Automation’s Deeper Meaning

Automation is an ancient technology for putting things in motion. As humans, we tend to think of automation as just something that replaces human labor. But to prepare for what lies ahead, we need to see automation as something far bigger.

At a basic level, you can think of automation as “goal-oriented, code-driven work.” Biology, from this lens, is the original version of automation. Its DNA coding embeds the goals of adaptation and reproduction into carbon-based life forms. That bird outside your window is fully automated. So are the trees you see when you walk down the street. They grow and heal and reproduce without any help from you. And when you really tune into the magnificence of that intelligence, you catch an inkling of the deeper meaning of automation. Life buzzes, groans, creaks, and whirs in an amazing cacophony of goal-oriented, code-driven work.

Why should we bother expanding our view of automation in this way? Because today’s machines are evolving into something approaching the uncanny autonomy of biology. We’re not there yet, but make no mistake—many of us will live to witness the rise of a new, automated intelligence well beyond our our current understanding of automation. It will transcend the human understanding of “work.”

Automation is intimately intertwined with intelligence. That was true in the evolution of biological automation and the genetic intelligence that gave rise to it. It is also true with machine automation and the artificial intelligence that is now accelerating it.

To really get where automation is headed means fully appreciating the growing synergy automation and machine learning. Here’s a simple yet powerful way to understand this connection. Imagine, a sensor that enables the system to take in data from the outside world. That data then generates an abstracted model to help the system understand the external world. That model then drives an actuator that sets in motion some automated activity. In short, the sensor feeds data into a machine learning model that then drives the system’s automation.

This integrated system for sensing, learning, and taking action largely mimics the biological automation that preceded. But these systems scale in important new ways. As the quantity of data increases, the fidelity of the machine learning model improves. As the model improves, the effectiveness and efficiency of the automation improves. That, in turn, increases the attractiveness of the system to end users, which attracts more usage and data and leads to better models and automation. The result is a virtuous cycle and an upward spiral in intelligence and performance—a fusion of automation and machine learning.

Despite the cosmic scale of changes now underway, humans still play a vital role in automation and we do so for two reasons. The first is that we were the ones that set generation of automation in motion. Our designs and engineering gave rise to its underlying technologies and our goals guide its ongoing operations. One of the central questions I explore on this site is whether human governance over these systems will hold as they increasingly learn to design themselves.

The second connection between humanity and automation is that our subjective experience gives them meaning. While biological automation pulsates within biological ecosystems, machine automation whirs within economic systems. In market-based economies, economic “utility” determines survival. That utility is set ultimately by end-users. Even in command economies where economic decisions are controlled by the state and selection pressure consolidates into a few hands, subjective human judgment still determines the fitness of automated systems. This is a critical and often underappreciated point. Barring the emergence of some form of artificial machine subjectivity, humans control the feedback through which these systems learn and grow. Herein lies the simultaneous promise and peril of automation.

Automation leaves a distinct footprint within each economic sector, from resources extraction to manufacturing, services and government spending.

Automation’s greatest impact on Earth’s ecosystems happens through resource extraction industries. As agriculture, fishing, and forestry automate, they accelerate extraction to increasingly unsustainable levels that risk widespread ecological collapse. Resource extraction systems feed the manufacturing, services, and government sectors where poorly designed processes lead to accelerating waste in the form of plastic islands in the Pacific and the blanket of greenhouse gasses now enveloping our planet.

If the automation of resource extraction impacts the planet and the automation of manufacturing impacts machines, it is the automation of services economy that most impacts humanity. Service industries largely depend on contact with end users to generate economic value. Automating services thus primarily means automating interactions with end users through automated self-service and automated service platforms like Google, Facebook, and Amazon. These platforms operate at “Internet scale,” engaging billions of people around the world in their work, which is why these systems are evolving into humanity’s primary interface with machines.

Governments play a variety of roles in connecting us with automation and artificial intelligence. They can deploy automation to improve the flow of traffic through city streets and the flow of energy into our homes, but they can also use it to surveil citizens and even deploy lethal force against them with automated drones. The rise of authoritarianism around the globe is now making the fusion of automation and artificial intelligence particularly scary within the government sector.