Hi. Welcome to this
final video of Week 3. We are going to conclude a conversation of our
ecosystem simulation, really looking at how could
we expand this ecosystem? As you could imagine, some
of the initial ideas of this ecosystem suggested
that we would actually create also a carnivore
class or a predator class, and we would have a whole
series of behaviors. Well, we're going to
leave it here and really allow you to continue expanding on this
ecosystem project, because it really it's outside the scope of this
class to teach, basically repeat the same
concepts that we've been using to just expand
it in the ecosystem. I think it's a
fantastic exercise to really go through and continue in the projects
that we start here. Let's discuss what
you could be doing. The first thing as
we've just discussed is introduce a new
class to the ecosystem. You could have more species. We've called Wendy herbivore. But we could actually
start including a carnivore or a predator. One that might not be
feeding on our food, which is vegetable, but might
be feeding on herbivores. That would yield a
chase relationship, a sense of a different kind of population control
of the herbivores through them becoming the food supply for
a different species. Again, you could have
a lot of circularity of predator prey relationships. You could also start
thinking that there's multiple types of food, maybe different species of food. Some of them that grows slower, some of them that grow faster. The food system
that we currently have if you start
thinking of them as vegetables or trees or
plants of different types, you could actually speciate
that system altogether. Again, different ratios of
growth just the way we've been doing it and
maybe different kind of caloric interests. Maybe certain animals prefer certain vegetables or
certain food supply. There's a lot to
do in this domain. You could also start expanding
the simulation of how do we introduce the food propagation or
plant propagation? How you could start thinking
of a system of pollination where maybe a new food entry, maybe within a
certain time frame of the first one appearing, it would actually
start flowering and perhaps spreading seeds. Then there would be new plants around a cluster of plants. Those would have their own
kind of growth patterns. That might be quite interesting because you're not
forced to think that you are seeding
the system with food, but rather there's
a logic of growth. That could be done.
It's a little bit of an idea of recursion where you want a food entry can
create another one over time, and then that one
would actually create another one over time. The simulation would
actually grow on its own. You have to be careful here, because if you do
it too quickly, you will end up with
a lot of plants. But again, we could
also do the same thing or a principle of reproduction, where two herbivores
or two kind of species of animals can
actually create an offspring. Instead of mechanically as we are doing currently creating a new
herbivore in the world, that new herbivore only
happens when well fed, maybe two well fed
herbivores have a state of mating and
actually create an offspring. I love to think about how
would the decision tree of our animals would
look like perhaps. This is again a very
small increment to the decision
tree that we have. We have and maybe am I awake? Am I resting? Am I being chased? This could be a priority. It's like if you're being
chased by a predator, perhaps you have a flee state. Its survival logic kicks in
first before anything else. If I'm being chased,
the first thing I want to be doing is probably fleeing the animal
that is chasing me. Then we could ask the
question, are you hungry? Well, perhaps it's
time for seeking food, but if none of those
are actually present, maybe there's a
reproductive level or reproductive urge level that suggests that we could
either wonder or mate, mating and finding a
partner and so on. All these ideas that we have discussed in terms
of adding new species, discussing forms of reproduction of a species within
the ecosystem, those could also be fitted into a slightly more complex decision tree as
you can see here. We're not going to be writing
the code for all these. The invitation is that
you use the knowledge of this week to continue
expanding on these ideas, practice and really see if you
can take it a bit further. But basically going through the same logic that
we've written so far. It would be a lot
of repetition to really build up some
of these ideas. With this in mind, the next week we're going to see
brand new projects. We have new concepts that we really want to introduce to you. I'll see you in
next week's videos where we're going
to start looking at path finding.
I'll see you then.