have only just become available to all, certainly in the last few years.
I've been dreaming of robots for more than 30 years, but only really in the last year has the hardware become available, or come down in price to put it into the range of a casual hobbyist. I'm a bit more serious than that, especially now I can get my hands on tiny camera boards, and tiny servos that can lift kilos, and more importantly access to things like Arduino, RasPi and their communities.
The hardware was twice the size a few years ago, lifted only a kilo or so and needed a thumping mains PSU, but this year I've built a biologically-inspired robot containing 6.4GHz mobile processing, 20 servos, stereo cameras and the batteries to run it all, and it stands about 1 foot 6 inches on all fours. I've spent less than 100 on its hardware too, the CPU board was scrap and the servos were 2 each from a Chinese eBayer, the rest is aluminium sheet (25, the most expensive), various plastic bits and nuts, bolts and rivets. Webcams cost under 10 each, and microcontrollers half that.
I'm now working on making it smart. It was hard enough making it see - judge distance and recognise objects - but getting context out of it is harder than I thought.
There's things like OpenCV to handle vision as an API, Kinect, neural modelling software and all sorts on the web, but nothing to do something as general as thinking.
I'd buy Kurzweil's book, but I'd like to bet theres not a lot of code in there among the philosophy - and you need a liberal helping of both to do what the title suggests...
Speaking of philosophy, one thing that Kurzweil mentions is that we shouldnt fear the future because we're already in a human-machine society. I saw an advert earlier on tv that highlighted just that. It was for a camera, billed as containing 'David Bailey' - or at least as much expertise. It wasnt so long ago that only expert humans could take consistently good photographs, and a machine that could was considered nothing less than AI. And yet this compact camera has absorbed everything that poor Mr Bailey was, and now all he can do is advertise it. Its just a camera, and most of us have a phone with more photographic knowledge built into it than average joe knows. I'd have been shocked 20 years ago to have bee told that we all carry video recorders as a matter of course today.
AI becomes invisible pretty quickly when its applied, and isnt scary for long. Unless its your job its after of course...
MARANA, Ariz. — Inventor, author and futurist Ray Kurzweil took to the stage here at the Techonomy conference in Tucson to offer his thoughts on a future where humanity is enhanced by technology.
Kurzweil spoke to Techonomy founder David Kirkpatrick about his new book on human thought, How to Create a Mind, and the various themes that stem from it. Their talk was varied and at times scattered — with a topic this big, you can imagine the temptation of tangents — but Kurzweil had a few choice things to say along the way.
The highlights:
- I’m very optimistic [about next five years] because there’s a lot of evidence that not only hardware is progressing exponentially but software [too].”
- IBM’s Watson and its performance on the television game show Jeopardy! is “viscerally impressive” in that people don’t understand how truly remarkable an advancement it is. (Why? Because it’s a computer coming to conclusions on its own, rather than searching a database and reiterating data stored within,
- We are so integrated with connected technology today that “During that one-day SOPA strike, I felt like part of my brain had gone on strike.” (SOPA legislation in threatened the autonomy of content providers on the Internet.)
- Holding up a Google Android-based smartphone, Kurzweil said that “these will be the size of blood cells by 2030.”
- Indeed, IBM’s Watson and Google’s autonomous car will become deeply integrated in how we live. “That kind of system will become a reliable tool that people will become dependent on,” he said.
- The very human capabilities of being funny or sexy? “These are not sideshows to human intelligence,” Kurzweil said. “That’s the cutting edge of human intelligence.”
- Artificially intelligent agents can be considered human “once they write a novel,” Kurzweil said. “They will be convincing in their ability to do human-like thing.”
- There’s no reason to fear the future. “We are a human-machine civilization [today]. Computers are doing things all the time that we can’t possibly do.”
- And things are progressing quickly. “A kid in Africa has more technology at his disposal than the president of the United States did 15 years ago.”
- Technology will push us to be more human, not less. “We’re going to use those tools to make ourselves more expressive and more intelligent.”
- And it won’t be conceptually different from the analog era, either. ”We do have ways to make ourselves smarter through collaboration,” Kurzweil said. “That was the value of language.”
- “I’ve been thinking about thinking for 50 years,” Kurzweil said, reflecting.
- Discussing his book, he said that only recently, “We can see our brain create our thoughts. We can see our thoughts create our brain.” His book’s publication was delayed four times because of new research advancing this idea.
- Understanding the brain better is important for three reasons: first, to fix it better; second, to provide models for humans to create more intelligent machines; and third, to further the science of understanding ourselves.
Finally, Kurzweil was asked by a member of the audience: “Are you an optimist?”
He chuckled, then replied: ”I’ve been accused of being an optimist.” Then he leaned back in his chair.
Photo: Asa Mathat
