www.mitchellhowe.com


The Singularity Q&A

Q: Does the Singularity depend on the continuation of Moore's Law?

A: This is kind of like asking whether Europeans would have discovered the New World without Columbus.  Of course they would have, but there is no telling how much longer it would have taken or how different history might have unfolded as a result.

The ongoing reliability of Moore's Law helps propel almost every scientific endeavor today, and thus indirectly effects each of the major technological approaches to the Singularity.  If a stubborn bottleneck in computer engineering preempted the introduction of better chips, however, there would be nothing to stop production of existing designs.  A nice feature of computer chips, particularly in the more powerful computers used for research, is that they can be linked together in large numbers to boost performance.  You can't stick a thousand engines in a car to make it go faster, but you can build a supercomputer this way.  A disturbance in the Law would thus result in greater use of this parallel architecture, in which computers are upgraded not by replacing the chips, but by adding more of them.  This is by no means the ideal solution, but the total computing power available to researchers and others would continue to increase by at least the rate of new "old" chips coming off the assembly lines.

But just as Columbus was not the only European who could have discovered the New World, any bottlenecks in Moore's Law are likely to be only temporary.  Computation is a very simple operation, which could be carried out in any number of ways within any number of different materials; transistors on wafered semiconductors just happens to be the most efficient one we know of right now.  Alternatives such as optical computing, nanotube "rod-logic", and true three-dimensional circuitry each offer theoretically superior performance if or when the manufacturing methods are worked out.




[Back to Futurism] [Back to Main]  ©2002 by Mitchell Howe