The internet, which became accessible to a broader public in the early 1980s, has fundamentally transformed human communication and the business environment.
Its development traces back to research conducted in the 1960s and 1970s, particularly to military projects initiated by the Pentagon, the United States Department of Defense. The goal of these projects was to create a decentralised communication network, distributed across the entire territory of the United States, capable of maintaining communication even in the event of a nuclear attack by potential adversaries at the time, namely the Soviet bloc. Another key feature of the internet’s original design—stemming indirectly from decentralisation and wide geographic distribution—was its ability to organise and reconfigure itself automatically, without human intervention. This meant that if part of the network were destroyed, the remaining system could reroute connections and continue to function, even at reduced capacity, since there was no single central point whose destruction would bring down the entire network.
A helpful analogy for understanding how this network was conceived is the human nervous system. If certain neural connections are damaged, the body attempts to find alternative pathways to restore communication between key points, such as between the brain and the hand. Similarly, if part of the internet becomes nonfunctional, the network identifies alternative routes to ensure that information continues to flow. Of course, there are limits to how much such losses can be compensated, just as there are in the nervous system.
Transforming data into bits
Before describing in more detail how the internet operates, it is necessary to clarify a few basic concepts from the digital world. A fundamental term is “digital.” What does it mean when we say that a value is represented in binary or digital form?
In the physical world, most measurements are continuous, meaning they can take on an infinite range of values between two points. Take temperature, for example: if we say it varies between 10 and 30°C, it can assume any value within that interval, such as 12.3 or 27.9°C. With highly precise measuring instruments, we might record a temperature of 21.98°C. In fact, any value within that range—no matter how many decimal places it contains—is valid and usable.
Even within this limited range (10–30°C), there is an infinite number of possible values. This poses a challenge for modern computers, which operate using digital numerical systems and therefore cannot handle values with infinite precision. When we speak of this “inability to operate with infinitely precise values,” we mean that there is always a limit to how precisely a number can be represented when it is processed or stored in a computer, due to its digital nature. The greater the desired precision—meaning more decimal places—the more bits are required to store the number, which in turn increases the use of computing resources.[1]
The solution lies in the “discretisation” of measured values, that is, limiting the range of values a quantity can take. In the case of the earlier example, this means that the possible temperature values are restricted to a distinct subset within the 10–30°C interval. For example, we might limit the values to the following set: [10.0; 10.5; 11.0; 11.5; 12.0; 12.5 … 28.5; 29.0; 29.5; 30.0]. If the measured temperature falls between two of these values, it is rounded to the nearest one. In other words, the measurement will have a precision of 0.5°C, and any recorded value must be adjusted accordingly. A temperature of 27.9°C becomes 28°C, while 12.3°C is rounded to 12.5°C. This process represents the first step toward digitalisation.
The second step involves converting these values into binary format—that is, into a sequence of 0s and 1s. For example, if the measured temperature is 12.3°C, we first round it to 12.5°C. Then this value is converted into binary as follows:
- The value 12.5 is transformed into an integer, namely 125 (by multiplying by 10).
- Any integer can be represented in binary (that is, as a unique combination of 0s and 1s). For 125, the binary code is 01111101.
- This eight-digit binary code is called a byte. In the digital world, values are often represented as multiples of a byte, so any integer is encoded in binary as one or more bytes, depending on its size.
- All temperature values within the previously defined discrete range (10–30°C) are represented in this binary format, each decimal value having a unique corresponding binary code.
- When performing operations on these binary values, the initial multiplication by 10 is taken into account, and the results are adjusted accordingly.
What is an algorithm?
Another foundational concept of the digital age is the algorithm. An algorithm is a finite set of clearly defined instructions or steps that must be followed in a specific order to solve a particular class of problems. For instance, in the earlier discussion on digitisation, the second stage outlined five steps for converting continuous temperature values (such as 12.98°C) into their equivalent discrete binary representation. Following this sequence of steps in the prescribed order (simplified in the example above) leads to the transformation of a decimal value into its binary equivalent. The class of problems addressed here is precisely this type of mathematical conversion, which applies not only to temperature but to any similar numerical value.
What is a protocol?
If an algorithm formalises and describes, step by step, the procedure required to solve a given type of problem, a protocol can be understood as a sequence of message exchanges between two or more technological entities. Protocols are also present in everyday life, particularly in routine human interactions, and this analogy offers a useful starting point for understanding them.
Consider two people meeting on the street. One of them wants to find out the current time from the other. How would this unfold? According to common rules of politeness, the first step is to greet the other person with an appropriate message, such as “Good afternoon.” The second person will typically respond in kind: “Good afternoon.” Once this initial contact has been established, the first person proceeds with the actual request: “Could you please tell me the time?” The other replies: “It’s 11:30.” Afterward, the first person expresses gratitude (“Thank you”), and the second may respond, “You’re welcome.” With this sequence of exchanges, the interaction concludes.
This brief dialogue captures the essential elements of a protocol.
It can be observed that, in human interaction, a protocol involves specific messages being sent (“Good afternoon”) and specific actions taken in response to received messages (replying with “Good afternoon” as well). We also respond in defined ways to other events—for example, if no reply is received within a certain time frame, communication is typically abandoned.
The messages sent and received, the actions performed when these messages are exchanged, and other events that may occur (such as delays in responses) are all key elements of a human protocol. If individuals operate using different protocols—for instance, if one is polite and the other is not, or if one understands the concept of time while the other does not—then their protocols cannot interact, and effective communication cannot be established. The same principle applies to the protocols used on the internet: all entities involved in communication must use the same protocol in order to accomplish a shared task.
In summary, a protocol can be defined as the set of rules that determine the format and order of messages exchanged between two or more communicating entities, as well as the actions taken when those messages are sent, received, or when other relevant events occur.
Building on this brief introduction, what, then, is the internet?
What is the internet?
The internet is an electronic network designed for data transfer, interconnecting hundreds of millions of computing devices worldwide. Initially, these systems consisted primarily of traditional personal computers (PCs), but in recent years technological advances have significantly expanded the range of electronic devices connected to the internet.
To name just a few widely used categories of connected systems, we can mention mobile phones, televisions, gaming consoles, alarm systems, and various types of sensors. All these heterogeneous devices are end systems, meaning they are connected to the internet. The transport of data between these peripheral devices is handled by the internet’s core network—composed of millions of connections via cable, fiber optics, or radio—as well as a similarly large number of electronic data switches (routers).
In traditional communication systems, such as landline telephony, there is a central switching node (the telephone exchange), to which all subscribers are directly connected through physical cables. When a phone call is made, the exchange’s role is to directly connect the caller’s line with the physical line of the recipient. In other words, a dedicated circuit is established through wires from the caller to the recipient. Once the call ends, the circuit is disconnected, and both parties are free to initiate new calls to other destinations, with the connection process repeating for each new conversation.
In contrast, the internet operates using what is known as packet switching. The information transmitted over the internet is stored in digital form—that is, as a sequence of binary values (0s and 1s: …00011010011110100010…). These binary sequences can be quite long, sometimes consisting of millions of values. To enable transmission from source to destination, the data is broken down at the source into smaller subsequences (typically thousands of bits long). Each of these subsequences constitutes a data packet. For example, the sequence 00011010011110100010 can be divided into five packets of four bits each: 0001–1010–0111–1010–0010. Within the internet, there is no direct physical connection (through wires) between the source and the destination, as is the case with landline telephony. Moreover, when the source of the data and the intended destination are geographically far apart, there are often multiple alternative routes through which the two entities can be connected. As a result, internet communication relies on a logical connection between end systems, without a direct physical circuit linking them. This type of connection is known as a virtual circuit. Data packets propagate step by step, moving from one intermediate node (data switch/router) to another, along one of the possible paths that virtually connect the source to the destination.
Packet-switched networks bear many conceptual similarities to a road transport system made up of national roads, highways, and intersections. Consider, for example, a factory that needs to deliver a large piece of machinery to a recipient located in a distant city. At the factory, the machine is dismantled into smaller components that can be loaded onto trucks, which are then dispatched toward the destination. These trucks travel independently through the network of roads and highways until they arrive. At the destination, the machine must be reassembled into its original form—a process that can only be completed once all the components have been received and properly reordered. In this analogy, the packets are the trucks, the communication links are the roads and highways, the routers correspond to intersections, and the source and destination are the factory and the recipient, respectively.
What is artificial intelligence?
Artificial intelligence (AI) refers to the intelligence of computers or the software that runs on them, as distinct from human or animal intelligence. Examples of AI applications include advanced internet search engines (such as Google Search), content recommendation systems based on user profiles and past activity (like those used by YouTube or Netflix), and autonomous navigation systems in vehicles. More recent generations of AI also include platforms such as ChatGPT or AI Art systems, which are generative in nature—producing content that did not exist prior to a user’s request. Although the theoretical foundations of AI date back to the 1960s or even earlier, it is only in recent years that advances in computing power have made it possible to fully harness the potential of existing algorithms.
Machine learning is a branch of artificial intelligence that enables computers to “learn” and improve without being explicitly programmed to do so. The core idea is that, starting from a training dataset, a machine learning system can generalise and abstract what it has learned by extracting essential features shared by all elements in that dataset. Consider a program trained on thousands of images of cats, learning what makes an image a “cat.” After training, the program can recognise cats in new images it has not previously encountered. In this way, machine learning is used to classify or organise information, such as identifying cats in photographs.
Deep learning is a subset of machine learning that focuses on identifying patterns within datasets. It relies on a system of layered information processing, where each layer is specialised in recognising specific features. For example, in identifying images of cats, one layer may detect the shape of the ears, while another may recognise colours. A deep learning system can also improve its performance by adjusting its internal parameters based on previous errors. In addition, it can discover new patterns or features in the data it processes, even if it was not explicitly programmed to recognise them. Thus, a deep learning system trained to recognise human faces might also learn, for example, to distinguish between male and female faces, even if it was not specifically programmed to do so.
We live in a world in which we are constantly interacting with the products of the digital revolution. Understanding the foundations of this digital environment—how the internet and artificial intelligence function—is no longer a luxury, but a necessity.
It is not enough to know how to use these tools effectively; we also need a clearer perspective on their impact on society and on our way of life. By uncovering what lies behind screens and lines of code, we expand our knowledge and, more importantly, become informed, active participants in a rapidly evolving digital age.
Radu Badea aims to open a window—if only slightly—onto the world of modern technologies, as a first step toward a better understanding of their impact on our everyday lives.
