Surrounded, as we are, by the digitizing of almost everything, it is easy to get swept along by the tide. As digital barrier after barrier comes crashing down, it is easy to succumb to the notion that everything can, in principle become digital in some shape or form. Some of a more theoretical bent often point out the possibility that the entire universe is really just a great big computer following some fundamentally simple, deterministic rules[1]. It follows, therefore, does it not, that we can encode anything in a form that a computer can understand?

No, it does not follow. Well, at least, not in any useful sense of the word "yes". Let me explain what I am thinking.

Humans are advanced social animals. We have very well developed faculties and strategies for getting what we want. These are not necessarily nice faculties...It is not always a matter of just using our opposable thumbs. Hmmm... I'm struggling to find the right words here...Let me just blurt it out. People all over the world make extensive use of untruths in order to get what they want. These range from the little white lie through to the gargantuan whoppers that change the course of history. There. I've said it.

Even the most innocuous human-to-human interactions have to be evaluated by the participants in the light of the fact that all may not be as it seems. For example, imagine that a person asks you for the time. Are they really just asking for the time? Is there some other goal? To start a conversation? To sell you a better watch? To buy your own? To steal yours? To see if you speak English? etc. etc.

In our digital age, more and more of these human-to-human interactions are taking place in digital environments. However, that digitization process has in no way impacted the fact that all may not be as it seems in terms of the intent behind the interaction. Intent does not get digitized when the surface level content of the interaction is digitized. Of course, we could add a digital "intent=honorable" indicator to all interactions but of course that just pushes the problem out one level. We can no more tell if this statement is honorable than the one it purports to reference. A computer saying "trust me" is no more to be trusted than the street hustlers doing the 3 cups and ball routine.

Two examples spring to mind: mashups and wikis. In a mashup, someone is the producer and someone is the consumer. The producer expends resources providing content to be consumed. A consumer may make very heavy demands on the producer. How can the producer tell if the intent of the consumer is honorable? Is there a clear distinction between a well-intentioned mashup consumer and a denial of service attack?

In a factual WIKI, lots of users congregate to collectively maintain the database of facts. In the world we live in - especially the human affairs aspects of the world we live in - separating fact from opinion is rarely uncontroversial. How can a WIKI tell the difference between fact and spin? Between, illumination and misdirection?

I think you will most likely agree with me that there are no hard and fast rules, no simple answers to these questions. And yet I regularly come across technological silver bullets aimed at addressing questions of trust and intent.

'Intent' is not a number people. It cannot be digitized. There is no computation that can be performed to compute the intentions of one computer talking to another. No amount of CPU power or bandwidth or operating system upgrades will change that fact.

[1] http://en.wikipedia.org/wiki/A_New_Kind_of_Science

ITworld.com

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine