Bottom Up Brain Building: The Missing Links

There is a way to build thinking machines that might work. Use as your model the only proof of concept of higher intelligence known to the universe: the human brain / mind. Start by developing a machine nano-architecture that can do almost as many things that the human brain can do. To do this, you must understand the brain and you must invent new machine tools for imitating what the brain does.

Berkeley professor of electrical engineering Leon Chua, has made a good start. His "memristor" devices have gotten a lot of people excited about the possiblity of designing a working machine brain from the bottom up.

When two metallic wires are separated with a few nanometers of memristive material (such as certain transition metal oxides), an electronic device is formed that acts much like a nonlinear resistor, but with a twist. The resistance varies over time as a function of the currents flowing through it. In other words, it is a resistor with memory.
The rate at which their resistance changes is extremely nonlinear in the voltage applied. Small voltages hardly perturb the resistance at all, while somewhat larger voltages can induce fast changes. _Memristor Cortical Computing

Because memristor devices are able to change their electronic characteristics based upon their electronic history, they can be seen as analogous to neuronal synapses, which change their characteristics based upon their synaptic firing history. This dynamic adaptation on the nano scale presents the possibility of building very dense self-organising electronic networking chips far beyond anything previously possible.

Intelligence is not an algorithm. It cannot be "programmed." But given the proper "neural architecture" and appropriate experience, intelligence can be evolved over time. The memristor needs to be joined by other advanced memory-electronics in order to provide for machines the extremely subtle learning that the human brain can do.

The electronic brain will be a time coming. "We're still getting to grips with this chip," says Williams. Part of the problem is that the chip is just too intelligent - rather than a standard digital pulse it produces an analogue output that flummoxes the standard software used to test chips. So Williams and his colleagues have had to develop their own test software. "All that takes time," he says.

Chua, meanwhile, is not resting on his laurels. He has been busy extending his theory of fundamental circuit elements, asking what happens if you combine the properties of memristors with those of capacitors and inductors to produce compound devices called memcapacitors and meminductors, and then what happens if you combine those devices, and so on.

"Memcapacitors may be even more useful than memristors," says Chua, "because they don't have any resistance." In theory at least, a memcapacitor could store data without dissipating any energy at all. Mighty handy - whatever you want to do with them. Williams agrees. In fact, his team is already on the case, producing a first prototype memcapacitor earlier this year, a result that he aims to publish soon. "We haven't characterised it yet," he says. With so many fundamental breakthroughs to work on, he says, it's hard to decide what to do next. Maybe a memristor could help. _NS

It should have been obvious that the tools for creating machine intelligence were inadequate to the task. Just as it should have been obvious that the models used to predict climate over multi-decadal scales are completely inadequate to the task. Many things become obvious once one throws off the blinders of conventional groupthink. If one allows oneself to think outside the box, the possibilities suddenly multiply wildly.

More: The image at top is of a slime mold, which is sometimes a single-cell organism and sometimes a multiple cell organism, depending on environmental conditions. Each biological cell is in essence an incredibly complex computing machine. Multicellular organisms possess unbelievable biological computing power, compared to non-biological computers. Imagine the computing power of a human being, complete with brain and nervous systems. It is literally beyond the power of human made hardware to emulate. But as humans develop more subtle technologies of computation, they will be able to build -- from the bottom up -- machines that have the intelligence first of insects, then of higher and higher animals.

Will humans eventually be able to evolve machines with the intelligence of humans and higher? Of course. It is also true that if humans do not choose to become more intelligent themselves, that they are finished on this planet, long term. How humans go about the enhancement of their own intelligence will make for some interesting work in the not so distant future.

Brian Wang provides more quotes and images on this topic