Author: lampsauce

lampsauce-Reading01

Kate Compton’s 10,000 Bowls of Oatmeal Problem describes the potential for a generative algorithm to produce content or artifacts that are too similar for humans. The resulting mundaneness makes any output of a generator look too bland (ie. like oatmeal). This may be a good thing if you need to generate something like trees; each should be slightly unique, but 10,000 trees together becomes a forrest. If a given tree stands out too much, the forrest will not look right.

However, the problem of perceptual uniqueness arises when an algorithm’s output lack’s diversity. More precisely,  consider a car generator. Cars are deeply personal items, with drivers often decorating them with bumper stickers, dashboard dolls, etc. It is difficult to explain why, but a human can tell if they’ve seen the same car drive by twice.

Like Kate Compton writes, considering dealbreakers of artifacts at scale is one way to try to combat a lack of perceptual uniqueness. Another strategy is to categorize artifacts into types/tiles (ie. SUVs, German, electric, diesel, sports, etc.) From a more artistic point of view, adding certain defects (ie. dust, rust, scratches, peeling paint, etc) to an artifact  may make it more perceptually unique.

lampsauce-Map

My imaginary world, much like our own planet, has several land masses over a vast ocean. It also depicts several cities with varying populations. The ocean level varies with each generation, allowing for both huge continents and small island clusters to occur.

CODE

Variations of Generator

Process

This map generator creates a 3D terrain using Perlin noise. The map’s scale and ocean level is chosen at random. The points are stratified by whether or not they are below the ocean level. The points below the ocean are colored by mapping each depth to a corresponding HSL color range. I wanted the land masses to look like a treasure map, so I mapped the land points to a specified color range.

The number of cities is directly proportional to the amount of land. I was curious what it would look like if I could raise or lower the ocean level of a given map. If you use the left/right arrow keys the ocean level will raise and lower. The cities & populations will adjust accordingly.

lampsauce-lookingoutwards01

Refik Anadol’s Melting Memories is an installation in Istanbul.

I admire the complexity and synchronicity of the work. The way that certain components melt into one another and reform or reappear later has a very aesthetic quality. I also admire the algorithms and dataset it took to generate this work. Anadol’s work often takes massive datasets and expresses them visually. One thing I admire about his works is how he uses data as a pigment, to express what is normally unseen.

Anadol partnered with the Neuroscape Laboratory at UCSF and collected EEG data from participants’ hippocampus region to create Melting Memories. Anadol’s creative team (comprised of AI and CS experts) developed a neural network to convert the EEG data into procedural noise which could be visually interpolated. Given the uniqueness of this dataset, it is likely that Melting Memories was created using custom software.

The work was inspired by Anadol’s uncle’s getting Alzheimers which making him consider the computational foundation of memory. Future works that may take inspiration from Anadol’s work may make clever use of data from different kinds of signals of the body.