Chapter 3, in which I attend the Accelerating Change Conference
I'm back from the ACC 2003, which lasted all weekend. I met several intriguing people & learned a lot, mostly about AI & nanotech. I also had some interesting experiences while working around the despicable lack of decent public transit in Silicon Valley. Here's my story.
After the weekly Bash on Friday, Mike Godfrey dropped me off at Tressider Union on the Stanford campus in Palo Alto, where the conference was being held. I signed in at the registration desk & then proceeded to mingle with the other attendees, all of whom were conveniently wearing name tags. It was during this time that I had a chance to initially meet some of the individuals with whom I had several invigorating conversations over the course of the weekend. At around 9:30 I got a ride to the nearest VTA transit station with Peter Voss, who has been working on an Artificial General Intelligence project for some time now. I caught the 22, which took me to the Santa Clara transit centre, from which I was supposed to get home via the 60. Unfortunately, the last 60 for the night had already left. Stranded, I tried calling my roommate on my cellphone, whose battery was almost dead. Unfortunately, however, he was not in. Next I tried Maria in San Jose, who wasn't sure where the transit centre was but promised to find it & come rescue me, which she did. By the time I finally got home it was well past midnight & the conference was scheduled to resume at 9am the next day. I needed a better travel plan.
By the time I was done with breakfast on Saturday morning I had come to a very useful realization: the transit systems in Silicon Valley made up in bike-friendliness what they lacked in coverage. Inspired by this new perspective, I biked down to El Camino Real, caught the 22, loaded my bike onto the rack at the front, rode it all the way to Stanford & then biked across campus to the conference venue, getting there with minutes to spare.
As it turned out, the begining of the day was plagued by technical difficulties, arising from a last-minute attempt to get a complicated & relatively unproven holographic projection system running in a room filled with EM interference from a panapoly of comunication systems. The keynote by Ray Kurzweil began almost an hour late and we were never able to actually see him. He spoke about the increasingly quickening pace of technological development & what it heralds for the next couple of decades. This was followed by a debate between him & 2 physically present naysayers whose objections he was able to vanquish for the most part.
I spent lunch engaged in a lively discussion about the state of the global economy with a pair of fascinating women from LA whom I'd met the previous evening and an impressively knowledgable elderly gentleman who never actually introduced himself to me. After lunch I attended a talk about genetic programming by a Stanford CS prof. It seems that this approach to programming can trivialize conventional engineering problems, although it does nothing to help with problems that cannot be formally stated or whose potential solutions cannot be automatically verified. Nevertheless, designing new circuits, logic controllers, algorithms and the like can now be done via evolution of randomly generated code.
He was followed by Ben Goertzel, whose talk about AGI (Artificial General Intelligence) forced me to begin rethinking my plans for grad school. I had planned to conduct research into the creation of machine intelligence that would be able to learn how to understand & use human languages. However, Ben's talk made me realize that this would be impossible until we have an AGI that has acquired the basic human life experiences of a 2-3 year-old human child, after which picking up language would likely be trivial.
The next few hours consisted of 3 speculative speeches about the origins of the universe, followed by a debate between the 3 speakers about the various theories that are currently prevalent. Some of it was interesting but , as expected, there was no progress on the problem of trying to prove theories that seem to be inherently unprovable. Then we had a dramatic switch of gears as Sun Microsystems CTO Greg Papadopoulos spoke about how networked devices are continuing to change our lives. As expected, he got in a few good digs at the beast of Redmond. He was certainly the most entertaining speaker at the conference, peppering his talk with more than enough hilarity to ensure that everybody stayed awake & paid attention. He even managed to sneak in an allusion to My Big Fat Greek Wedding when he joked that the Greek words for large numbers were much better than the latin ones so beloved by the computer industry.
The dinner tables were tagged with questions to stimulate discussion so I went around looking for the most interesting one. That was when I met the legendary (well, at UW, anyway) Ka Ping Yee, who followed my lead & did the same thing. Apparently he is now pursuing a PhD in CS, exploring software for group discussions. I ended up having dinner with Ben Goertzel & a friend of his from a former AI company as well as a few other students also interested in AI. Oddly enough, we almost ignored our assigned topic & instead talked about all manners of other interesting things, including the nature of human language - one of my favourite topics of late - and Jef Raskin's discovery that aeroplane's do not really fly because of the Bernoulli effect. After dinner John Smart, the conference organizer, gave us a light talk about accelerating change, which was pretty insightful & I wish I hadn't been so sleepy during it. By the time he was done, around 11pm, I had decided that I was in no shape to face the long trip back home & began thinking of alternatives. Fortunately, I had brought along a change of clothes so I wouldn't have to wear the same outfit the next day.
I sauntered over to a nearby student house, where a bunch of Stanford students were chilling on the front porch and hung out with them for a bit, sipping a beer. After hearing about my plight, they offered to let me crash on a couch in their living room, so I did. They explained the Stanford student housing system to me. Apparently frosh live in dorms and then move out into these student houses for the remainder of their undergrad career. However, the student houses there are massive compared to the ones in Waterloo, often holding as many as 60 people. This was one of the smallest, with a capacity of only 30. As expected, the place was a mess, with random junk strewn all over the landing, piles of unused furniture stacked on the far side of the main hall and a dining table that wasn't hired for its looks.
The last day of the conference was pleasantly free from technical disruptions. It began with a moving talk about why treating everybody well is good for business & followed that up with a presentation by the founder of a startup pursuing a software replica of the human cortex before the lunch break. After lunch, the two interesting events were an introductory talk on nanotech by Christine Peterson of the Foresight Institute and one on social software (like blogs). I had to miss a talk by Tim O'Reilly in order to attend the nanotech one but I did have a chance to meet him briefly after the talk. He's a really friendly guy. Well, unless you happen to work for a company whose legal division is hell bent upon destroying any goodwill that people may have towards them, as a pair of SCO employees found out when they approached him and were met with some vocal vitriol.
I made it home in time to get in on my roommate's weekly trip to the grovery store.
0 Comments:
Post a Comment
<< Home