[Continuing coverage of the 2009 Singularity Summit in New York City.]
The last talk of the morning is from Brad Templeton of the Electronic Frontier Foundation on “The Finger of AI: Automated Electrical Vehicles and Oil Independence.” (Abstract and bio.)
He starts off bemoaning the horrors of human driving, mostly for reasons of safety. Then he complains about how on the way to the conference today, under the sidewalk, he heard workers installing honest-to-goodness nineteenth-century transportation technology! Subways! This is what governments today are spending their money on! The future, he says, is in robot-driven or autonomous cars. (Yeah, replacing mass transit with cars for everyone will work great here in Manhattan.)
He’s going over the DARPA autonomous vehicle project now — showing an extended clip of a documentary video. Really? I think everyone here already gets the idea. (I have very conflicted feelings about autonomous cars, incidentally. I have a few friends who participated in one of the DARPA contests, and the technical challenge is amazing and must have been really fun to work on. But man, do I love driving a car myself, and that’s something now pretty thoroughly integrated into the modern American psyche. It may be interesting to watch how the A.I. and environmental movements converge or diverge on this topic; indeed, the tension is already evident from Templeton’s previous comment.)
Now Templeton is touting autonomous vehicles as a political necessity because it will eliminate our dependence on foreign oil. If robots are driving cars, we don’t need to own them ourselves anymore, and they can become specialized and lightweight. We press a button on our cell phones and near-instantly, a robot taxi pulls up. He says this will bring about drastic increases in efficiency, so that it’s even more efficient than mass transit.
Templeton is talking about potential problems with this vision. He notes, “People are very scared, for some reason, of being killed by robots.” That may be the single best quote of the conference so far. But he also noted earlier that robots will be better drivers because they don’t get drunk. Oh really?
A questioner asks what will happen to the car insurance industry when robots are driving and crashes disappear. Templeton says, “Yeah, f*** ’em.”
Another questioner asks how a society increasingly dependent on robotics will be affected in terms of its perception and concentration capabilities. She seemed serious and inquisitive, but Templeton sort of starts badgering her, thinking she’s disguising a critique as a question (she says she was just asking). His answer is just that he’s surprised to hear that question in this town, and that he thinks of these innovations as advancements, not “bugs.” At least one Twitterer agrees that he should have handled it differently. [UPDATE: A picture of the questioner asking Templeton.]
I’m skeptical of Templeton’s plan, but I must admit that his was one of the more entertaining and engaging talks at the conference. He speaks in a rather rapid-fire way, but everyone’s completely following along with him on everything. It’s like he’s trying to sell us a car…
And that’s a wrap on the morning talks.
I was hoping for a better presentation by Brad. It is not just Robocars. Congestion pricing, for example, requires a digital wireless device in the car for $100. But once a car has digital two way, then suddenly, software in Manhattan, for exampel, can organize cars, buses and commutes with software, resulting in up to 60% less oil use.
Auto, bus and truck automation of various levels is coming next year.
It'll be interesting to watch that develop, if it does, and watch the self-reliance of automobiles that's so written into the American psyche disappear.
I do find fascinating, though, the work done by some computer scientists at my own alma mater, The University of Texas, in which cars that are controlled by a centralized computer system (or at least, all agree to cooperate with its directives) can eliminate the need for stoplights. They all go through the intersection at the same time, and it's just perfectly timed so they don't hit each other. It's amazing to watch these simulations. If this innovation ever came to pass, it would dramatically improve efficiency and speed, and without the need to change the existing road infrastructure.