I went with Derek yesterday to the Singularity Summit at Stanford. To say it was an interesting colloquium would be an understatement— it was full of interesting speakers, awkward moments, polite debates and impolite arguments. Though I didn't take great notes, I hope to tell you a little bit about the event here.
Ray Kurzweil was the keynote speaker, in that he spoke first and last. He was full of interesting ideas about "the singularity," a supposed event-horizon in the predictability of technological development. His argument is that the pace of many technological improvements has tended to move at an exponential rate, and that at some point, the pace of technology will surpass our ability to predict its emergence, let alone its social impact. 
Kurzweil showed many, many graphs of straight lines in log-space (i.e. exponential growth). Personally, I'm still a little skeptical of his cherry picking of significant technological breakthroughs; there will always be hindsight bias, plus the passage of time further biases our ability to pick out important events. What seemed to be a very important event 100 years ago may not even make it into today's history books, so there may just be a natural tendency to highlight more recent events as important and understate those that occurred long ago. Furthermore, what about the effect of population? Human population grows exponentially— one would naturally expect inventions to grow with the number of inventors. One would get exponential increase in the number of inventions as populations grow, only to continue at a linear (sub-exponential) rate when the population saturates. So although the idea of a technological singularity is exciting, I'm less than convinced that it is imminent.
I'm also a little confused by Kurzweil's log-log graph of paradigm shifts (above). He showed it, and nobody ever questioned it, but the math savvy should recognize that a straight line on a log-log graph is not indicative of exponential growth but instead illustrates a power law relationship. So does that chart contradict the exponential developments that are the cornerstone of his arguments? What does it mean to have a power-law relationship between time and technological developments? Of course nobody brought up these concerns, so these questions remain unanswered.
Next up was Douglas Hofstadter, professor of Cognitive Science and Computer Science at Indiana University. Hofstadter is also the author of Gödel, Escher, Bach. His talk was basically a criticism of the singulatarians' penchant for demagoguery and over-enthusiasm, and suggestion that those who want fair-minded academics to take "the singularity" seriously must renounce any association between "the singularity" and science fiction. It was gutsy of him to take the stand and say these things with Kurzweil on stage, and I agreed with most of what he said.
There were a few other notable speakers, Nick Bostrom (Director of the Future of Humanity Institute at Oxford University), Cory Doctorow (Co-editor of Boing Boing), Sebastian Thrun (Director of the Stanford AI Lab). All of them made some pretty good points, but the last few speakers (those at the various "acceleration studies" institutes) came across as kooks to me. There were also a few notables in the audience; apparently the chairman of the Electronic Frontier Foundation (EFF) was also in attendence.
The last speaker, Bill McKibben, was a skeptic. McKibben is the author of The End of Nature, the first book about global warming to target a general audience. McKibben suggested we might better spend our money on trying to make the Earth a better place, rather than distancing ourselves from our own humanity by further abstraction of the world through technology. He made some excellent points; judging from the level of applause from the audience, others must've thought so as well.
All in all, it was pretty neat to see these guys. I still have a lot of unaddressed concerns, but it is fun to think about how humanity appears to be getting accelerating returns on our investment in technology. I suspect we'll see many, many disruptive technologies, but a technological singularity? We'll have to wait and see.
UPDATE: I thought I'd give one positive example of an interesting idea somebody said at the meeting. I forget who it was, but one speaker suggested that because we are increasing the density of our connectedness, we ought to eventually have some phase transition. In statistical mechanics, we've long known that at some critical degree of connectivity, almost every system of interacting particles will go through a chaotic phase change into some other new phase. Often this new phase is associated with some unusual new property of the system (net magnetization, in the classic Ising model). So if global person-to-person connectivity reaches some critical density, what will happen? What will the new "phase" look like? We could soon find out if human networks behave like most other networks of interacting particles.
UPDATE (2006-06-08): After doing some math, I have decided to qualify my criticism of Kurzweil's chart. It turns out it is possible to make an exponential trend appear as a straight line on a log-log graph. The curious can read more here.
Last night, we went to a barbecue at some first-years' house up in the Berkeley Hills. The boyfriend of one of the first-year students is a product designer who works in Cupertino for a technology company called Apple. They make iPods. He designs what's coming next. So after talking with him for a bit, he tells me that part of his job involves making buttons(†). Not software— the physical construction of buttons onto the plastic/aluminum. He said he went to Best Buy and just played around with every imaginable consumer product to find out what the best "feel" is for buttons. Turns out, a user can be easily frustrated if things don't feel responsive... the whole interface needs to feel "just right."
