Beginning with a sad ending

After much prodding, pleading and pushing, the Push the Future blog is finally launched! Here you’ll find more of the behind-the-scenes news and ideas from the Push Institute, latest reports on the upcoming PUSH conference, and smatterings of observations, inspirations and musings from my life as a futures researcher.

I start, however, by pointing to an article in this month’s WIRED on the uncanny parallels between the lives and deaths of two promising researchers in AI (Artificial Intelligence), Chris McKinstry and Push Singh. Both had a dream of programming human-like intelligence in robots, that would finally make the idea of HAL, from the 1968 film 2001: A Space Odyssey, a reality. To achieve this, human intelligence must first be understood, and second be reverse-engineered.

Which is the greater challenge is hard to say, but the person who has probably spent the most time thinking about thinking is the much-revered AI researcher, Marvin Minsky, founder of MIT’s AI lab. Minsky was also a consultant on 2001: A Space Odyssey, so HAL was created from what he understood – and expected – the future of AI to be.It’s fascinating stuff, really. A combination of mathematics, psychology, computational linguistics and philosophy, AI gets at the heart of mind. And depending on where you weigh in on the matter, it can be argued that mind is at the heart of soul, and soul the heart of life. Or something like that.

Which is just a quick glimpse at how interwoven and theoretical this stuff is. But back to Chris and Push, because they were both dedicated to translating all that understanding into simple “thought” patterns of the mind, such that they could ultimately be programmed for the most elusive of all human traits, common sense.

Common sense, they reasoned, was a function of binary processes, a succession of yes/no switches that allow us to make “sense” of our surroundings and make decisions based on learned patterns. These patterns define perceptions of comfort, emotion, morality and more.

I’m no expert in AI, by any stretch, so I can’t be confident that how I’m representing it is 100% accurate, but here’s what I believe it boils down to: collect as many yes/no phrases, constructs, terms, rules as possible, and build a data base that a program can scan such that the final output is a behavior, reaction, emotion that is something akin to thinking.

The huge wall to scale in this challenge is amassing all that information. Both Chris and Push used the relatively new open source platform and invited anyone anywhere to add their own “binary propositions,” i.e. are rocks edible? Are babies brought by storks? Is Madonna an accountant? This was the best way to create the enormous data base required to simulate intelligence.

The difference between these two men, however, was that Chris McKinstry was doing this on his own, and seemed, from his online rantings and violent history, to be rather unstable, while Push Singh was conducting his work at MIT, with Minsky as his mentor and PhD advisor.

But the parallels are profound. Both were Canadians, both trying to solve the same problem in the same field, and both took their own lives, 4 weeks apart, in early 2006, in the same manner.

All this is a long paraphrase of the fascinating article that details the story in WIRED. And it’s all to serve as a background for the story that cuts much closer to home.

Push Singh was a speaker at PUSH 2005: The Geography of Change (June, 2005). He was featured along side Helen Greiner, co-founder of iRobot and Loretta Hidalgo, President, Space Generation Foundation. Respectively they represented significant advances in AI, robotics and space exploration.

Push had just earned his doctorate and had terrific enthusiasm for the future of AI and the promise that his OpenMind Common Sense project held for further advances in the field. He was very likable -warm and kind, sweet – and very impressive as well. His presentation was captured beautifully by Ethan Zuckerman, another presenter at the conference (a very smart man whose work focuses on the use of technology in international development, and probably the best blogger around). You can read/revisit it here.

Ethan called me just after Push died to give me the news. It was a terrific shock, I’m sure for everyone. He asked for permission to use a photo of Push from his presentation at PUSH (the shared name was a happy coincidence) for his memorial service at MIT (credit: Jim Nihart, Shortpantz Photography).

I didn’t know of the back story before I read the WIRED article. It is fascinating, if not a bit spooky. But mostly it’s just very, very sad. I’m so grateful to have met Push and for all of us who were there, to have known something of his brilliance and grace.