Consider Google GO as a possible implementation language.
------------------------------------------------------------------------------------------------------------------------------------------------
> That leads on to the next question: how do we allocate CPU time? The
> recursive search structure offers a starting point: each argmax node
> can be a process, continually searching for better answers to send to
> its callers. Of course there will be too many of these to make each
> one a process at the operating system level let alone give each its
> own CPU core, so now we need to decide how to allocate a small number
> of CPU cores to many candidate processes. In the worst case this could
> be regarded as a problem requiring general computation in its own
> right, but our list of open problems is tediously long as it stands;
> is there anything more definite we can say about this one?
>
> Perhaps there is. There are general arguments that the expected
> utility of producing some result should be (something isomorphic to) a
> real number. Further, we can argue that if node A calls B as a
> subgoal, the utility of B can't exceed that of A. (If a node has
> multiple callers, its utility is bounded by the sum of caller
> utilities.)
>
>
> Comments so far?
>
>
-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com
