Re: A Race of humanoid drones for labor.

Eugene Leitl (
Fri, 30 Aug 1996 14:34:25 +0200 (MET DST)

On Thu, 29 Aug 1996 wrote:

> On Thu, 29 Aug 1996, Chris Hind wrote:
> > [ dimwits for slave labour by genetech ]
> >
> What could a creature "with little or no grey matter" do that a machine
> or AI could not do faster and cheaper? It appears to me that the only

Currently, robotics does not provide us with machines capable of robust
navigation, vision, etc. AI is a somewhat unfortunate term, since it is
still loaded with strong AI things like Lisp workstations, expert
systems, inference engines & the like. This is brittle stuff which has
nothing to do with robust robotics.

Furthermore robots are far from being cheap. They do not autoreplicate.
On the other hand, they are tireless & accurate (no labor's union
membership, too) if it comes to doing well defined, tasks requiring little
sensorics nor complex realtime processing of input. Most are still
dedicated automates, not all-purpose reprogrammable robots. Statistics
lie, sometimes.

But you just can't trust idiots running a car assembly line. (This is not
a 3-banana problem).

> reason to build any tool is to get the job done quicker, cheaper and more
> thouroughly than the old tool. Why assign tasks to a new tool that is
> inferior to the one it replaces. By the time a race of humaniod drones

Hmm, I failed to notice a dramatic increase in the use of robots just
lately. Manual labor has still a fair niche. Only large hardware ANNs
will change this situation, and these are unlikely to appear too soon.

> hit the market, AI should already have a considerable marketshare of
> brainless labor. , Machines already out perform humans, horses and oxen

If AI=dumb robots, then yes.

> at many brainless tasks, and better machines will(be able) to out perform
> them in the remaining drudgeries.

You know, most people don't like being unemployed. Trivial things like
housing, food, clothing, and such. They might get pissed sufficiently to
start an insurrection. There are some areas in Krautland with 15%
unemployment, trend's still up.

> A big question is how much will it cost to maintain a drone work force.
> I'm sure that genetic engineers could make more efficient organisms than
> those currently available, but they'll still need food and waste

I know a bit about DNA tinkering, and I think this damn unlikely. This is
one of the reasons why I don't fear genetically modified organisms
release in ecosystems _which are in equilibrium_. Alas, most things
humane are not. Woe! Fear the engineered virus. Welcome to Project Blue.

> removal. Who will provide the mindless service of feeding and clean up
> after the drones-- machines, humans, or more drones?

If they are sufficiently smart to operate the machinery, they sure are
smart enough to feed themselves.

> Moral Question: How is bulding an race of drones in a laboratory for
> labor different from just building designer humans to perform those same
> tasks? Would these humans be created or enslaved?

This is pure slavery, of course. Tinkering with human DNA to produce a
race of slaves is immoral as hell, imo. Currently, _any_ tinkering with
human DNA is a strong no no since the yield is so low. Site directed
mutagenesis or DNA snipping/insertion is easy on bacteria where you
select automagically e.g. by ampicillin resistancy, but just try this on
human embryos. 99.999 % nonviable. Hoo, boy.

> Michael Bowling

| | cryonics, nanotechnology, |
| | >H transhumanism, [...] |
| | "deus ex machina, v.0.0.alpha"|
| | >H: "alpha-->omega" |