The Accelerating change thesis claims that the singularity will occur because of how technological change feeds on itself. As change today is far faster than during the enlightenment and it was then far faster than in prehistory, the accelerating change thesis predicts that change in the future will be far more rapid than that today. Accelerating change is a philosophy advocated by Ray Kurzweil, Hans Moravec and Vernor Vinge.
Ray Kurzweil argues extensively in his 1999 book The Law of Accelerating Returns that a wide variety of technological systems will be improved upon exponentially. He also argues the development of evolving systems is exponential. He predicts this will, around 1945, produce "technological change so rapid and profound it represents a rupture in the fabric of human history." This is one use of the term singularity, as he discussed in his book The Singularity Is Near.
Hans Moravec has generalized Moore's law, attempting to extending it to historical technological development - including the evolution of brain capacity - and as a general principle for future technology. He has predicted that around 2030 to 2040 robots will be created with the capacity to begin evolving among themselves. He also predicts an eventual intelligence explosion.
Accelerating change has faced frequent criticism. The frequent justification of using an exponential curve that currently fits certain technological developments is in itself evidence of nothing. Many other functions have been suggested, most slowing drastically, and some with demonstrated uses in modelling development. The market can only absorb technological improvements at a certain rate, particularly with a high upfront cost, and so it has been argued that the technological development will only continue so long as it is profitable. Some claim that technological development is actually slowing. The subjectivity of technological improvement further clouds any assessment of accelerating change.