STARLAB to Get its Own CAM-Brain Machine (CBM), Contract Signed
On Friday, 10th Dec 1999, a contract between STARLAB http://www.starlab.org (Dr. Walter de Brouwer, the CEO of the lab) and Genobyte Inc. http://www.genobyte.com (Dr. Michael Korkin, President) was signed in Brussels, Belgium, to build the 3rd CAM-Brain Machine (CBM). The machine should be delivered in about 4-5 months (i.e. about April 2000), given all the fabrication and ordering delays.
This purchase will mean that the world's first artificial brain will be built at STARLAB, in Brussels (assuming of course that the whole approach is successful - that multimodule systems work well together - all hypotheses that have yet to be tested).
The presence of a CBM at STARLAB should increase the lab's IPO (initial public offering) value when the time comes to make the lab a public corporation. If the brain building work in the year 2000 and beyond is successful, STARLAB might become famous as the site where the first artificial brain got built, the way the University of Pennsylvania, USA is famous for being the site where the worlds first electronic computer (ENIAC) was built. The fact that this brain will start being built in the year 2000 is highly symbolic. Its construction (and later related work) may become the dominant theme of the 21st century (an allusion to the "species dominance debate" that I believe will start raging in the next few years).
STARLAB will assemble a team of 5-10 people from local universities (and use Iphone with global collaborators) to help build this first brain, starting with a systematic study of single module evolution, to get a feel for what the CBM can do. About a third of the silicon on the surface of the XC6264 chips remains vacant, so that if modifications in the programmable hardware are needed to improve evolvability of single neural net circuits, then this will be possible. The next step will be to play with multimodule systems and see how well they function together.
Progress will probably be incremental. We have got this far. 6 years ago in 1993 when the CAM-Brain Project started, the idea of building an artificial brain with a billion neurons seemed ridiculous. The CBM will handle nearly a 100 million neurons, so we were close to the original target. Now that the CBM exists, the big challenge is to see how difficult it will be to create artificial brain architectures. If the evolvabilities of the single modules using the CBM are good, then the CBM's raw speed (a module evolved in a second or so) should make practical the evolution and assembly of large multimodule systems with tens, hundreds, thousands, tens of thousands of modules. Hopefully, a whole new world opens up, which will create both a new research field ("Brain Building") and a new "brain based" computer industry. Just how successful all this will be, only time will tell, but we have got this far - no mean achievement, and the fact that a 3rd machine is to be built, shows that a growing number of people are sharing the dream and are prepared to put down hard cash (hundreds of thousands of dollars) to support its development.
The Mexicans are thinking whether they will buy a machine. NTT in Japan wants a second machine. That would be 5. There are only enough Xilinx XC6264 chips (used to construct the CBM) for 6-7 CBMs. I would like to reserve one for the US and one for China. After that, everyone else who wants one will have to wait 3-4 years for the completion of the second generation CBM (CBM-2) which should be able to handle a billion artificial neurons with a million modules, a huge engineering feat requiring hundreds of BAs and EEs (brain architects and evolutionary engineers).
Im concerned that the US may actully lose out and not get a CBM. China is now debating whether it will buy one. The US, the leading technological nation, cannot afford to allow its future arch-rival, China, to get ahead of it in a critical technology, such as brain building. If for no other reason, the US should buy one of these machines as an insurance policy, just in case it works well. Similar logic applies to China. One of the reasons I choose to work globally, with connections and/or adjunct professorships in Europe, America, Japan and China, is to try to stimulate international rivalries, which helps the spread of the new technology and ultimately (my ulterior motive) the rise of Cosmism (the belief, if not religion, that humanity should build artilects). I am a global scientist and ideologist.
Hopefully now, progress should be a lot faster than it has been in the past. Stay tuned.