entry: it can help to visualize concepts on a 2d plane and try to find solutions there
entry: if you get stuck, try to change the representation of the concepts
entry: if you reason a lot in a specific way, you can speed things up by
building a more specific, procedural way that worries less about alternatives
entry: if something is sometimes false, you need to find out when
this generally happens and alert units that use this information
or disable it from being used.
entry
if you think about non-useful things, you waste (cpu) time
entry
if nothing changes or gets added to the system, you cant learn things
entry
all useful units should get some cpu time to process things
entry
units that have something important to add to the processing, should
be activated somehow so they can send their messages.
how do you know which areas will be important (you cant activate them all,
yet you dont know in advance)? (see copycat for a way to handle this)
entry
you can make sure all units get cpu time if you update them
randomly with an equal chance
entry: dont waste time on unimportant things
entry: Detector - you will want to run small, inexpensive units to detect which larger units
are useful; they will give little results, but they will enable larger units
entry: units are small if they dont require a lot of cpu time
entry: units are small if they dont have much code in them
entry: if you have goals, you should probably solve them
entry: entries are usually a string of characters
entry: you can use "critics" to looks for a certain condition and select an appropriate action
entry: you can make critics that look for a certain pattern in
the current situation and a story (see: entries-stories). And try to adapt the
story so it matches the current situation and propose actions similar to the story.
For example: cause-effect patterns.
