[agi] "doubling time" revisted.

[Mike Deering]

Doubling time - the time it takes for the power of a computer to double for the same dollar.   example:  suppose you purchased a computer on a certain date and at a later date you found another computer with double the power for the same price, or you found another computer with the same power at half the price, or some mathematical computation resulting in double the power per dollar.   The reason we would be interested in this doubling time value is that available computer power is an important factor in the timing of the Singularity.  For instance, Eliezer writes, "The intelligence required to create AI.  Decreases with increasing computing power." in CFAI: 4.1.2.  Eliezer also writes, "The total processing power available to an average research project will increase faster than chip clock speeds (i.e., maximum parallel speeds increase faster than maximum serial speeds).  The total networked processing power on the planet will increase even faster than that; a doubling time of nine months is probably an underestimate." in CFAI: 4.1.2.  Ben Goertzel writes,"We are limited tremendously by CPU speed and RAM capacity.  Either greater CPU speed or greater RAM capacity would be valuable, but the biggest boost would be both together.  We could utilize essentially any amount of CPU speed or RAM capacity.  No limit in sight.  Having a CPU with (for example) 10x greater speed would have a HUGE positive impact on some of the work we're doing." on 1/4/2003 [AGI].  Ben also writes, " If we had vastly better CPU's and vastly more RAM, the amount of time to get to a complete working implementation of a Novamente system might be reduced to 2/3 what it is right now." on 1/4/2003 [AGI].  I accept these experts opinions that the availability of more powerful computers would significantly push up the date of the Singularity.   So what is the doubling time?  How do we calculate it?  On the http://www.bjklein.com/sing/default.htm website it says, "Computer power is doubling at a slowly accelerating rate: every 18 months currently."  This seems to be the general consensus based on a little google statistics:   google search                                                number of sites singularity  doubling  "24 months"                        60 singularity  doubling  "20 months"                        4 singularity  doubling  "18 months"              163 singularity  doubling  "12 months"                        78 singularity  doubling  "11 months"                        8 singularity  doubling  "10 months"                         8 singularity  doubling  "9 months"                          28 singularity  doubling  "6 months"                          49 singularity  doubling  "3 months"                           25 singularity  doubling  "1 month"                            10 For purposes of assessing the effects on the Singularity we need to use the computer with the fastest doubling time out there.  AGI will come out of the computer market with the best power to dollar ratio.  This market is the entry level desktop.  The current entry level desktop sells for $600 and has a 2 gig Celeron processor with 256 meg of ram and a 40 gig hard drive and other standard periferals.  As you can see from this graph the price of desktop computers has been dropping since 1990 and continues to drop:     While at the same time the power has been steadily increaseing:     Based on available data how are we to calculate the doubling time extrapolation into the future?  On 1/6/2003 Stephen Reed writes. "Progressing from -50 db HEC to 0 db HEC in 22 years is equivalent to Moore's Law doubling every 16 months. [ 2^16.61 = 100025, 22/16.61*12 = 15.9 ]"  A careful examination of this formula shows that Stephen is merely averaging the doubling time over the past 22 years and applying that constant to the next 22 to arrive at his crossover date of 2021.  A constant extratulation of an average doubling time is not the correct method to project an exponentially changing value.  Unfortunately I haven't been able to get good historical data on entry level computer market.  I would welcome any assistance.  This is my current extrapolation:     DATE        DOUBLING TIME        DROPPING RATE   1900          48 months                     1915          42 months                    6/180    (6 months in 180 months) 1930          36 months                    6/180 1945          30 months                    6/180  1960          24 months                    6/180 1975          18 months                    6/180 1980          17 months                    1/60 1990          15 months                    2/120 1999.5       12 months                    3/114 2001          11 months                    1/18 2002.3       10 months                    1/15 2003.3        9 months                      1/12 2004           8 months                      1/10 2004.7        4 months                      1/8 2004.6        3 months                      1/6 2004.9        2 months                      1/5 2005.2        1 month                        1/3 2005.3        <1 month                      1/1            Singularity!     Mike Deering. www.SingularityActionGroup.com    <--new website.