On Tue, Aug 13, 2002 at 04:12:14PM -0300, Jecel Assumpcao Jr wrote:
> On Tuesday 13 August 2002 17:12, Alan Grimes wrote:
> > If AI is even 1/10th as powerful as the projections I've been
> > reading,[....]
> 
> 5 years before the projected date (Smalltalk's 60th birthday?) they will 
> be 1/10th as powerful :-)
> 
> Of course, Ray Kurzweil would say this is far too pessimistic since we 
> are talking about a double exponential growth and not simply Moore's 
> law (don't try http://www.kurzweilai.net in Netscape/Linux or you will 
> get into an infinite recursion).
> 
> Jaron Lenier claims that we have already reached the limit of what we 
> can build in software and that faster hardware won't help. 
> http://www.edge.org/3rd_culture/lanier/lanier_index.html
> I don't agree with him - computing power does matter. Alice Squeak isn't 
> a photorealistic 3D world because we don't know how to do it (we 
> learned about ray tracing, radiosity algorithms and so on) but only 
> because our machines are too slow.

I had read this before, and skimmed it again, and didn't see where he
made this claim.  Could you point it out to me?  The closest thing
that I saw was that after 30 years, Unix (Linux, MacOS X) is again the
Next Big Thing.  However, he attributes this to bad design rather than
bumping up against intrinsic limits.

Joshua


> 
> -- Jecel