AI View of AI, Scaling Wars

The artilects, as they have been conceived so far in this book, have been largely "nanoteched" creatures. But nanotechnology may be unnecessarily restrictive and far too large a scale to be suitable for advanced artilects. It may be possible that a "femtoteched" creature could be built. Such "femto-artilects" or "femtolects" as they will be called from now on, would be vastly superior to "nano-artilects" or "nanolects," thus setting the stage for a new "species dominance war" all over again.

Perhaps I should call these wars "scaling wars." The first such scaling war is indirectly the topic of this book, concerned with the potential for a war between the nano and the meter scales, i.e. between nano based life forms (the artilects) and humans. A second such war, relating to the femto and the nano scales would need to be described by some future creature. Since history tends to be written by the victors, in this case the femtolects, I have little idea what such a creature might be like, other than to say it would probably look a bit like a neutron star.

A femtolect could perhaps signal with gluons. A gluon is a theoretical entity in modern particle physics that is hypothesized to glue quarks together. Quarks combine in various ways to form larger particles such as protons, neutrons etc. We are now talking about sizes in the femto-meter range, i.e. a million times smaller than nano-meters. If we assume for the moment that femtolects still have to respect the speed of light limit, then they will be able to signal between quarks at a rate that is a million times faster than nanolects could signal between molecules. Femtolects could "think" a million times faster. Note, that the same type of reasoning applies when human electronic engineers shrink the size of their chip components to make them signal faster.

Not only that, femtolects could pack themselves at a density 10^18 times greater than nanolects, i.e. 10^6, cubed. In the same unit of volume and unit of time, a femtolect could process a trillion trillion (102^4) times more information than a nanolect, and thus outclass the nanolect by as much as the nanolect could outclass human beings.

All these powers of ten may not mean much in terms of human intuition, so let me attempt to translate them into more meaningful terms. A nanolect consisting of a mere trillion atoms would be so small that it could not be seen by the human eye. It would be the size of a bacterium. A nanolect consisting of a million trillion trillion (10^30) atoms would occupy about a cubic meter. An asteroid sized nanolect, of dimensions 10 kilometers cubed, would contain roughly 10^42 atoms. A planet sized nanolect of dimensions 10,000 kilometers cubed would contain roughly 10^51 atoms. But, if femtoscale entities were used, i.e. nucleons, quarks, etc, then all of the above numbers would become too small by a factor of a trillion trillion, i.e. one would need to add 24 to all of the above powers of ten.

Maybe there is a trend here. Perhaps true godlike intelligence resides well below the elementary particle level. If we continue this line of reasoning then the so-called elementary particles may not be elementary at all but be whole godheads!