This thinking avoids the reality of the situation in software as well as the 
biology in question. The result may not self-organize into something we could 
physically wrestle and which has an inherent understanding of knife usage for 
instance. As fairly fragile biological units ourselves, harm can come from lots 
of directions. Through evolution, and the separation of continents, we adapted 
to 'attacks' from external things such as disease (biological adaptation 
through selection over generations) or other agents, such as UV radiation 
(melanin production).

Where we are is the joint evolution of everything that's here and now, having 
come the same way in the same environment, slowly honed into it's current form 
by standing the test of time so to speak, as well as slowly adapting, mutating, 
etc over the ages.

Now you create something in the lab that has no relevance to our environment, 
and we have no knowledge of it's pending evolution. Most will fail at first, 
failure being defined as not acting as expected in controlled cause-effect 
relationships. However, failure or not, these introduce combinations of complex 
variables into an even more complex system (the ecosystem and us). A system 
where we can't even agree on whether it's getting warmer or not, much less the 
effect of some new life form on the complex system.

The best analogy would be our system as a nuclear missile and it's control 
cabling (an analogy within a reality perhaps...). It presents us with a mass of 
control cabling - our interaction with the 'system'. We've studied it and even 
sent some signals back and forth, but have nowhere nailed down what everything 
and every wire does. This new life form is like a super giant PIC perhaps, or 
even a PC, and in introducing it, we connect it to a large group of wires (it's 
interaction with the environment or us). What we don't know is if there are 
'bugs' in the code, or even how we should structure the code in the computer 
such that some combination doesn't launch the missile (and many combinations 
could). Who among us would just randomly attach IO lines to these wires and 
boot up? Of those that might, there would be big "nothing happened" events, but 
not all the time. And as we got cocky, any controls that might have been in 
place (put a scope on the line first to see if there's a signal before 
connecting) get loosened (use a DVM and just check for voltage perhaps). And, 
more people decide to do the same, seeing it might be safe... the danger still 
lurks there though...

If we're so susceptible to pandemic like a flu, which is merely a variant of 
something our bodies have been through for eons, then so much worse it is for 
something entirely new and lethal. Consider it might not kill for 5-10 years, 
perhaps by cancer, and it were deemed 'safe' in the lab and released for use in 
the 'system'? At least the organisms we live with we have a known 
'relationship' with us - avoid plague, keep the antibodies for measles, etc. 
Look at the effects of various synthetic chemicals, which are so much simpler 
to model and contain the an organism.

The other problem is mutation. It will have very little biological 'experience' 
in our environment, so it will be fragile, but HIV is also fragile (it can only 
live for a short time outside the body) and look how deadly it is. It will no 
doubt mutate. If it doesn't reproduce, it will still be reproducing it's own 
cells for growth, unlike a chemical, and perhaps it's own 'cancer' will be the 
real problem - found after the fact that it makes it lethal to us.

We are also not the target. Any thing that upsets the balance out there can be 
a target. Like a dirty bomb, the destruction isn't the 'bomb', it's the 
residual effect. Perhaps the decay of this 'body' creates substances that 
radically kill off plankton. No one thinks of that... So, a few dead 
cyber-pet-fish get flushed (or many) and we don't notice it until higher orders 
of sealife (seafood that is) dwindle and we're stuck with food shortages and a 
diseased ocean...

And consider the simple effect of protein molecule twist direction in prions - 
the cause of Mad Cow. Something that wasn't even detectable or looked for... 
And, they're not destructable.

Lastly, even if one argues their way past all of this and is convincing that 
with the right caution and small steps, we can, over a long time, cautiously 
develop helpful life forms, from microbes to pets, the reality is that the 
industry will be funded by thousands of competitive venture capitalists and 
investors, in hundreds or thousands of companies who want immediate big gains 
and are in it to rush the latest biological gizmo to market. All in a hundred 
different countries who will see it as their way to economic success (or 
military success) with various degrees of laxity in whatever controls do happen 
to be put into place. In other words, no matter how bad an idea, there will be 
money and a place that will be willing to 'risk it' and foist it upon the world 
with whatever longer consequences there may be.

-Skip

David VanHorn wrote:
> Well, my experience with software is that most mistakes do not result
> in a powerful AI computer that takes over the world.
> Usually they just sit in a corner and drool, sometimes with smoke.
> 
> So one would hope that the genetic software would be much the same,
> "legend" movies notwithstanding.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist