Losing the race.

From: Mike & Donna Deering (deering9@mchsi.com)
Date: Sat May 04 2002 - 12:17:35 MDT

The race between those building an AI by reverse engineering the human brain physically and those reverse engineering the human mind functionally will be won by the brain builders.

I am sorry to say this. I wish it wasn't so. I think the implications for friendliness are not good.

The reason the brain builders will win is simple. What they are doing is completely mechanical and will therefore benefit from technological exponential advancement. Look at the Human Genome Project as an example. The project started in 1988. By 1995 5 million base pairs had been sequenced. 1998 100 million base pairs. June 1999 1 billion bp. June 2000 2.91Bbp. Over half of the base pairs were sequenced during the 9 months prior to June 2000. And actually Celera completed it in 6 months by February but agreed to delay until the joint announcement in June so as not to embarrass the official project. Everyone was surprised by the exponential growth curve of base pairs sequenced by time. We shouldn't be surprised when the same thing happens with the brain builders. What Ben Goertzel, Peter Voss, Eliezer Yudkowsky, and others are doing is at the opposite end of the scale from mechanical, and will advance relatively linearly until completion of the working prototype, too late.


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:38 MDT