[ExI] nick's book being sold by fox news

Anders Sandberg anders at aleph.se
Mon Oct 27 23:34:53 UTC 2014


Brent Allsop <brent.allsop at canonizer.com> , 27/10/2014 10:46 PM:

http://canonizer.com/topic.asp/16

There are very strong arguments for why concern over AI is just dumb, a complete waste of time, and so far at least, there is more consensus for these arguments, than there are for the fear mongering camps.
I think this shows the problem with Canonizer. The "Concern over "Unfriendly" AI is a Mistaken Waste of Time" camp's claim "Intelligence is necessarily Moral" does not engage with existing (peer reviewed) literature (c.f. http://link.springer.com/article/10.1007%2Fs11023-012-9281-3 - free versions exist online too) or known counterexamples (AIXI-style paperclippers). 
I know, I know, the proper response is of course for me to form yet another camp and add these arguments. And so on. Fine. But I do not have the time. And I happen to *like* the idea of Canonizer. 
The problem is that you need a significant critical mass, and it needs to include the people who actually know stuff. This is a hard problem when trying to get an entire debate onto a single forum: the LessWrong crowd are pretty busy over there, and we at FHI/MIRI are writing papers (or grant proposals) over here. So you get the same situation as for information markets with too little liquidity: very biased and chunky information because it only represents whoever happens to be posting or the imprints of somebody with social capital sending out a call for activity. 
It seems to me that in order to get debates like this done properly we need better ways of feeding existing literatures into the Canonizer framework to produce initial scaffoldings so there is no need for the locals to reinvent the wheel in a biased way. And then there is the vexing question of how to ensure a steady flow of commenters - in my Wikipedia models that turned out to be more important than their average quality. 

Anders Sandberg, Future of Humanity Institute Philosophy Faculty of Oxford University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20141028/b54b5d10/attachment.html>


More information about the extropy-chat mailing list