Self-replicating machine
-
Who says the SRM needs to be that complex? It could be so simple that an actual memory storage isn't needed. This doesn't mean that a 'DNA' like substance isn't nessecary. It is, but if it is simple, a UC can be smaller and more simple.
@Box-Poorsoft: That's the attitude i'm talking about!
@Greymatter: A machine that grows could..somewhat be possible using pistons and paste to make brick. I'm sure there are plent of other ways to make the SRMs grow.
-
well, the memory isn't necessary - GOL sometimes just generates pixels of itself right beside itself. This way the particle, by definition, is a SRM. Fnding a machine that just happens to replicate would take forever, whereas constructing a predictable machine would be easier. The universal constructor is VERY useful. why not, after duplicating, have the SRM make computers which would be put to good use?
Isn't the information singularity the moment in time that computing power is growing at it's full potential? That woud require SRM's which double as computer factories.
-
The device needs to be complicated enough to be 'predictable'. You could just brute-force search for a pattern that replicates, but with hundreds of elements and a 700 X 350 px canvas, plodding through formal concepts would actually be easier. How would the your machine create material? or would it find material? The bray would also just decompose. Man people have thought about replication in the past. It has been so complicated that some thought that Biology is magical.
