Original article can be found here (source): Artificial Intelligence on Medium
Pokemon and the Technological Organism – Earth Ship Control
^Now that is a graphic novel title if anything…
I wanted to expand on a talk I gave for the DGov council earlier year, where I explored the combination of AIs and DAOs in the governance of earth ships. We’ve seen how different “colours” of organisation have remarkably different behaviours depending on their internal structure. The 3rd characteristic of a teal organisation from the Wikipedia page, it reads:
“Evolutionary purpose” introduces a Teal organisation as a living organism with a direction of its own where its members are invited to listen and take note of the purpose it wants to serve.
This treatment of the organisation as a living entity that exercises its desires through us requires us to acknowledge both a form of life wider than ourselves, and an implicit purpose in our existing social systems. The multi-personal organisation lives between different physical, symbolic and economic spheres just as we do.
As a multi-cellular organism, at a certain scale I describe the action of many cells as my action; the impersonal information flow through my brain as “my thoughts”. In organisations, decisions are made on different levels of how to best respond to the daily inflows and outflows of life.
We’ve been speaking about the associated consciousnesses of complex systems. On each level — person, company, city, country — we can describe the entity as a single cohesive organism, with its own conscious will, and with the overall behaviour determined by its internal culture and motivations. It is wonderful to look to hive insects to see how simplistic behaviours at the individual level can create very complex emergent behaviours.
So what does this mean for the (organisms/)organisations we create? Let’s take the task of managing a largely autonomous earthship. It is relatively simple to imagine a timer for watering the plants, automated bills and grocery payments. We can set some rules to change these throughout the year and have a system that simply needs occasional tweaking. What I would like to see, however, is an entirely autonomous earthship, that regulates its own day/night cycle, can open every window, run a bath, monitor food growth and what it means for my groceries. So how do we do this?
I imagine raw sensor and actuator data connected to a learning algorithm, the “earthship pilot”. “Behaviours” here constitute finite or cyclical paths in the earthship phase space, which includes all possible states of the sensors and actuators. One path might be to move the windows from open to closed, and we can reward this behaviour at the appropriate times to reenforce it. This means that the earthship would go through an infantile phase where absolutely everything needs to be learned, and I’m imagining quirks and unexplained behaviours in the process. Perhaps there is an intermediate possibility, where we can set target parameters such as temperature, or even guide the program through a behaviour and have it learn it. Certainly to get to the science fiction possibility I want, there would need to be a common language that lets me describe what I would like in a way that the program understands, and some way for me to understand that sensor6 at 38 means that the food is cooked and it’s dinner time. As the earthship learns its own quirks, we arrive at an efficient method to manage environmentally symbiotic homes with little management.
Function aside, I would like my house to have a personality that I love as much as a tamagotchi. At 6pm when the water from the solar pipes moves into the house-and-hot-tub circuit, I receive a text letting me know it is time to go home and I am greeted with hot water for tea. My house assistant’s voice and face are unique, as generated by a genetic algorithm and preserved in its uniqueness on a blockchain. This identity also serves as its financial arm, and bills are handled directly by the house. While generated by an open-source algorithm, I picture house assistants being effectively bred or cloned and traded in much the same way as CryptoKitties permitted. We can infuse our houses with the same sense of soul which we know our pets to have. I want to see my house dance along to the music I play, and to remind me when I’ve gone too long without an evening off, or jamming with friends.
What happens when we have a network of these? When our houses can leverage data from the neighbourhood and inform us of anything that needs our intention. It begins to sound very similar to Nature 2.0 as proposed by Trent McConaghy which I link again here. When houses are financially sovereign entities, 2 entities can cooperate together to buy land to grow additional food for their occupants. If we are able to accurately marry our desires to those of our pilots, then we can allow for some very unconventional solutions to grow.
This is especially relevant for me as I think of how I would like to design the earthship I’m building in Guatemala. If we remove all historic preconceptions about how we should live, then what is the new optimum? I’m currently reaching out to people living in similarly unconventional arrangements to hear about their insights, please get in touch! For those interested, you can find out about the project here: knowhere.live
Much love folks! We’ll now return to our regular “Pokemon and the Technological Organism” programming.