I am not aware of any published papers proposing pure RSI (without input) as a 
path to AGI.  But in 2002 there were experiments with AI boxing to test the 
feasibility of detecting and containing unfriendly AI, discussed on SL4. 
http://www.sl4.org/archive/0207/4935.html (The results showed that unfriendly 
AI would almost certainly escape).

The implicit assumption of RSI in this approach seemed important enough for me 
to debunk it.

I could also argue that the limitations on RSI would constrain a hard-takeoff 
singularity to an explosion of computational power, not of knowledge. But I 
think that might be a stretch. Not everyone agrees that there will even be a 
singularity in the first place.


-- Matt Mahoney, [EMAIL PROTECTED]


--- On Tue, 11/25/08, Trent Waddington <[EMAIL PROTECTED]> wrote:

> From: Trent Waddington <[EMAIL PROTECTED]>
> Subject: Re: [agi] JAGI submission
> To: agi@v2.listbox.com
> Date: Tuesday, November 25, 2008, 4:25 PM
> On Wed, Nov 26, 2008 at 12:30 AM, Russell Wallace
> <[EMAIL PROTECTED]> wrote:
> > It certainly wasn't a strawman as of a couple of
> years ago; I've had
> > arguments with people who seemed to seriously believe
> in the
> > possibility of creating AI in a sealed box in
> someone's basement
> > without any feedback from the environment.
> >
> > If nobody believes that anymore, well, then that is
> significant progress.
> 
> One does not write scientific papers to address
> "arguments with
> people".  If someone has said, in print, that this is
> a good direction
> then getting them to nail down specifics is the next step..
> then
> addressing the feasibility is acceptable - well, sort of,
> these
> "negative" papers are never well received, as
> they don't make forward
> progress in the field, but if *everyone* is looking at
> closed box RSI
> and you feel the field isn't moving forward because of
> that, then it
> might make sense to publish a paper pointing out
> everyone's
> blindness.. but that's not that case.. not even close.
> 
> Trent
> 
> 
> -------------------------------------------
> agi
> Archives: https://www.listbox.com/member/archive/303/=now
> RSS Feed: https://www.listbox.com/member/archive/rss/303/
> Modify Your Subscription:
> https://www.listbox.com/member/?&;
> Powered by Listbox: http://www.listbox.com


-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=8660244&id_secret=120640061-aded06
Powered by Listbox: http://www.listbox.com

Reply via email to