Friday, June 23, 2006

One Step Closer to World Domination


Clifton and I were sitting at the Hole in the Wall one Friday evening after work. After our CRAZY, long days at work we would spend just about as long at the Hole dissecting everything that was happening while solving all of the problems of the world. Good times.

It was this Friday evening at the Hole when Gene introduced us to Aubrey who turned out to be one of the oddest men I’ve ever met. Aubrey is an older guy, a UT professor of physics and mathematics, and the original Austin hippie, complete with the long, dirty beard, tattered tie-dyed t-shirt, and a wild look in his eyes that hinted of being way too smart and way too stoned. That night, Aubrey entertained Clifton and me for hours talking about how math will ultimately solve all of the mysteries of the universe. He was totally engaging.

At one point I asked Aubrey if he thought there would be life on Earth one million years from now. He said “yes” but it won’t be life as we know it – it will be artificial life. That led to a completely sci-fi conversation about computers and how man (via math!) is building smarter and smarter machines and how someday we are going to build the right programs that will allow computers to take over the computing – in other words, computers will begin to “think” and “make decisions” for themselves. Aubrey believes that when this happens, over time, computers will begin to “evolve” on their own, becoming more and more intelligent. Then, in our effort to foil mother nature, as we humans continue to try to live longer, healthier, stronger and smarter, man and machine will ultimately fuse in some “bionic man” kind of way (Aubrey used a much more sophisticated explanation than that!), and artificially enhanced life will be born.

Shiner Bocks aside, it was one of the most fun bar conversations I’ve ever had.

But it got me thinking. Maybe hippie Aubrey is right. Will part of our natural human evolution be the incorporation of technology to the point that it ultimately takes over and a new life form is produced? Homo Erectus, Homo Sapien, Homo Machinetus.

Actually, I have proof that this day is drawing near.

I am fascinated with the “thinking” responses of on-line websites that are based on my choices. Actually, what I’m really fascinated with are the people who developed the algorithms that the computers’ programs use to make educated guesses about me. You see it everywhere, now.

Example number one. I rent my movies via NetFlix, the wonderful service that allows me to rent as many movies as I want, to keep for as long as I want, for the low, low monthly price of $9.99. Each time you go back to that website to add another movie to your queue, NetFlix tells you what you want to see based on what you’ve already rented. Sometimes I think I’ll trick it because my movie selections are so completely diverse. But – most of the time – it gets it right! Based on three completely diverse movie choices I made, 1) The Passion of the Christ, 2) The Blue Butterfly and, 3) Capote, NetFlix suggested that I would like 1) Celsius 41.11 which counters Michael Moore’s left wing rhetoric in Fahrenheit 9/11, 2) Tuck Everlasting about a reclusive family in the 1800’s who discover the fountain of youth (also one of my favorite books of all time), 3) Braveheart about the leader of Scotland’s independence movement and 4) Dreamer about a broken down race horse who makes a heroic comeback. It was right – I’d like all of these movies. Now, I’m smart enough to realize that somehow NetFlix’s algorithms include common identifiers like directors, actors, and story themes, but I also think they somehow must tag more general characteristics like conservative, liberal, romantic, independent, etc. It’s pretty amazing to me that the system processes all of the “tags” associated with my movie choices and from that information decides what else I might like.

Same with example number two, When you buy a book or something else from that site, you’ll get their famous line, “Others who bought XXX (the thing you are buying) also bought XXX”. So, not only are their algorithms figuring out what you like, but Amazon’s marketers have honed in on our competitive desires to “keep up with the Jones’s”. So, if “others” are reading Book Y because they bought the same Book X that I’m buying, then I’d better buy Book Y, too! Weirdly, I usually think I’d really like Book Y. is example number three. A website devoted to “shared space” for people to display their digital photos – and they’re not really trying to sell me anything – it’s all free. Their algorithm is something they call “Interestingness” where they sort through the thousands of photos people upload onto their site every day and pick the most interesting to display/rotate on their home page. And the photos they pick are really cool! I’m guessing the programming equation must include something about the image’s lighting, or shapes, or composition, or something obscure like that, but every picture Flickr has chosen for the home page is totally right on in terms of its “interestingness”.

PBS stations are trying to tap into this “personalization”, as we call it. If you visit our website for information about certain programs we want to send you an enewsletter every week promoting programs of a similar nature that you would enjoy. This would all be automated through the programming.

You see this stuff everywhere online.

Maybe it’s all coincidence. Maybe it’s all a big scam. Maybe it’s pure randomness developed by savvy marketers to get me to buy more things. But the science/math behind the process amazes me, and I think it’s the first step towards computers being able to “think” on their own, just like hippie Aubrey told us over beer at the Hole in the Wall on that Friday years ago.

So, will there be life on Earth one million years from today? Mind boggling stuff to think about on a Friday...


Post a Comment

Subscribe to Post Comments [Atom]

<< Home