How Can Electricity Perform Computation?
How Can Electricity Perform Computation?
How Can Electricity Perform Computation?
Computation is the transformation of information according to rules. Clean definition. But it raises an immediate problem. Information is abstract. Rules are abstract. How does a physical machine made of silicon, conducting electricity actually execute them?
The Machine Beneath Everything
Open a spreadsheet. Type numbers. Apply a formula. A result appears almost instantly. Somewhere beneath that interface, information is being transformed. Computation is taking place. But there are no accountants inside the machine. At the lowest level, there is only electricity moving through circuits. So the question stands: how does electricity perform computation?
The answer begins with a device most people have never seen even though billions of them exist inside whatever device they're reading this on right now.
The Transistor
A transistor is a switch. Its entire job is to control whether electricity flows through a circuit. Electricity either flows or it doesn't. The transistor is either on or off. On equals 1. Off equals 0. That is the complete behaviour of a single transistor.
This seems too simple to matter. A light switch in your home does the same thing. What makes a transistor different? Scale and arrangement.
When Switches Start Enforcing Rules
A single transistor controls one flow of electricity. When transistors are connected together in specific configurations, they stop merely controlling electricity they start enforcing conditions. Electricity might flow only if two inputs are both active. Or if at least one is active. Or it might be deliberately inverted flowing when it previously wouldn't, stopping when it previously did. These configurations are called logic gates, and they are where electricity stops being a current and starts following rules. A logic gate is a physical rule. Transistors arranged into gates are computation made material.
From Gates to Everything
Logic gates combine into circuits. Circuits combine into processors. Processors execute programs. Programs build systems. At every layer, the underlying mechanism is identical: transistors switching on and off, enforcing conditions, producing outputs. The spreadsheet, the search engine, the AI that generates text all of it traces back to this foundation. There is no separate layer where something more mysterious is happening.
Why Transistors Changed the World
The transistor itself wasn't the breakthrough. Miniaturisation was. Once engineers learned to place millions then billions of transistors onto a single chip, computation became fast enough, cheap enough, and reliable enough to transform society. The internet required that scale. Smartphones required it. AI requires it. Modern AI is not a new kind of system. It is computation at a magnitude that was previously impossible made possible because transistors became small enough and numerous enough to support it. The AI revolution is not a philosophical breakthrough. It's an engineering one.