction, and thus justified an unnecessary war.

Apparently, we know every person in the world, in their heart of hearts, prefers a capitalist economic system, and thus seriously underestimated how quickly China’s mixed economy would rise to become a financial powerhouse.

The Internet is built on the same flimsy foundation. Somehow we believe spam can be eliminated by reading every email and determining the sender’s intent. But the best email filters still misfile a few important messages each week. As a result, I have to skim all the email in my "junk email" folder to avoid insulting important correspondents.

Search engine companies believe (as do buyers of the public stock) that they can target ads more efficiently based on invading my privacy and analyzing my last hundred search queries and emails -- and thus charge a premium for each ad served. But last week, while I was seeking information on car recalls, I was flooded by ads to buy the very same lemon from the same company I was investigating. This hardened my resolve to never look in their showroom again. (See Cory Doctorow’s blog: The Future of Ignoring Things.)

The Recording Industry Association of America (RIAA) snoops around our computers to see what music files we're posting and trying to guess our intent. Do we own the track we posted, and are we just backing it up to the net? Are we letting a friend listen, or borrow a track, or are we incenting a stranger to steal? As with child pornography, the RIAA assumes everyone is guilty -- by their version of intent.

The system (and our patience) is getting stressed, and you can watch how intent’s foundation is crumbling. Search engines now present a "universal" panel of results -- some algorithmic, some video, some blog -- because they aren't sure what you are looking for, and because you aren't completely sure either.

There is a practical alternative -- a "transparent" Internet based on actions and responsibility rather than mental telepathy.

A transparent Internet would:

• Add cost to spam. As many people have noted, spam is primarily an economic behavior. If you simply charged a few pennies for every email, spam would wither away. No Bayesian filters, no email filters, no inferences about intent -- just fighting cash with cash. The small cost would produce enormous gain.
• Eliminate anonymous addresses. Then spam, illegal file sharers, and the providers of patently illegal, underage nude photos could all be stopped based on their actions, not their imaginary intentions. These changes would probably require modifying the DNS and SMTP protocols to require verifiable addresses.
• Take cookies out of hiding. Redesign browsers so, if you clicked on a site, you could view on a side panel that reveals "these companies are following your activities." You could either allow this automatic surveillance via cookie, have your browser block it, or trace back through the chain of that cookie to monitor your footsteps across the Web.

These suggestions are all eminently realistic and realizable. A number of standards groups are already working on similar programs and I sense that the momentum is building for change. Like many of the waves that form on the Internet, the rising demand for transparency won’t take the form of new demands for legislation, just increased grassroots pressure that will cause companies to respond.

— Greg Blonder, VC at Morgenthaler Ventures