[extropy-chat] I'm back.

Dan Clemmensen dgc at cox.net
Sat Nov 15 19:55:10 UTC 2003


Eliezer S. Yudkowsky wrote:

> Dan Clemmensen wrote:
>
>>
>> Will this happen by 2006? Well, "I still have no viable justification 
>> for this assertion."
>>
>> Sorry for the weasel words. As you see, I'm mostly holding this 
>> position for old time's sake.
>
>
> *head hits keyboard with dull thud*
>
:-)
OK, try this. I can find no compelling argument against the "SI-by-phase 
change" scenario. This scenario logically becomes more probable as time 
goes on if you accept it at all. When will the environment become 
supercritical for SI? Is it supercritical already? What is the shape of 
the probability curve for probability-of-SI versus the technical 
richness of the available technological resources near the supercritical 
transition point? Since all of these are unknown, it is very difficult 
to predict the time of the singularity with any accuracy, and "May 1, 
2006" is as good as any.

Note that the shape of the singularity probability function is unrelated 
to the shape of the singularity function.
(Sincere apologies for lack of rigor. If my meaning in unclear, I'll try 
for some precision later.)  The singularity function itself describes 
the rate of technical advance. I can find no valid argument against hard 
takeoff, so for me this is a step function.

To rephrase my 2006 "prediction" let's say that my best guess is that 
the probability of the SI phase-change increases rapidly near the phase 
transition point of the technological capacity. Furthermore, the 
technological capacity curve is "Moore's Law" or better and will remain 
so for at least the next decade, absent a singularity. If you look at 
both of these together, you see that we will have little or no warning 
of the singularity. I have no logical reason to retract my "prediction."

Unfortunately, this non-quantitative logic does not permit me to assign 
a probability to 2006. It only allows me to assert that the probability 
increases with time.

When I look at today's technologies, I see a great many pieces that 
might be building blocks for the SI. I see more of them that I did in 
1996, and I see a general increase in the ways the pieces can be put 
together. thus, I have a gut feeling that we are already supercritical.

I will stick with my original prediction: on or before May 1, 2006.

I note with interest that the consensus predictimn among futurists 
appears to be coming in rather than stretching out.




More information about the extropy-chat mailing list