Posts Tagged ‘kurzweil’

Robot Language and Data Transmission Speeds

One demarcating line between sentient creatures and non-sentient creatures is the ability to use language, or otherwise communicate, with other members of the same species. Dolphins, ants, bats, and dogs (among others) all seem to be able to communicate within their own species, and they obviously don’t any human language to communicate. Thus far, only Koko the gorilla has been able to communicate directly with humans using a language that both understand (with apologies to owners of expressive pets.)

Now, researchers at the University of Queensland have created robots that are, themselves, independently creating a unique language. Finding human words too complicated for the efficient transfer of information, the robots now generate a random sequence of sounds to name a new place, then share that sequence with other robots to allow those robots to find the same place. Because each robot travels independently, they each make up their own random name for a place, and when two robots have both come to the same place but given it a different name, they seem to essentially negotiate with each other until they agree on a common name. Robot A, having traveled to some place, can tell Robot B (who has not been to that place) about that place: Robot B can then tell Robot C about the place that Robot A went to, even though Robot B has not been there. The ability of a robot to talk about places that that robot has not been to (indeed, they seem to be able to speculate about places that no robot has been to) seems like an important step toward more robust consciousness.

Considering that we don’t have a perfect explanation of how language maps onto representations of the world for humans, recreating a similar experience within robots also allows for realistic hope that we could create a sentient robot without fully understanding how our own sentience works. There is danger in that as well, because we don’t have a proven method of recognizing consciousness researchers might inadvertently destroy a sentient creature (or a great many) without realizing what they’ve done. Hopefully we can create such a reliable test quickly, or barring that, that sentient robots are more forgiving than people have a tendency to be.

If robots can create and share their unique language, they will soon be able to transfer that information across vast distances much more quickly. Scientists in the UK have broken a single laser information transmission record by transferring data at an astounding 26 terabytes per second. That’s roughly 700 DVDs, or 400 million phone calls, every second. Japanese researchers have used a multi-laser setup to transfer data at an even more amazing 109 terabytes per second. Confirming some of what Kurzweil has been saying, data transmission rates, according to researchers, are nearly doubling every 18 months. That means in about six years we should be able to transfer nearly 400 terabytes per second on a single laser using mathematical algorithms to encode and decode data. My fuzzy math says that’s about the data encoded in 10,000 DVDs transferred via single laser every second – at that rate much more complicated things like virtual reality ought to be transferable quickly enough to create decent representations of the ‘real world’ without too much lag (or stuttering, as the computer attempts to display the graphical information.)

Both discoveries are impressive in their own right, but the combination of technologies means that robots can potentially begin to create their own language and then transfer that information to other robots at incredible speeds. Moreover, there is little reason why language needs to be tied exclusively to location – robots should be able to use their language to describe anything in the world, and then transfer that information to other robots who can speak about the thing describes without ever having directly experienced it – not at all unlike what people can do.

Better, as computer components miniaturize and become integrated with people, we too ought to be able to take advantage of the data transmission rates. Of course, a biological brain could never process that much information that quickly, but with some mechanical augmentation we might well be able to translate the optical signals into chemical reactions that the brain can understand.

As a formatting note, I’m going to move to a once-a-week publishing schedule. Although today is Saturday, I expect to publish on “Futurist Fridays” from now on.


Biotic Hands and Programmable Brains

May 20, 2011 2 comments

Every so often I come across an article that really illustrates how near the future is. This week, I came across two of them.

The first article, by Singularity Hub, is about Milo and Patrick, two men who have chosen to have their hands removed and replaced with biotic hands. Both went through extensive surgery (Milo for around 10 years) but eventually they decided that because the surgeries were ineffective, replacing their hands with biotic hands was a sensible alternative. These stories are important for at least three reasons.

First, both surgeries were elective procedures in the sense that neither man had his hand replaced to save his life, or as part of a traumatic incident. Both men had biological hands, although they were damaged beyond reasonable use. Elective replacement for limbs is on tricky ethical ground because, for many people, replacing limbs is largely a procedure of last resort. Previously, limbs were removed to prevent the spread of gangrene, or to save the person from further infection spreading, or for reasons otherwise necessary to protect the person’s life. Here, for at least two men, given one hand of lesser functionality than normal each, replacing a less functional human hand with a biotic hand more functional than their damaged hand (but, seemingly, less functional than a ‘normal’ hand) made sense.

Second, if two men are able to choose to replace a biological hand with a more functional biotic hand, then others should be allowed to make the same decision. Despite the amazing progress made by Otto Bock (creator of the biotic hand), the hand still doesn’t provide all of the benefits of a normal human hand, and offers only one benefit that a human hand does not: 360 degree range of motion at the wrist. However, the limitations on the hand are technological, and with a sensory feedback system like the one Bock is currently working on those limitations ought to be cured quickly. Once a biotic hand is as functional as a biological hand, scientists ought to be able to craft improvements for the biotic hand that include all sorts of functional improvements (increased grip strength, additional sensory inputs like the ability to sense electric currents, more range of motion, enhanced durability) and some more cosmetic improvements (perhaps a small storage space, or wifi, an OLED screen, or other patient-specific enhancements.) Very quickly, a biotic hand will be superior to a normal hand, and not just a severely damaged hand,

Finally, one cannot escape the naked truth that this is what people have had in mind when they used the word “cyborg” for decades. Although it’s true that eye glasses, pacemakers, and seizure-reducing brain implants are all mechanical augmentations to a biological person such that the term cyborg is properly applied, few people tend to think of their uncle with the pacemaker as a cyborg. In part, that’s true because pacemakers are not visible, and even hearing aids are more like eye glasses than biotic hands because they are removable and not an integral part of the human body. These hands, however, are replacing a major part of the body with a clearly mechanical device. The article is unclear whether these hands come with some sort of synthetic skin that masks their metal servos and pivot points, but from the pictures there is just no mistaking that these men are now part robot.

We have reached a point where we can program mechanical devices so that they can communicate with the brain through the nervous system. But what about programing the brain itself?

Ed Boyden thinks he has created a solution for that too. I highly recommend watching the video yourself. The gist is that neurons in our brains communicate via electrical signals. By using engineered viruses to implant DNA encoded with photo-receptor cells taken from parts of algae into brain cells, Boyden can then shine a light onto parts of the brain and only those cells so implanted with photo-receptors activate for as long as the light shines. By activating particular groups of neurons, Boyden can stop seizures, or overcome fear responses that would otherwise cripple an animal. Using fiber optics and genetic encoding, Boyden has found a way to direct the brain to act just as he wants: He has, in essence, figured out how to program a brain, or at least how to hack the brain to add or remove particular functionality.

Further, when photo-receptor cells are implanted and neural activity can be activated via light, the human brain begins to look even more like a computer. By regulating light inputs the cells implanted with photo-receptors activate and produce particular effects depending on the type of cell that they are. With a basic on-off activation scheme, neurons become a lot like the chips in our computer that we activate with electricity turned either on or off. This on-off sequence is represented by 0’s and 1’s in binary code and upscaled to more complicated programming languages. With an implanted light array, programmers ought to be able to create flashing light sequences that affect the brain in preset ways, essentially writing a code that controls parts of the brain. Even if scientists simply read the signals of the neurons, all of human experience ought to be reducible to groups of neurons firing or remaining dormant in complicated patterns. If that is so, then there is no reason why we couldn’t download a stream of our experiences in complete detail, and perhaps eventually upload them as well.

The viral DNA distribution method has also been used to restore sight to mice with particular forms of blindness, apparently at the same level of functionality as mice who had normal sight their entire lives. This distribution system ought to be able to introduce whatever bits of DNA seem useful, essentially taking parts of the DNA from other animals and implanting them into human cells to augment our own biology. The color-shifting ability of a chameleon, the ultra-sensitive scent glands of a snake, or the incredible eyesight of a hawk are certainly products of their DNA, and conceptually ought to be transferable with the right encoding. Boyden is quick to point out that the technology is just getting started, but given the exponential increase in technological progress I suspect that we will see vast progress in their fields, and perhaps even human testing, in the next five to ten years.

Despite the exciting prospects of viral DNA introduction, I can’t help but flash back to the beginning of movies like Resident Evil and I Am Legend. Even for technophiles like myself, some of this technology is a little unnerving. That’s all the more reason to start taking a hard look at what seems like science fiction now and figure out what ethical lines we are prepared to draw, and what the legal consequences for stepping outside of those lines ought to be. Much of this technology, if used correctly, is a powerful enabler of humanity for overcoming the frailties of our haphazard evolutionary path. The very same technology used incorrectly, however, could have dramatic and catastrophic consequences for individual patients and for humanity as a whole.

These two stories indicate the dual tracks of transhumanism: The mechanical augmentation side replaces biological hands with mechanically superior components while the biological enhancement side introduces bits of foreign DNA into our own cells to provide additional functionality. If the rate of progress continues, both of these tracks ought to be commonplace within the next 20 years or so. At the point where we can reprogram the human brain and replace limbs with mechanically superior prosthetics, Kurzweil’s Singularity will be here.

I, for one, am very excited.

Ray Kurzweil visits the Colbert Nation

April 13, 2011 3 comments

Last night, Ray Kurzweil appeared on the Colbert Report to discuss his new video, Transcendent Man.  Because Transcendent Man discusses transhumanism generally, the discussion opened with a film clip and then allowed Kurzweil to discuss transhumanism generally.

I’ve seen the entire Transcendent Man video, and the snippet shown on the Colbert Report is a fair summary. The film itself has (to my mind) a strange turn toward the religious and slightly creepy; I’m not sure if the film really gets across a message about the virtues of transhumanism (though it is discussed throughly) and instead puts the focus on Kurzweil and makes him seem like a creepy prophet. Other interviews suggest that Kurzweil is not nearly so obsessed as Transcendent Man makes him seem (with reviving his father, anyway.)

The Colbert interview, on the other hand, struck more of a techno-idealist tone with the usual Colbert humorous undertone.  As Singularity Hub points out, this interview on the Colbert Report will probably reach a younger (and larger) audience than Kurzweil’s longer interviews on shows like Charlie Rose. Because of the Colbert format, Kurzweil wasn’t able to get too in depth with his research and arguments for The Singularity, but Kurzweil was still able to get the high points across and build some excitement. Given the much vaunted ‘Colbert Bump’ I’d expect to see articles in the next few days saying that downloads of Transcendent Man skyrocketed.

While I’m all in favor of transhumanism gaining more mainstream support, I worry that people who are excited by the Colbert interview will get introduced to transhumanism via Transcendent Man, write off Kurzweil and his associated ideas because of the tone the film takes, and associate believers with the fringe beliefs (at least without a little information as prep) that seem to take center stage in the film. I wish there were a better film to introduce the mainstream to transhumanism; or that the film had mentioned other sources of information like H+.