For this project we set out to explore the subject of artificial intelligence and how we humans domesticate that intelligence. How can we go beyond something simply being connected or more effective? What does being smart actually mean? Within this mindset our team decided to explore these ideas within the area of politics and governance, and what the world might look like if there was such a thing as a ruling AI.
Course: Domesticating Intelligence
Team: Me, Luuk Rombouts, Lars Kaltenbach and Bjørn Karmann
Governance decisions are made on our account everyday. We elect representatives that make these decisions for us. We vote for them, based on programs that may or may not be respected once they reach power. A lot of us feel disconnected from the decisions being made. Sometimes we even stop paying attention to everything that's happening and our participation becomes very limited, either by choice or by the system itself.
With a future where the government is ruled by AI, would people be more engaged or even more disconnected?
On one hand, as confidence in the system is gained and knowing that all decisions are made for us by this smart entity, will we still pay attention to the decisions being made? Or will we only notice them once they affect us directly and immediately? On the other hand, this new system allows us direct access to information about everything that's happening and being decided. We can just ask, and information will be delivered in an individual and catered fashion. Will this make us more involved? More empowered?
- What if some of the choices we made on our everyday lives affected the way the government makes its decisions?
- What if politics became an even more mundane thing, invaded our lives and our homes - became domesticated?
- What would be the primary goals of this AI? And how do you translate qualitative goals into quantifiable data?
Based on current events, and the more predictable near future and scientific advancements, we constructed a fictional and somewhat funny timeline of future events that led to that status quo, and started raising and collecting different sets of questions we wanted to address and reflect upon.
We raised a lot of questions about this possible future, and started exploring different scenarios of interactions between people and this smart government.
What can be used as input to inform the ruling AI’s decisions?
How does it communicate with the general population?
How do we know what decisions are being made and how?
How transparent is this system?
How do people interact with it and with what purposes?
With this in mind we built a fictional system and 4 different objects that represent our attempts at exploring some of these questions.
Democracy is now mundane.
Voting is comprised of a set of everyday activities.
Reaction and readjustment are immediate.
1. Fine-tuning the President
As a first exploration we built an everyday object, resembling a kitchen appliance, that people could use to fine-tune the personality and priorities of the AI. Each knob had a different area of governance attached to it, and people can choose to assign it more or less importance in the grand scheme of things. A display shows a composite image of the President’s face, made up by different parts of past presidents faces, each part directly correlated to one of the knobs.
2. The Red Phone
A ruling AI can listen and reply to everyone, at any time and use that to inform its decisions. Re-using the idea of the red phone to call the president, we took it to the next level. The working prototype can listen to and record the citizens concerns, turn their speech into text and analyze it to find what entities were mentioned and determine the emotions present (we connected our processing sketch to the Watson API), and then display that information.
3. Everyday choices
In our future scenario we envision that the everyday choices we make could serve as input to inform the AI’s decisions. Because sometimes the unconscious decisions we make are as important or even more informative than what we say or articulate. We built a working prototype of a shelf, to be used in retail, that detects what product the person chose and once more uses this data as input to inform the AI’s decisions.
4. DISPLAYING DECISIONS: The pyramid
This distributed system of inputs communicates with a core, where decisions are made. But how can we visualize those decisions being made, and at such a fast pace?
A decision is never the product of a single input. As an experiment we built a pyramid of decisions that would show the decisions being made through several nodes, leaving behind a trace of the path that a decision went through. In this case the decision being made is represented by an augmented glass sphere that contains in itself the question being answered.
Behind the scenes