Away from intention and wet networks

<Naomi>You can only draw from the experience you have already.</Naomi>
<Cesar>Well, this is a very safe method of acquisition of knowledge.
Let’s describe knowledge as a territory, in order to extend it, you might not build
upon what you already know, but you will rather jump out of it, as far as possible,
to eventually come back to what you know with “little steps” (Teodora), from the
unknown back into the known. This hypothesis of knowledge acquisition is being
verified in the neurosciences, in the pattern of knowledge propagation in neural
network.
</Cesar>

<Teodora>So if you have this (neuron A), you can have this (neuron B). But the
question is : would you want to jump, once you have created some stable (net-
work) here (pointing neuron A) and it probably varies from person to person :
some people want to go like that (small synapses drawn as dashes from neuron A
to neuron B) small steps and then get here (pointing out neuron B); some peo-
ple would sample (creating many new neurons C, D, E, F, G etc) starting from
everywhere and then group them together. In infancy you would se more this kind
of model, where you start with many points at the same time, this is only the way
they can advance because they have to learn so many things. But we do simplify
things that’s the only way. The ultimate goal is to understand how we make deci-
sions, how we interpret the world, but it is impossible to study the world! You need
to start somewhere, this is why I study categorization, because categorization will
never happen on its own in the real world.</Teodora>

<Cesar>An Open Architecture often uses this method : multiplying the possibili-
ties (anchor points) and the connexions before they start making sense, a sense
that is made in the process, the architecture is the process. A simpler definition
emerges for a set of complex ideas and relationships.</Cesar>