How many tries to give prior to getting frustrated and giving up?
couple issues
1. Giving Up
The entity should not "give up." Giving up implies that the
potential opportunity is written off, discarded and not to be considered
again. Why would one ever do that? Rather, let the opportunity linger
and resurface when conditions might have changed.
Nit picking? yes, but in my opinion, better to say "set aside."
2. Frustrated
If the AGI or entity has this emotion (I believe it should,) then
it is likely that the frustration will be the trigger for the
consideration to "set aside." Without the frustration, which would come
first, the intention might not be questioned. Don't give up or set
aside until after you are frustrated - frustration is the "evidence" you
detect that leads to reconsideration.
3. Set Aside
What gives the entity the ability to set aside? It is the
cognition that the recipe being used does not produce the results
expected. This creates a new opportunity - to come up with a better
recipe to produce the expected results. One might label this "creation
of a sub goal" but I find that less descriptive than to call it a new
opportunity.
4. probabilities
Matt states a fairly simple equation (utility * probability < cost)
that he calls the answer. I'm not sure it's as simple as indicated - in
fact I'm sure it is not. But he is right that there is a dependency on
estimates. A "model" is mentioned as the way to get an estimate of
probability. If we wait for this imaginary model to be developed, we
wait too long to make vital choices.
The alternative is to accept probability from any source that's
available. Better sources of probability increase ones "chance" of
making a good choice.
Consider that even the model of a physical process is built with
probability factors that are often "accepted" and not proven - in other
words, numbers that come from observation and not by a realistic
sub-model. (example, what's the probability that water will boil at 100 C)
...
On 09/25/2014 07:51 PM, Matt Mahoney via AGI wrote:
On Wed, Sep 24, 2014 at 11:10 PM, Piaget Modeler via AGI
<[email protected]> wrote:
How many tries should you, a child, or an AGI give to an intention before
getting frustrated
and giving up?
3, maybe. The better answer is when the utility of succeeding times
the probability of succeeding is less than the cost of trying. So it
depends on how your model estimates probabilities.
-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com
