7 Comments

IMO, it is unreasonable to expect people to get away from using "ancestor", "descendant" "parent", "child", "sibling", etc. in the context of memetic evolution. Calling cultural evolution "revolutionary rather than evolutionary" seems generally dubious. You can have revolutions in evolution - see "Genetic Takeover" by A. G. Cairns-Smith. In the future, there will likely be a "memetic takeover" - along the lines that Hans Moravec described in "Human Culture: A Genetic Takeover Underway". In that respect, cultural evolution is likely to result in a revolution. It will still be evolution, though. Maybe what Richard Dawkins described as "a new kind of evolution".

Expand full comment

I start with Robin's first sentence, "If competition, variation, and selection long continues, our worlds will become dominated by artificial creatures who take a long view of their future, and who see themselves as directly and abstractly valuing having more distant descendants.", with a simple question, is this a coherent and logical assertion? I would submit that it is not. The if condition of the assertion is underspecified, relative to the then condition. Robin need to fully specify his system before he can make declarations that are causal in nature. There are no reasons to accept the if condition approximates "our worlds" or a semi well specified system. Further, the "long view" need not follow as posited. I am not saying there may be an interesting thought in this posting, but I am a loss to see that as Robin has articulated his mind experiment that there is anything coherent at the present. As Jonas and others are positing how does our standard evolutionary model fail, and how does this idea step beyond it. A suggestion, develop a simulation that can support your mental experiment.

Expand full comment

Yes. For all the parallels between memes and genes, there are big differences in the constraints on their evolution. Big enough to use a different words for memetic vs genetic relationships IMO.

Expand full comment

Saltation. It is a widely-discredited idea, though. Hans Moravec used the term "Mind Children" to indicate that future engineered organisms would be descended more from our memes than from our genes. It is not bad terminology, IMO.

Expand full comment

The production of descendants by evolution is gradual, whereas the relationship between a human being (or a committee of human beings) and an autonomous machine that he produces (or they produce) will be abrupt--revolutionary rather than evolutionary. I think that for most people the term 'descendant' is associated with the evolutionary process.

Expand full comment

I don't see why that's any more misleading than calling you the distant descendant of an amoeba.

Expand full comment

It is potentially misleading to refer to machines produced by human beings as the latter’s “descendants”--even more misleading so to refer to machines produced by other machines that were in turn produces by other machines that were in turn produced by other machines that were . . . produced by human beings. As for what kinds of machines other machines will produce—that seems quite unknowable at present, orders of magnitude harder than, say, predicting the terrestrial climate a century from now. But I agree that the prevalent machines at any given time will be geared to produce other machines (not necessarily copies of themselves) that will have a relatively high probability of, in turn, producing successful machines.

Expand full comment