[ExI] Could the Singularity be 1 year away?

[Travis Porco]

> From: Jason Resch <jasonresch at gmail.com>

> While doing research to make this presentation:
> https://docs.google.com/presentation/d/1erbKPY7V7fTgY61n-BONBPvdhJn1GEdE8HAuuDH8zO8/edit?slide=id.p#slide=id.p

> I discovered that the underlying macrotrends (for the economy, human
> population, the pace of history, and computing technology) aren't
> exponential, but rather, are best described by hyperbolic functions.

Yes, the phenomenon of 'finite escape time'...the standard equation giving
exponential growth is dx/dt = kx, with k>0 of course; if you write dx/dt=kx^2
with k>0 you get finite escape time solutions.  I think even Kurzweil
had this in the appendix to The Singularity is Near (or was it Spiritual
Machines).

> Unlike exponential functions, hyperbolic functions reach infinity within a
> finite amount of time.

> Curiously, two independent models, based on different data sets, put the
> singularity for these exponential models at a time near 2027.

>From your lips to God's ears...ominous forces of war are unfortunately
gathering out there as well.

> In the above linked presentation, I speculate at what could be the ultimate
> driver of these trends, and trigger for shooting towards infinity. Based on
> what I show to be possible with existing AI tools, I think we really may be
> just a year or two away from seeing truly incredible progress.

> Jason

--tcp