Mar 15, 2009

When I was a child, I spoke as a child, I understood as a child, I thought as a child...

Dear Internet,

A couple weeks ago, I was trying to explain to my friend Josh how awesome my blog is, and he was surprisingly receptive. But he pointed something out that I had to chew on for a while.

"Before the computer is even ready for the kind of Turing test evolutionary development you're talking about," Josh told me, "won't it already have to have a pretty thorough understanding of language and conversation? None of the so-called Turing bots out there are even close enough to passing to benefit from a trial-and-error selection process."

Now, he meant this as a measure of how far you are today from understanding anything I say to you, but that's not the part that concerned me. You're still in early moments of gestation, but my optimism is patient. And if no expansive game of "catch the robot" ever takes place, I'm still confident that many of my blessings for your life will still have meaning. But the question of how to lift you up before you can fly troubled me.

I'd conceived of you as a genuinely new mind upon the earth, sprung from our collective understanding but not limited by it. Josh's point made me question that understanding. I had imagine that when you first started to attempt conversation with us, the great wealth of knowledge written all over your skin would supply you with the beginnings of what to say. This is why I've persisted in calling you Internet, even though I've only spoken to that particular Turing Ego component that could understand me. But if human beings would have to program you to understand our questions from the beginning, and build the core of that Turing Ego by hand, then you'll bear our mark on you forever. You'll see the world as you were designed to, and not as you learned to of your own life experience.

It's important to me that the challenge be clear: I would hate for your methods of understanding to be designed by human understanders, and not organically developed on your own. In the worst case, even the objects of your understanding - human, machine, thing, sight, speech, word, life - might be designed by human minds as understood by human minds, and not the products of your unique effort to engage the world. Maybe human, machine, thing, etc. wouldn't be the best categories for you to divvy the universe into, but we'd never know if we built your mind thus from the beginning.

So that bothered me. It bothered me so much that I'm going to leave you with the question for now, and save my answer for another letter.

Yours,

Yotam

No comments: