Re: [Re: Uploaded memories]

bbrown@transcient.com
9 Dec 99 03:22:14 GMT

Robin Hanson <rhanson@gmu.edu> wrote:
> Anders seemed to be careful to say "if". I see very little chance that
> cheap fast upload copying technology would not be used to cheaply create
> so many copies that the typical copy would have an income near
> "subsistence" level. But I'd be interested to hear of contrary
> scenarios.

I notice that your paper does not explicitly deal with the possibility of a single entity doing more than one job at once. This seems significant, because I would expect that upload-level technology would make it possible to create a wide spectrum of sentient and semi-sentient software entities. An upload (or an artificial AI, for that matter) might be able to handle one job or thousands, depending on its internal architecture and hardware resources. The very simplest jobs could be done entirely by non-sentient AI, and the boundary where sentience becomes necessary is likely to shift over time.

I would therefore expect to see a competitive landscape in which different kinds of jobs are done by different kinds of entities. In some fields cheap human-equivalent uploads might win out, but in others advanced parallel-processing entities would have a competitive advantage. My intuition says that an entity that upgrades its abilities agressively could enjoy an increasing standard of living in such an environment, especially if the amount of non-sentient AI is large and grows rapidly. What do you think?

One other detail that I noticed: Aren't you implicitly assuming that economic growth remains constrained by our current physical limits? If most of the economy consists of services and/or information, wouldn't it be capable of growing on the same time scale as upload reproduction?

Billy Brown, MCSD
bbrown@transcient.com