Remember This

First a movie of the thing in action before we start talking details:

The idea behind this project was to create a machine that would have a memory that mimicked in some way how a human mind remembered things. From what I've gathered, when memories are formed the neural network that contains that memory is activated by a series of neurons creating a specific protein that is sent down the synapse either inhibiting or stimulating the neurons it is connected to. This resulting network is the actual memory. However, as with any sort of transcription, each time these proteins are formed, there are slight alterations which cause the resulting memory to shift.

The program I wrote takes a snapshot from a webcam and creates a memory of this scene via data-points which represent bright spots in the image. A rudimentary "memory" for sure, but a memory none-the-less. These data points are sent to my wall drawing machine, which goes through the process of moving from one data point to the next in the order the points where seen. Upon reaching the end of the data set, a new picture is taken from the webcam and this new set of points are compared to the memory set. Points are shifted via an inverse power law. In the pictures below you can see this process take shape in the debug screen of the program. Yellow dots are the memory set, large white dots are the current camera image and the small dots just show the x/y coordinates that are being sent to the machine.

The machine itself draws very slowly (the movie above is sped up about 75%). This is something that was done on purpose, as it forces the viewer of the project to either wait a very very long time to see massive change or they must leave and come back to see progress. In either instance, however, the viewer must deal with the memory of what they think they saw last.

There is also another factor that I enjoy about this project: the feedback loop of digital algorithm and analog output. At the heart of biology there are very algorithmic processes that dictate the interactions of the development of life. From our perspective, however, we see a shifting and complex world -- something that escapes in large ways this formalism that created it. For me there is something about this machine motoring around that seems child-like -- even with the algorithm that dictates its direction, the motors slip or the machine housing sways, and what gets created is anything but formulaic.