Computer

A computer is a machine that carries out instructions. These instructions are very simple: add two numbers, compare two values, remember a result. The computer doesn’t execute them particularly cleverly, but incredibly fast and without errors. A modern processor manages several billion such steps per second. Everything a device does – displaying text, playing music, generating an answer from an artificial intelligence – ultimately consists of such small steps. The decisive point: which instructions are carried out is interchangeable. That’s why a computer is not fixed to a single task, unlike a washing machine or a pocket calculator.

The machine that replaced everything else

Before the computer, every task needed its own device. A typewriter for letters, a record player for music, a calculating machine for invoices. A computer replaces all three without anything changing in its construction. You just swap out the instructions, in other words the program. Experts call this property universality.

This universality explains why computers are everywhere today. A modern car contains dozens of small computers. They control the engine, the brakes, and the screen on the dashboard. A washing machine, a traffic light, and a wristwatch now also contain one. The device on the desk is only the most visible form.

For artificial intelligence, the computer is the basic prerequisite. An AI model is not an independent being but a very large collection of numbers and computational rules. Without machines that carry out these calculations millions of times, there would be no AI. That is precisely why companies that manufacture computer chips have become so valuable on the stock market.

Processor, memory, and the path of data

Every computer consists of three basic building blocks. The processor calculates, the working memory holds data ready in the short term, and the hard drive preserves it permanently. Added to this are input and output devices such as keyboard and screen. A good comparison is a desk: the processor is the person working, the working memory is the desktop surface, the hard drive is the filing cabinet in the next room.

The processor works in a fixed loop. It fetches an instruction from memory, executes it, and stores the result. Then it starts over. For this to work, everything is represented as a number: letters, the colors of a pixel, the volume levels of an audio recording. Internally, these numbers consist only of zeros and ones, because two states can be distinguished most reliably in a circuit.

A common misconception is that a computer understands what it does. It doesn’t. It doesn’t distinguish between a love letter and a tax return, because to it both are just sequences of numbers. Meaning only arises with the human who reads the result. This remains true at the core even with AI programs, even if the answers often seem astonishingly sensible.

From smartphone to data center

The computer that most people use most often is their smartphone. Technically, it is a fully-fledged computer with a processor, memory, and screen. Game consoles and smart TVs also belong to this category. Anyone who reads about chips or semiconductors in the news is reading about the components in exactly these devices.

At the other end of the scale are data centers. These are halls with thousands of computers working for companies like Google or OpenAI. When you use a chatbot, it’s not your device that computes the answer. Your question is sent over the internet to such a data center, and the answer is sent back. The electricity consumption of these facilities has since become a political topic in its own right.

In economic news, the term usually appears indirectly. It’s about supply bottlenecks for chips, graphics cards for AI training, or computing power as a scarce resource. Behind this is always the same question: who has access to enough computing capacity? Anyone who understands computers for what they are – fast but dim-witted instruction processors – can make much better sense of such reports.