I believe I answered the exponential limit claim already...exponentials reach limits only until surpassed by a new paradigm. My example was processor technology. Something claimed by critics for many years...that there would eventually be a materials limit in Moore's Law, but which has again been surpassed: http://www.trekbbs.com/showthread.php?t=153184
RAMA, you make the mistake of assuming that new paradigms will keep appearing, based on the recent past, on the scientific/technological revolution.
In other words, you make the mistake of assuming you can extrapolate forward indefinitely - a mistake Christopher already pointed out to you.
Indeed, one can prove logically that there are not an infinte number of paradigm shifts in our future:
There are a finite number of laws of nature AKA there are a finite number of combinations one can make using them.
Almost all these combinations are useless - they have no useful result, are not technology.
The few combinations that are useful are finite AKA they will not appear ad infinitum.
One erroneous assumption of singularity proponents is forward extrapolation ad infinitum - that there is an infinite number of paradigm shifts/advances posible.
Another one, made by some of them, is the assumption that the frequency of appearance of these infinite paradigm shifts/advances will increase exponentially (which is how Kurzweil came up with 2050 as the date for singularity).
In many fields, this assumption was already proven wrong.
It's not a mistake, there are 3-4 different technologies already being explored that can make faster processing speeds possible, the first one was predicted and has already occurred. The mistake is in expecting that we won't
come up with new technologies, which is contrary to prior historical and current demonstrable evidence...as long as there is a theoretical limit to processing speed(in fact no one has claimed this was infinite), but it is much higher than we can reach now, there will likely be away these new technologies can reach it.