Machine people (2)

In a remarkable experiment, researchers in North Carolina have installed brain-interfacing implants in the brains of rats and demonstrated that the animals can use the devices to share information directly from mind to mind (and yes, rats have minds).  In one case, one rat was in the North Carolina lab while its partner was in a lab in Brazil, and their brains were connected over the internet.  The information-sharing still succeeded.

The work has obvious implications for combining the power of multiple brains to solve problems that individual brains alone could not solve -- the technology would probably work even better with humans than with rats, since humans would understand the process and be actively trying to make the best possible use of it.  But given the onerousness of receiving brain implants, I suspect the human applications will be limited, unless it turns out that this artificial telepathy has some spectacular advantage over the system we already use to share information between brains (language).

The most important revelation from this work is that it demonstrates, to a new level, the ability of the organic brain to interface in functionally useful ways with machines, even at our current low-resolution understanding of the brain.  It's clear enough by now that the brain is essentially a data-processing system -- that the operations it performs which we experience as sensory awareness, thought, consciousness, and so on, are fundamentally the same kind of thing that a computer does when it runs complex programs.  The brain is organized very differently than a man-made computer, is made of very different materials, and has different strengths and limitations, largely because it evolved through natural selection rather than being designed -- but we can still say it is the same kind of system as a computer, to the extent that we can make practical use of the resemblance.

Making the brain interface usefully with man-made systems, in the same way that man-made computers can interface with and work with each other or with other devices, is the most important form that this "practical use" takes.  It demonstrates the feasibility of ever-stronger functional augmentation of the brain with machines, eventually leading to the full integration of computer capabilities into human intelligence -- the key technology of the Singularity.

The Nicolelis Lab in North Carolina which carried out the rat experiments has already done pioneering work along these lines, allowing paralyzed people to control computers and artificial limbs directly with their brains -- an artificial replacement for natural capabilities they have lost.  The logical long-term goal of such technology is what I discussed in this post -- complete artificial bodies to enable people whose organic bodies can no longer support them, due to age or disease, to continue living as normally as possible.  The Russia-based 2045 Strategic Social Initiative has set itself the goal of developing this very thing over the next dozen years.

(In the long run, of course, we won't truly have escaped the curse of biological death until the fragile organic brain itself can be replaced with more durable systems which can replicate all its capabilities well enough to continue "running" an individual mind exactly as the organic brain did -- but that's a whole further leap in technological sophistication which will likely take a further couple of decades.)

As I pointed out in the earlier post, developing artificial bodies which can replicate the capabilities of the organic human body is a very tough problem.  One of the most difficult aspects of it is walking, since the upright bipedal human gait requires a sophisticated ability to balance.  Robotics is making impressive progress in that area, however:



(HRP-4C can walk rather better than I can at the moment, due to a chronic hip-tendon problem.)  Here is another robot, not designed to imitate the human body specifically, but impressive in its ability to keep moving on difficult ground (link supplied by reader Bacopa):



Notice how strikingly animal-like the robot's self-stabilizing reactions are when pushed or slipping, illustrating how well machines are already able to replicate the capabilities of organisms.

As I said in the earlier post, there are other areas that need work, such as replication of the full range of human senses including touch, taste, and smell.  But by the time we're able to start transplanting the brains of elderly people into new bodies to give them new lives, the bodies will be ready.