The more things change, the more they stay the same, with Ray Kurzweil’s talks. At the TEDx Conference in Manhattan on Monday, December 3, he and the biotech pioneer Juan Enriquez were the keynote speakers.
They were the only speakers given a lot of time. The rest of the conference was “digital tech lite,” focusing on very basic concepts sprayed at the audience at high velocity. Each main speaker had only 18 minutes—10 for the small-fry, the local entrepreneurs—so it was kind of like consuming fast food: you bolted down the thoughts that were flung at you rapid-fire, often in disorganized fashion, in the hope that you could digest them later without too much heartburn.
Kurzweil and Enriquez each gave an 18-minute talk, then they sat down and fielded questions by a moderator, with some tweeted questions from the audience thrown in. Kurzweil spent the first half of his time talking about his well-worn theory of the exponential (rather than linear) progression of technology, which will in his estimation provide us with sentient AI by 2029. Then he launched into a short explication of the main idea of his new book, How to Create a Mind. If you’ve read the book jacket, you know what he said: the brain can be seen as a series of hierarchical pattern recognition algorithms. Using this theory of mind, he claims, would allow us to more easily reverse engineer the brain and make humanlike AI. In fact, he claims, our success to date has been due to an innate recognition of this fact. As he says in his book, “there must be an essential mathematical equivalence to a high degree of precision between the actual biology and our attempt to emulate it; otherwise these [A.I.] systems would not work as well as they do.”
An interesting theory, and very seductive. But he left some questions poorly answered or unanswered. First, as Gary Marcus, a neuropsychology professor from NYU, says in his review of the book in The New Yorker Magazine, “What Kurzweil doesn’t seem to realize is that a whole slew of machines have been programmed to be hierarchical-pattern recognizers, and none of them works all that well, save for very narrow domains like postal computers that recognize digits in handwritten zip codes.” And second, he really doesn’t address in his talks (I’ve now seen two of them) how we are going to power all of our increasingly complex and advanced devices in a world constantly on the brink of an energy drought. (He did address this briefly in the question and answer period by saying he is confident that solar energy will progress geometrically as digital progress has, but didn’t explain it any further.)
Enriquez talked about big data, riffing on the burgeoning immortality that is caused by wide availability of information on the web about each one of us. He especially pointed to the possibilities and potential of combining this information with facial recognition technology. You could know about anyone you meet in a bar before you actually talked with them.
The question and answer period was a bit slow and didn’t allow the audience an active role (beyond tweeting, which left out those of us without accounts or who, like me, suffered from lousy reception in the cavernous vault of the venue, which for some reason didn’t have open wireless access). Also, Kurzweil and Enriquez appeared at the end of a long day, and many people got up and left during the lengthy and slow Q&A.
There were about 400-500 people at any given time inside “Terminal 5,” a large concert venue rented for the daytime event. Organized by Chris Grayson, a Digital Innovation Strategist who previously worked for H+ magazine and Mondo 2000, the event featured the two big names I’ve just discussed: the biotech pioneer Juan Enriquez and techno-futurist Ray Kurzweil. Other speakers of note were Bre Pettis, Founder of Makerbot, a maker of small 3D printers for consumers; Jincey Lumpkin, sex columnist for the Huffington Post; and Francesca Ferrando, an Italian novelist and academic specializing in the philosophy of the posthuman. Also giving short presentations onstage throughout the day were representatives from various local tech startups.
I was at the conference almost the whole day (it went from 9:30a.m.-9:30 p.m.), and it was of varied value. The presentations by the principals of the various local startups were really just 10-minute marketing blurbs, and didn’t present anything that I considered really new or astounding. One that was impressive was Sameer Parekh’s presentation of his company’s autonomous flying bots. These units, made by Falkor Systems, which are essentially four fans arranged in a square, with a digital processor in the middle, might be familiar to those who’ve read about the biggest project to emerge from last summer’s consortium at Singularity University. This project, which has been in the news recently (see The Economist, Dec. 1 issue: “An Internet of Airborne Things”), is planning to use a vast network of machines like Parekh’s as a sort of robotic, airborne pony express. The idea would be to deliver smallish objects—like medicines—to hard-to-reach locales via “a drone-powered internet.” Parekh presented his machine in an impressive manner having it fly onto the stage and follow an assistant around using simply its own processing power. Jim Kovach, COO of CrowdOptic, gave an interesting, 10-minute talk about using "clustering" of personal tech, like smart phones, to present crowd-sourced news.
Lumpkin discussed theoretical legal issues connected with the prospective development of sex robots (eg., what legal rights would they have, if sentient? Should they be programmed to have the right to say “no”?); Bre Pettis, owner of Makerbot, promoted use of his small, microwave-sized 3D printers as a "factory at home.” His aim: to make it easier for entrepreneurs to launch companies with rapid and cheaper prototyping. Another idea of his is to put them in schools, like Lego sets. And he thought that the common consumer could make replacement parts for anything, at home, on the fly. Examples: some hard to get part for an espresso machine, or a part for an iPod. What he glossed over was how expensive the CAD graphics programs and material substrate for these printers can be…and how difficult to use. Ferrando’s talk on the posthuman was a quick summary of how the increasing incidence of implanted, digital devices are making moving us toward a cyborgian existence, and how moving toward a posthuman state might actually create a more equitable society by forcing us to engage with even more diverse forms of humanity than we have now. I found myself wondering, afterwards, how this might happen: we seem to have a hard enough time dealing with diversity now; how would more diversity lead to greater acceptance of it? Wouldn’t it just provide haters with more things to hate?
To be fair, I could have asked her about this quibble afterwards, since one of the good things was that the speakers often hung around after their talks, and I had a chance to speak with her. But I didn’t think of it until I had time for tranquil reflection.