Sunday, December 04, 2005

AI and human evolution transcends government

Autonomy (the experience and discussion of) in early metaverse worlds like Second Life is most interesting in the sense that this is presumably a precursor to what autonomy in digital environments will be like with fully uploaded human minds.

In the crude early stages of these metaverse worlds, an unfortunate theme is facsimile to reality. Digital facsimile to the physical world is evident in the visual appearance dimension; how avatars, objects and architecture look, in the dynamics of social interaction and community building, and in conceptual themes. The tendency is to recreate similitudes of the physical world and slowly explore the new possibilities afforded by the digital environment. Presumably, dramatically more exploration will occur in the future and in freer digital environments without as many parameters established by the providers.

Analogous to the physical world is the theme of the check and balance between autonomy and community in digital environments. There is freedom to a degree and norms and enforceable codes if norms and laws are not maintained. This is seen in all existing virtual worlds; Amazon, eBay, Second Life, etc.

In the near term, humans will likely continue to install and look to a governing body for the enforcement of laws. If their power base can be shaken, governing bodies will hopefully become much more representative and responsive (say the full constituency votes daily on issues via instant messenger). Governing bodies could also improve by being replaced by AIs who would not have the Agency problem.

Post-upload, physical location as a function of governance will be quite different, and, in fact, human intelligence can presumably evolve to a point of not needing external governance.

A separate issue is whether non-human intelligence will be the governor of human intelligence and this is probably not the case for several reasons; first, the usual point that machine intelligence finds human intelligence largely irrelevant, and secondly, machine intelligence understands that system incentives not rule governs behavior.

3 comments:

Timothy Moenk said...
This comment has been removed by a blog administrator.
Timothy Moenk said...

If you haven't caught wind of it yet, you might be interested in reading Beth Noveck's A Democracy of Groups.

Beth is one of the minds working on the concept of e-governance, and is very forward thinking in that she includes virtual worlds as part of her calculus. :)

I've been wonderinghow efforts towards granting legal rights to transhuman entities such as agi's might be affected by a change in landscape in the direction of Beth's ideas on providing a legal framework for granting legitimacy to groups.

I'm also finding it interesting that this discussion is beginning to take root around current developments in virtual worlds.

LaBlogga said...

Thanks for the comment, Timothy, you make some good points. Yes, I have been following Beth Noveck's work & agree, she is doing some quality thinking on the topic. I still think government diminishes due to obsolescence in the future except regarding law enforcement. In the early days of AGIs, we would expect them to operate somewhat within the human-developed world and therefore demand legitimization through legal recognition but that need will soon pass as they supersede operating in the human world.