On Google’s Perception of Semantic Content

Fitting, isn’t it?  Advertising built the bridge to A.I.

Today, I gmailed myself a prospective blog entry.  Google’s “Sponsored Links” to the right of my email, while not “comprehending” (or reflecting) the entirety of the entry’s content, at least picked out some salient aspects it found marketable.  Picked out.  Isn’t this what we do, “pick out”?

Google is working on a “Universal Translator” for real-time translation of spoken language in cellphone technologies.  Suppose I only speak English, and you, French.  I speak in my language.  There’s a few-second lag.  You hear what I’m saying in your language.  While such technology does not yet exist, ponder for a moment what algorithms it requires.  It requires semantic recognition.  You can’t just translate “word for word” from one language to another.  You’ve got to get the “gist.”  Talk about language games.  You’ve got to know what the game is.  What the stakes are.

For many years, I’ve been interested in intelligence.  I’ve coined a predictive aphorism I hope catches on:  “After the Information Age comes the Intelligence Age.”  These days we’re inundated with information.  The recent commercials for “Bing” (the “Decision” engine) highlight this fact.  Ready to move beyond the “Conduit Metaphor” for human interaction?  Consider this:  when you and I speak, you’re not just “packaging” “data” that gets “encoded” and “transmitted” through language, which I “unpack” and “decode” and understand.  Rather–especially in the best instances–listening to you gives me new algorithms for thought.  I resonate with the intelligence in your communication and learn from it.  Moving beyond Chomsky’s “Generative Grammar”…this is Deleuzian “Generative Pragmatics.”  Our intelligences interact…not just our sets of information.

Advertising is building the bridge to A.I.  My guess is that the algorithms used to generate the “Sponsored Links” on your gmail page are pre-cursors to the “Universal Translator” now in development.  Google is listening.  It’s listening to your emails.  It has learned that certain keywords are indexes of marketable content.

Is that so different than what we do?  We’re all phenomenologists.  We all live in a lifeworld, a “lived world,” a world that we and only we can know what it’s like to live.  When someone is talking, you’re highlighting and hiding, “picking out” semantic content that fits the goals and structures of your lifeworld.  Afterward, you can summarize what topics were discussed…and the story you tell of conversations past will be rooted in the interests you took…and are taking.   Google’s interests happen to be to market you on products and services.

The orchid “advertises” to the wasp.  Fitting, isn’t it?


One thought on “On Google’s Perception of Semantic Content

  1. The aim of education is to teach the young how to think rather than what to think.* As the focus shifts from the what of data to the how of processing–without changing the metaphor–new capabilities emerge. Any good parent will tell you, this is the joy of rearing a child. Any good child will tell you, this is the joy of growing.


    * Clement & Lochhead, Cognitive Process Instruction, 1980.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s