Any progress towards AI at all? was Re: [extropy-chat] Futures Past

Brian Atkins brian at posthuman.com
Sun Oct 9 23:00:07 UTC 2005


I think when discussing all this, it's interesting to see how sometimes the year 
of adoption or development of a particular technology can be influenced 
significantly by relatively simple organizational techniques (or lack thereof).

A good example I think is the successful DARPA Grand Challenge event yesterday. 
Prior to this the government had funded this type of idea for quite a long time. 
All the necessary research into having computers work with various sensing 
systems had been done. The hardware necessary for a success, from sensors, 
drive-by-wire, computers, etc. was already here years ago. Yet we didn't end up 
with working autonomous offroad vehicles until yesterday. It took a well-funded 
competitive prize to finally get the major universities and groups to sit down 
and write the complete software and do the offroad testing necessary to complete 
the working vehicles. By organizing a public prize, DARPA found a way to scare 
the established funded groups into worrying that they might be upstaged by new 
outsiders, and this proved to be a very effective motivational strategy.

Without such organizational efforts? We would not have those vehicles right now. 
Who knows how much longer it would have taken.

There are tons of other examples where a technology could have technically been 
completed years earlier, but just didn't for various non-technical reasons. I 
think at this point in history there are many such things that could be 
significantly accelerated if organization efforts could be focused on them. It 
could be simply funding something unfunded, developing prizes (MPrize is a good 
example), etc.

Overall in regards to this thread, I would say that the feasibility and 
prerequisites are more important to figure out rather than simply trying to 
guess a particular year of creation or mass adoption.
-- 
Brian Atkins
Singularity Institute for Artificial Intelligence
http://www.singinst.org/



More information about the extropy-chat mailing list