Silicon + Neurons Perfected, Massive Bio-Computers on the Way...

"Brain Cells Fused With Computer Chip"
"NACHIP uses special proteins found in the brain essentially to glue the neurons to the chip, but the proteins act as more than a simple adhesive, the professor explained. "They also provide the link between the ionic channels of the neurons and semiconductor material in a way that neural electrical signals could be passed to the silicon chip," he said."



I'm not surprised by this one bit, in fact they've been publicly doing work like this for several years now. However, these are important results because this proves that effective and lasting connections can be made between silicon and neurons. The ultimate applications of the technology are "potentially limitless", according to researchers involved with the NACHIP project funded by the European Commission's Future and Emerging Technologies initiative within the IST programme.

I have some science journal printouts that detail some other successes with neuron to silicon interfaces, and stacks of printouts about neuron stimulation using MEA's. Neuron stimulation has been going on for decades using MEA's, but they're far more limited in their ability to effectively manipulate multiple neurons, and more importantly being able to form and fire multiple synapses with each individual neuron.



MEA technology is rather effective and has brought some respectable achievements such as "brain in a dish flies flight simulator" (PDF), 'semi-living artist "MEART" and the "animat". They have been making profound MEA advances in recent years, in several ways important to this presentation.

First, they have built MEA's capable of firing multiple synapses, but the only MEA project of this sort involved an 8 electrode array designed for testing firing on one single neuron. This could be applied on larger scales, but not as effectively as using silicon. Second, they're now developing 3D MEA's, to study firing the live neuron networks from different sides. Third, they've been advancing control technologies. That is somewhat expected, but it demonstrates the advances in capabilities from custom upgrades beyond the standard equipment, and more importantly the advances that 'simple' university research teams can achieve imagine what multi-billion dollar government science teams can do.



I must add that (beyond this all being what is actually public) that those technologies and experiments are mostly just that: experimenting. The results of these large-scale experiments are important for 2 reasons here: One, understanding utilization of the entire neurons, instead of just interacting with the edges of networks of neurons. Two, understanding how to manipulate the neuron networks from multiple sides. Silicon chips, if properly interfaced with neurons, can control multiple synapses on individual neurons, but also in larger arrays with much of the required circuitry built right into the silicon architecture, which would be far more effective than separate circuits connected to the neuron interfacing hardware by wires.

Those who realize that these experiments are being conducted across multiple labs and universities may not consider these realities to be a monumental issue or threat. Likewise, those who read the new brain-chip articles, and read the part where the scientist said "It could still be decades before the technology is advanced enough to create living computers" might feel the same. I find that claim to be absurd (it's more like his team is decades away from), as (at least in the US) all national and university laboratories feed all of their science findings into what's titled the "NSF TeraGrid", which is a major WAN database that's feeds all of the science findings into one easy to mine source. It's like a science pyramid, it's existence motivated by the NBIC initiative.

At the top of that pyramid sits DARPA, who mines all of that data while also issuing major contracts with desired teams to complete critical projects to acquire findings, which are important to their cognitive computing programs. Those many programs are all just individual pieces in the same goal, the same beast. It should be noted that this particular program was part of the EU, who has their own NBIC initiative that clearly attempts to follow that of the US (as is probably true of the China/Russia/India superpower).



It fascinates me how conventional AI proponents still desperately hang onto the notion that conventional AI theories will rule first. Conventional AI has been around for decades, biological I has been around longer than anyone can really say for sure. These people think that they can build math models of live neurons and operate virtual networks of neurons to effectively out-do live neurons. This will probably be possible in the future, but we have a long way to go before we truly know exactly what "math" a neuron is capable of. On the other hand, we can already control live neurons, and if properly done we could control large arrays of them in advanced configurations.



Originally, I figured that they would have large 'brains' in massive MEA arrays, probably in cube or sphere like shapes. I considered the sphere shape based on the shape of the infamous Echelon domes (that people still speculate about what's inside them to this day), but the cube or rectangle shape seemed more feasible. Several months ago, after I learned about the effectiveness of then silicon-neuron interfacing technologies, I determined that the ultimate super 'brain' would be a large cube shaped device.

The inner walls of said cube would be these silicon chips, a 3D neuron chamber with interfacing silicons on all sides. I reached this conclusion by assessing two important schools of thought. First, the science and manufacturing school (science); silicon chips are typically formed in square shapes in large molds that produce sheets of semiconductor chips. Each 'sheet' typically furnishes a large set of identical chips (but often in surprisingly circular shaped molds). Second, the occult/esoteric school of thought (psi-ence); the square, and more importantly the cube symbolize perfection (most notably according to Freemasons), as it also does in mathematics.





[edit on 24-4-2006 by IgnoranceIsntBlisss]

Interesting stuff.

You know, before now, I thought that we would eventually end up with AI through extremely complex network links, and a few glitches... but now, well... it looks as though we may be applying Natrual intelligence to artificial pathways.

I knew it was possible... but I didnt think we would do it this soon.

Its kinda creepy though... enough to put hair on a mans chest.




See?

I just had a thought...

What if Bio-Computers are the step we need to achieve Artificial Intelligence? A kind of short-cut around programming difficulties.

Why do I mention this? Because of the unique ability for neurons to form their own new neural connections. That's how a person learns and grows - by forming new neural connections.

So, if you could construct a micro-brain of neurons and silicon, and give the silicon pre-made instructions (ie, programs) - could the neurons create new connections and be the heart for a computer that learns?

Now, obviously it'd be more complicated than this, though it would make for an interesting movie concept "how the machine came alive", and you'd have to have programs in place that could interact with slightly mroe random signals from the neurons - but the heart of the concept is there. Do you agree?

Exactly, that's what I've been trying to tell everybody since last Sept, but it's been an uphill battle to get people listen. Check out 'my' "Beyond AI" thread.

WICKED! I think I had dreams of stuff like this as a kid. Pyrimid and cube shaped multifaceted chips. Not sure about the braincells part, maybe in nightmares after watching faces of death. People had better be careful with that sort of stuff if you know what I mean. That could really spawn a "Terminator" scenario. We no need any mo governators running around! LOL. It could turn out like in hitchhikers guide where the happy gay ship computer hates the pessimistic sad robot! ROFL. This could spawn smarter robotics and androids "in the near future"

~MASTERODUNIVERSE