Every time they die they remember how they died and do less of that. When they get a kill they start doing more of what got them that kill. with every time they get a kill their memory also expands so they "mature" over time
if there are enemy's at the end of it and you gave them long enough they would learn to traverse it as long as there is a reward in place and the time to learn
unfortunately not, instead of being some badass malware attacker or blackhat hacker, or even working a stable job in software or web development, i am just a broke ass gamedev trying my luck at becoming the next big time success
interesting game. This is not quite a genetic algorithm simulator, but it's close. I'm sure you must have heard of GAs. What was the impetus to move away from them for this kind of game and come up with this more classical approach?
