Curtains lifting, worlds simulating, heaven descending

[Brian Atkins]

Kent Spaulding wrote:
> 
> > You are stereotyping. You are also thinking linearly ("not in our
> > lifetimes..."). Our current rate of technological progress is
> > much much faster than 100 or even 20 years ago, and it will be
> > much much faster in 2010, 2020, etc. So extrapolating out your
> > view of the future based on the current rate of change is a
> > mistake. Frankly, even based on our current rate you are far too
> > pessimistic/unrealistic. You don't have to look far to spot many
> > companies working right now on anti-aging biotech. So even if
> > AI and nanotech are impossible (unlikely) there doesn't appear to
> > be a good reason you can't live way past 100 as long as you aren't
> > too old right now.
> 
> I've heard this 'think exponentially' thing many many times before. Okay, so
> technical progress is exponential.  I can see how there is a network effect
> between technologies and knowledge, in that the things learned in one area
> are applicable to others.  But still, I have a question though, because I'm
> not really sure what this means.
> 
> Does this mean that progress within any one technology is exponential?
> 
> Or does it mean that our acquisition of knowledge that is then applied to
> technology as a whole, increases exponentially?

It means your first answer- the tech itself grows exponentially. The best
know example is probably Moore's Law, although actually Moore's Law was
only about chips. In fact, computing power has been exponentially increasing
since at least the end of the 19th century when they used a primitive
device to calculate the US census. The "integrated circuits" portion of
the curve is so far the 5th stage of this particular technological trend
if I remember my kurzweilai.net info correctly.

> 
> So - can we apply exponential thinking to individual technologies?  If yes,
> how do we know where we are on that technology's curve?  Clearly, we aren't
> really very far along the fusion-reactor curve.  Of course, we've thought we
> were for the last 50 years...
> 

I am not sure exactly what you are asking here, can you clarify? You
can't really say where you are on a curve until you actually get some
data points. For fusion one set of data points might be the amount of
energy you get out compared to what you put in... recently a couple
weeks ago I read a news blurb about how they now have quadrupled the
output in a test reactor. So progress is being made I think. Fusion
though is still going to be a post-Singularity technology I think, at
for real world commercial use.
-- 
Brian Atkins
Director, Singularity Institute for Artificial Intelligence
http://www.singinst.org/