[ExI] Ray Kurzweil's predictions regarding the Singularity

[John Grigg]

As with many of you, I am utterly enthralled by Ray Kurzweil's
predictions about the march toward AGI that is many times faster and
brighter than humans.  But I am just not so sure about his 2045 date
for a Singularity.

Michael Annisimov and others have challenged his claims, but I would
say generally Kurzweil has a sound track record.  But I do think his
statement that the global Google network makes his crucial prediction
of roughly human brain computing power (but without actual AGI)
actually being achieved by 2010, as really grasping at straws.  Or am
I being too hard on him?

And so not counting the Google network, when do any of you see a
supercomputer existing that can do 30 petaflops a second, or whatever
level of computational power that would be required to equal the human
brain?  2015? 2020?

How many of you still consider 2045 as a good date for the
Singularity?  I am beginning to think it will take at least one
additional decade for his vision to come true.  And this worries me
because I may not be alive in 2055...

But then considering the *possible* rate of take-off if a human level
AGI (but thinking so much faster than humans) is allowed to
exponentially improve itself, I can also envision (though unlikely) a
Singularity right around 2035.

This really matters to me.

John