/lit/ - Literature

>>12036883
A logistic curve still has a singularity though (that is, the point where the slope is infinite, the center of the "S"). The main premise is also more about the large-scale change that will occur -- the time after it will be fundamentally different than before. It is essentially more of phase transition than anything else (which also follow similar parameter curves e.g. the temperature-energy curve for boiling water or tempature-magnetization curve for a paramagnet).

In this case, the two defining things are computation and time. It's a phase transition visible on a plot of time vs computation, in which the state (this time, a state of intelligence, as opposed to say a state of matter) will be qualititavely different than the time before it. There are many correlates we can tie to this- intelligence explosion, self-replicating machines etc; the main thing is all about computation though.

I've never seen anyone argue that the singularity will be a time of eternal exponetial growth, most of the main proponents will readily admit the maximum theoretical computational ability of matter (for example Bremermann's limit, which realistically far exceeds the actual possible rate).