The digital revolution: computing and the internet

Intuition Beginner

The digital revolution is the most recent and perhaps the most far-reaching transformation in the history of technology. In less than a century, computers went from theoretical concepts and room-sized machines to pocket-sized devices more powerful than the computers that guided the Apollo missions to the Moon. The internet, which barely existed before 1990, now connects over five billion people and has reshaped commerce, communication, politics, and culture worldwide.

The story begins with mathematics, not engineering. In 1936, Alan Turing (1912-1954) published a paper titled "On Computable Numbers, with an Application to the Entscheidungsproblem" (decision problem). Turing was trying to answer a question posed by David Hilbert: is there a general procedure that can determine whether any given mathematical statement is provable? Turing showed that the answer is no. To prove this, he invented a theoretical device now called a Turing machine — an abstract machine that reads and writes symbols on an infinite tape according to a set of rules.

The Turing machine was not meant to be built. It was a thought experiment designed to clarify what it means for a function to be "computable" — that is, capable of being calculated by a mechanical procedure. Turing showed that there exist well-defined mathematical functions that no Turing machine can compute, which means they are not computable by any mechanical procedure whatsoever. This result — the existence of uncomputable functions — has profound implications for the limits of computation and the limits of what machines can do.

But the concept of a Turing machine also suggested something positive: a single machine that could compute anything that is computable. This idea — a "universal" computing machine — was the theoretical ancestor of the stored-program computer. Turing showed that a single machine, given different instructions, could perform any computation. This was a radical idea in 1936. Before Turing, computing machines were designed for specific tasks (calculating artillery tables, encrypting messages, tabulating census data). Turing's insight was that a general-purpose machine could replace all special-purpose machines.

The first electronic computers were built during and immediately after World War II. The Colossus, built in Britain in 1943-44 to decrypt German military communications, was the world's first programmable electronic computer. The ENIAC (Electronic Numerical Integrator and Computer), built at the University of Pennsylvania in 1945, was the first general-purpose electronic computer. It weighed 30 tons, contained 17,468 vacuum tubes, and could perform 5,000 additions per second.

John von Neumann (1903-1957) made a crucial architectural contribution: the stored-program concept. In von Neumann's design, the computer's program (instructions) is stored in the same memory as the data it operates on. This means the computer can modify its own program during execution, enabling conditional branching, loops, and other programming constructs that make complex computation possible. The von Neumann architecture — a central processing unit (CPU), memory, input, and output — remains the basis of most computers today.

The transistor, invented at Bell Laboratories in 1947 by John Bardeen, Walter Brattain, and William Shockley, was the key hardware innovation. Transistors replaced vacuum tubes as the switching elements in computers. They were smaller, more reliable, less power-hungry, and cheaper to manufacture. The invention of the integrated circuit (Jack Kilby at Texas Instruments and Robert Noyce at Fairchild Semiconductor, both in 1958) made it possible to place multiple transistors on a single chip, beginning the trend toward miniaturization that has continued for over six decades.

Gordon Moore, co-founder of Intel, observed in 1965 that the number of transistors on a chip was doubling approximately every two years. This observation, known as Moore's Law, has held remarkably well for over fifty years and has driven the exponential increase in computing power. A modern smartphone contains billions of transistors, each smaller than a virus.

The development of the internet has its own trajectory. ARPANET, the precursor to the internet, was created by the U.S. Department of Defense's Advanced Research Projects Agency (ARPA) in 1969. ARPANET was designed to connect research institutions and allow them to share computing resources. It used packet switching — breaking data into small packets that are routed independently across the network and reassembled at the destination — rather than the circuit switching used by telephone networks.

The concept of packet switching was developed independently by Paul Baran at RAND Corporation and Donald Davies at the National Physical Laboratory in Britain, both in the early 1960s, illustrating how multiple researchers can converge on the same innovation when the technical conditions are ripe.

The TCP/IP protocol suite, developed by Vint Cerf and Bob Kahn in 1974, provided a common language for connecting different networks. The adoption of TCP/IP as the standard for ARPANET in 1983 created the internet — a network of networks using a common protocol. The World Wide Web, invented by Tim Berners-Lee at CERN in 1989, provided a user-friendly interface for accessing information on the internet through hyperlinked documents viewed in a browser. The combination of the internet (the infrastructure) and the Web (the application layer) transformed a research tool into a global communication platform.

The personal computer revolution of the 1970s-1980s, led by Apple (founded 1976), IBM (which entered the personal computer market in 1981), and Microsoft (which provided the operating system for IBM-compatible PCs), put computing power in the hands of individuals. The smartphone revolution, initiated by Apple's iPhone in 2007, put internet-connected computing in the pockets of billions of people.

The social consequences of the digital revolution are still unfolding. The internet has democratized access to information, enabled new forms of commerce and communication, and created platforms for social and political organization. It has also enabled mass surveillance, the spread of misinformation, the erosion of privacy, and new forms of economic inequality. Understanding the digital revolution is essential for navigating the present and anticipating the future.

Visual Beginner

Era	Period	Key technology	Impact
Theory	1936	Turing machine	Defined computability; universal machine concept
First computers	1943-1955	Vacuum tubes (Colossus, ENIAC, UNIVAC)	Military and scientific computing
Transistor age	1947-1965	Transistors, early integrated circuits	Minicomputers; business computing
Integrated circuit	1965-1980	Moore's Law, microprocessors	Personal computers become possible
PC revolution	1975-1995	Apple II, IBM PC, Macintosh, Windows	Computing for everyone
Internet age	1990-2007	World Wide Web, broadband, search	Global information access
Mobile era	2007-present	Smartphones, cloud computing, AI	Ubiquitous computing

Worked example Beginner

To understand the power of Moore's Law, consider the following calculation. In 1971, the Intel 4004 microprocessor contained 2,300 transistors and cost about $60. If transistor density doubles every two years, how many transistors would you expect on a similarly-priced chip in 2023 — 52 years later?

Number of doubling periods: $52/2=26$

Expected transistor count: $2,300\times 2^{26}=2,300\times 67,108,864\approx 154$ billion transistors

The actual Apple M1 Ultra chip (2022) contains 114 billion transistors. The Nvidia H100 GPU (2023) contains 80 billion. The numbers are in the same ballpark, confirming that Moore's Law has held remarkably well over five decades.

This exponential growth has consequences that are difficult for humans to grasp. We think in linear terms: if something grew by 10% last year, we expect it to grow by about 10% this year. But exponential growth compounds. A technology that doubles every two years grows by a factor of a thousand in twenty years and a factor of a million in forty years. This is why computers have gone from room-sized machines that cost millions of dollars to pocket-sized devices that cost hundreds of dollars and are millions of times more powerful.

The end of Moore's Law has been predicted many times but has not yet arrived, though the rate of doubling has slowed somewhat. Physical limits — transistors cannot be smaller than individual atoms — will eventually halt the trend. The search for alternative computing paradigms (quantum computing, neuromorphic computing, optical computing) is motivated partly by the recognition that the current approach cannot continue indefinitely.

Check your understanding Beginner

Formal definition Intermediate+

A Turing machine is formally defined as a 7-tuple $M=(Q,\Sigma ,\Gamma ,\delta ,q_0,q_{\text{accept}},q_{\text{reject}})$ where:

• $Q$ is a finite set of states
• $\Sigma$ is the input alphabet (does not include the blank symbol)
• $\Gamma$ is the tape alphabet, where $\Sigma \subset \Gamma$ and the blank symbol $\bigsqcup \in \Gamma$
• $\delta :Q\times \Gamma \to Q\times \Gamma \times \{L,R\}$ is the transition function
• $q_0\in Q$ is the start state
• $q_{\text{accept}}\in Q$ is the accept state
• $q_{\text{reject}}\in Q$ is the reject state, where $q_{\text{reject}}\ne q_{\text{accept}}$

The machine operates on an infinite tape divided into cells, each containing a symbol from $\Gamma$. The machine has a read/write head positioned over one cell. At each step, the machine reads the symbol under the head, consults the transition function $\delta$ given the current state and the symbol read, writes a new symbol, moves the head left or right, and transitions to a new state. If the machine enters $q_{\text{accept}}$ or $q_{\text{reject}}$, it halts.

The Church-Turing thesis states that any function that can be computed by an effective procedure (an algorithm) can be computed by a Turing machine. This thesis is not a theorem — it cannot be proved, because "effective procedure" is an informal concept — but it is supported by the fact that every alternative model of computation that has been proposed (lambda calculus, recursive functions, register machines, etc.) has been shown to be equivalent to Turing machines.

The concept of a Turing machine also enables the classification of computational problems by their difficulty. A problem is decidable if there exists a Turing machine that halts on all inputs and correctly determines the answer. A problem is semi-decidable (or recursively enumerable) if there exists a Turing machine that halts and accepts when the answer is "yes" but may run forever when the answer is "no." The halting problem is semi-decidable but not decidable: one can simulate the machine $M$ on input $w$ and accept if it halts, but there is no general procedure for determining that it will never halt.

The Chomsky hierarchy provides a parallel classification of formal languages by their generative power, with each level corresponding to a class of automata. Regular languages (generated by regular expressions) are recognized by finite automata. Context-free languages are recognized by pushdown automata. Context-sensitive languages are recognized by linear-bounded automata. Recursively enumerable languages are recognized by Turing machines. This hierarchy connects the theory of computation to the practice of programming language design, compiler construction, and natural language processing.

A function $f$ is called Turing-computable if there exists a Turing machine that, given input $x$, halts with $f(x)$ on its tape. Turing proved that there exist functions that are not Turing-computable. The most famous example is the halting problem: given a description of a Turing machine $M$ and an input $w$, determine whether $M$ halts on input $w$. Turing showed that no Turing machine can solve this problem — there is no general algorithm for determining whether a program will eventually halt or run forever.

Key theorem with proof Intermediate+

Theorem (Unsolvability of the halting problem): There is no Turing machine that can determine, for an arbitrary Turing machine $M$ and input $w$, whether $M$ eventually halts on input $w$.

Proof (by contradiction):

Suppose there exists a Turing machine $H$ that solves the halting problem. $H$ takes as input a description of a Turing machine $⟨M⟩$ and an input string $w$, and outputs "halt" if $M$ halts on $w$, and "loops" if $M$ runs forever on $w$.

Construct a new Turing machine $D$ as follows:

$D$ takes as input $⟨M⟩$ (a description of a Turing machine $M$).

$D$ runs $H$ on input $(⟨M⟩,⟨M⟩)$.

If $H$ outputs "halt," then $D$ enters an infinite loop.

If $H$ outputs "loops," then $D$ halts.

Now, what happens when we run $D$ on its own description, $⟨D⟩$?

If $D$ halts on input $⟨D⟩$, then $H(⟨D⟩,⟨D⟩)$ outputs "halt," which causes $D$ to enter an infinite loop. Contradiction: $D$ both halts and does not halt.

If $D$ does not halt on input $⟨D⟩$, then $H(⟨D⟩,⟨D⟩)$ outputs "loops," which causes $D$ to halt. Contradiction again: $D$ both loops and halts.

Either way, we reach a contradiction. Therefore, the machine $H$ cannot exist, and the halting problem is unsolvable.

This proof is a variant of Cantor's diagonal argument and the liar paradox. It demonstrates a fundamental limitation of computation: there are well-defined questions that no algorithm can answer. The result has practical implications for software engineering (no general program verifier can determine whether an arbitrary program is correct) and theoretical implications for the philosophy of mind (if human minds can solve the halting problem, then minds are not equivalent to Turing machines — though this is a contested inference).

The halting problem has several important corollaries. Rice's theorem (1953) generalizes the result: any meaningful (non-empty and non-universal) semantic property of programs (does this program ever output 7? does it terminate in fewer than 100 steps? does it contain a buffer overflow?) is undecidable. This means there is no general algorithm that can automatically verify arbitrary properties of programs, which has profound implications for software verification and cybersecurity.

The relationship between computability and complexity is also significant. While the halting problem shows that some problems are undecidable (no algorithm can solve them), complexity theory classifies decidable problems by their difficulty. The P vs. NP problem — whether every problem whose solution can be verified in polynomial time can also be solved in polynomial time — is one of the seven Millennium Prize Problems and carries a $1 million reward from the Clay Mathematics Institute. Most computer scientists believe P does not equal NP, meaning there are problems that are inherently difficult to solve but easy to verify, but this has not been proved. The practical implications are enormous: many cryptographic systems, optimization problems, and logistical challenges depend on the assumption that certain problems are computationally hard.

Exercises Intermediate+

Advanced results Master

The digital revolution has been shaped by a complex interplay of technological innovation, economic forces, government policy, and cultural change. Understanding these interactions is essential for understanding both the history of computing and its future trajectory.

The role of government funding in the development of computing is often underappreciated. The first electronic computers were funded by military needs: Colossus for codebreaking, ENIAC for artillery table calculation. ARPANET was funded by the Defense Department. The mouse, windows, and graphical user interface were developed at SRI and Xerox PARC with government funding. The early internet backbone was funded by the National Science Foundation. The Global Positioning System (GPS) is a military system made available for civilian use. The pattern is consistent: governments fund high-risk, long-term research that the private sector cannot justify, and the resulting technologies are then commercialized by private companies.

This pattern has been called the "entrepreneurial state" by the economist Mariana Mazzucato. She argues that the most innovative economies are those where the government actively shapes markets through strategic investment, rather than merely creating favorable conditions for private investment. The development of the internet is a prime example: every major technology in the modern internet — from packet switching to TCP/IP to the Web browser to search engines to touchscreen technology — had significant government funding in its early stages.

The relationship between computing and warfare has been a recurring theme. World War II drove the development of electronic computing (Colossus, ENIAC) and operations research. The Cold War drove the development of ARPANET (designed to survive nuclear attack), GPS (developed for military navigation), and much of the early semiconductor industry (funded by military procurement). The War on Terror has driven the development of surveillance technology and drone warfare. The emerging field of autonomous weapons raises new ethical questions about the role of artificial intelligence in warfare.

The gender dynamics of computing are a significant and often neglected aspect of its history. The earliest programmers were predominantly women — Ada Lovelace (often called the first programmer, though this claim is contested), the women who programmed ENIAC, and the women who worked as "human computers" at NASA and elsewhere. As programming became more prestigious and better-paid in the 1960s, the profession masculinized. The percentage of women in computer science peaked at approximately 37% in the mid-1980s and has declined since, reaching approximately 20% in recent years. The reasons for this decline are debated but likely include the marketing of personal computers to boys, the cultural association of computing with masculinity, and the workplace culture of the technology industry. The underrepresentation of women and minorities in computing has consequences for the technology that is produced: AI systems trained on data generated by a non-representative population may encode the biases of that population.

The development of artificial intelligence (AI) is the most recent chapter in the computing story, and it is still being written. The concept of artificial intelligence was formalized at a conference at Dartmouth College in 1956, organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon. The conference participants believed that human-level machine intelligence was achievable within a generation. They were wildly optimistic — the problem turned out to be far harder than anyone anticipated.

The history of AI has been characterized by cycles of enthusiasm and disappointment, sometimes called "AI summers" and "AI winters." Early successes in the 1950s and 1960s — programs that could prove theorems, play chess, and solve algebra word problems — created great optimism. But these programs worked only in narrow, carefully constrained domains and could not be scaled to more general tasks. The limitations became apparent in the 1970s, leading to funding cuts (the first AI winter, 1974-1980).

A second wave of enthusiasm in the 1980s focused on expert systems — programs that captured the knowledge of human experts in specific domains (medical diagnosis, mineral exploration, factory scheduling). Expert systems achieved some commercial success but proved expensive to develop and maintain, and they could not handle uncertainty or learn from experience. The second AI winter lasted from approximately 1987 to 1993.

The current AI boom, beginning around 2012, is driven by deep learning — neural networks with many layers, trained on large datasets using powerful GPUs. The convergence of three factors — massive datasets (from the internet), powerful hardware (GPUs originally developed for gaming), and improved algorithms (backpropagation, transformers) — enabled breakthroughs in image recognition, natural language processing, game playing, and scientific research. AlphaGo's defeat of world champion Lee Sedol in 2016 was a watershed moment, demonstrating that AI could master domains previously thought to require human intuition.

Large language models (LLMs) like GPT-4, trained on vast amounts of text from the internet, have demonstrated remarkable abilities to generate coherent text, answer questions, write code, and engage in conversation. The capabilities and limitations of these models — they generate fluent text but can produce false information, they lack genuine understanding but can perform complex tasks — are the subject of intense scientific and public debate.

The question of AI alignment — ensuring that AI systems pursue goals compatible with human values — has emerged as a central concern. As AI systems become more capable, the potential consequences of misalignment grow more severe. An AI system that optimizes for an objective that is subtly different from what its designers intended could produce harmful outcomes even while appearing to perform well on its specified task. The alignment problem is partly technical (how to specify objectives precisely) and partly philosophical (how to define human values in a form that can guide machine behavior).

The social implications of AI are vast and unevenly understood. AI has the potential to automate many forms of labor, both physical and cognitive, raising questions about employment, economic inequality, and the distribution of wealth. AI systems trained on historical data can perpetuate and amplify existing biases (racial, gender, economic), raising questions about fairness and justice. The use of AI in surveillance (facial recognition, predictive policing) raises questions about privacy and civil liberties. The development of autonomous weapons raises questions about the ethics of delegating life-and-death decisions to machines.

The internet's transformation of knowledge production

The internet has fundamentally altered how scientific knowledge is produced, disseminated, and consumed. Before the internet, scientific communication was mediated by journals with long publication cycles (often 6-12 months from submission to publication), limited distribution (expensive subscriptions restricted access to wealthy institutions), and geographical constraints (researchers in developing countries often could not access the latest research). The internet has compressed these cycles dramatically.

Preprint servers like arXiv (founded 1991), bioRxiv (founded 2013), and SSRN (founded 1994) allow researchers to share results immediately, before peer review. The COVID-19 pandemic demonstrated both the power and the risk of this approach: crucial findings about the virus, its transmission, and potential treatments were shared within days of discovery, accelerating the scientific response. But the absence of peer review also meant that flawed studies received widespread attention, sometimes with harmful consequences.

Open access publishing, mandated by an increasing number of funders and institutions, has made research freely available to anyone with an internet connection. The Budapest Open Access Initiative (2002), the Bethesda Statement (2003), and the Berlin Declaration (2003) established the principles of open access. Plan S, launched in 2018 by a coalition of European research funders, requires that research funded by participating organizations be published in open access journals or on open access platforms. The transition to open access has been contentious, with commercial publishers (who control many of the most prestigious journals) resisting changes that threaten their revenue model, and some open access models (article processing charges) creating new barriers for researchers without institutional funding.

The platform economy and digital labor

The digital revolution has created new forms of economic organization that challenge traditional categories of labor, property, and regulation. Platform companies like Uber, Airbnb, and Amazon Mechanical Turk create marketplaces that connect providers with consumers, but they do so in ways that often circumvent labor protections, tax obligations, and regulatory frameworks designed for earlier economic forms. The "gig economy" worker is classified as an independent contractor rather than an employee, lacking benefits, job security, and collective bargaining rights.

The concentration of economic power in a small number of technology platforms has antitrust implications. Google controls approximately 92% of the global search market. Amazon controls approximately 40% of all US e-commerce and a dominant share of cloud computing infrastructure. Facebook (Meta) controls approximately 70% of the social media market through its ownership of Facebook, Instagram, and WhatsApp. This concentration gives these companies extraordinary influence over economic activity, information flow, and political discourse. The question of whether existing antitrust frameworks are adequate to address platform monopolies, or whether new regulatory approaches are needed, is a central question in technology policy.

Environmental and sustainability dimensions

The environmental impact of the digital revolution is often overlooked but increasingly significant. Data centers consume approximately 1-2% of global electricity, a figure that is growing rapidly as cloud computing, streaming media, and AI training increase demand. Bitcoin mining alone consumes more electricity than many countries. The production of electronic devices requires rare earth minerals whose extraction has significant environmental and social costs, often borne by developing countries. Electronic waste (e-waste) is the fastest-growing waste stream globally, with only about 20% being formally recycled.

The energy consumption of AI training has become a particular concern. Training a single large language model can emit as much carbon dioxide as five automobiles over their entire lifetime. As AI models grow larger and more complex, the environmental costs of AI research are becoming a legitimate sustainability concern. Researchers have begun advocating for "green AI" — developing more efficient algorithms and computing methods that achieve the same results with less energy consumption.

Connections Master

The digital revolution connects to computer science (chapter 25) most directly, as the content of that chapter is essentially the technical foundation of everything discussed here. Turing machines, algorithms, data structures, networks, and operating systems are the building blocks of the computing infrastructure described in this historical unit.

The mathematics of computation connects to logic (chapter 24) and the foundations of mathematics (chapters 00-01). Turing's work on computability was directly motivated by Hilbert's program for the foundations of mathematics, and the Church-Turing thesis has implications for the philosophy of mathematics.

The development of the internet connects to sociology (chapter 30) through the social transformations it has produced: new forms of community, new patterns of communication, changes in work and leisure, the emergence of the "attention economy," and the role of social media in politics. The concept of the "network society," developed by Manuel Castells, provides a theoretical framework for understanding these changes.

The role of government funding in the development of computing connects to political science and public policy. The question of how societies should fund and govern technological development — whether through market mechanisms, government direction, or some combination — is a central question in the politics of technology. The tension between open-source and proprietary software, between net neutrality and carrier control, and between privacy and surveillance are contemporary manifestations of this question.

The development of AI connects to psychology (chapter 29) through the question of whether machines can think, and to philosophy (chapter 20) through questions about consciousness, free will, and the nature of intelligence. The Turing Test, proposed by Turing in 1950, asks whether a machine can fool a human into thinking it is human — a question that is simultaneously empirical and philosophical.

The environmental impact of computing — the energy consumed by data centers, the rare earth minerals required for electronics, the electronic waste generated by discarded devices — connects to earth science (chapter 27) and the broader question of sustainable technology. Training a single large language model can emit as much carbon as five automobiles over their entire lifetime. The environmental costs of computing are growing as the technology becomes more pervasive.

The globalization enabled by the internet connects to world history (chapter 32). The internet has enabled truly global supply chains, remote work, and cross-cultural communication on an unprecedented scale. At the same time, the digital divide — the gap between those with and without internet access — has created new forms of inequality both within and between nations. Approximately 2.7 billion people remain offline as of 2023, concentrated in low-income countries, rural areas, and among women and older populations. The digital divide is not merely about access to technology but about access to the economic, educational, and civic opportunities that technology enables.

The development of cryptography and cybersecurity connects to mathematics (chapter 01, number theory) and to the history of warfare and espionage. Public-key cryptography, invented by Diffie and Hellman in 1976 and implemented by Rivest, Shamir, and Adleman (RSA) in 1978, depends on the computational difficulty of factoring large numbers — a problem in pure number theory that had no known practical application before cryptography. The irony is that the most abstract branch of mathematics became essential for the most practical application: secure communication on the internet.

The digital revolution also connects to contemporary science (chapter 33.08) through the transformation of research practices. Computational methods have become essential in virtually every scientific discipline. Bioinformatics, computational chemistry, climate modeling, and particle physics all depend on massive computational infrastructure. The Large Hadron Collider at CERN generates approximately 1 petabyte of data per second, requiring sophisticated computational filtering and analysis. The Human Genome Project would have been impossible without computational methods for assembling and analyzing the three billion base pairs of human DNA. The digital revolution has not merely provided new tools for science; it has transformed what questions can be asked and what answers are possible.

Historical & philosophical context Master

The digital revolution raises fundamental philosophical questions that are still being worked out. Some of the most important concern the nature of information, the relationship between humans and machines, and the politics of technology.

Claude Shannon's "A Mathematical Theory of Communication" (1948) established information theory by showing that information can be quantified and treated mathematically. Shannon defined information in terms of surprise: a message that tells you something you already know contains no information, while a message that surprises you contains a lot. Formally, the information content of a symbol with probability $p$ is $-{\log }_{2}p$ bits. This definition made it possible to analyze communication channels, design efficient codes, and establish fundamental limits on data compression and transmission.

The concept of information as a measurable quantity has transformed not only engineering but also biology (DNA as information storage), physics (information in black holes, the holographic principle), and philosophy (the nature of meaning, the relationship between information and knowledge). The "information age" is not merely a label for the current era but a description of a fundamental shift in how humans understand the world.

The relationship between humans and computers has been a central concern since Turing's 1950 paper "Computing Machinery and Intelligence," which asked "Can machines think?" Turing proposed replacing this question with the "imitation game" (now called the Turing Test): can a machine fool a human into thinking it is human through a text-based conversation? The Turing Test has been criticized on many grounds — it tests the appearance of intelligence rather than intelligence itself, it anthropomorphizes intelligence, and it is readily gameable — but it established the question of machine intelligence as a legitimate topic for scientific and philosophical investigation.

The concept of the "singularity" — a hypothetical future point at which artificial intelligence surpasses human intelligence and begins to improve itself recursively, leading to an intelligence explosion — was proposed by mathematician I. J. Good in 1965 and popularized by Ray Kurzweil. The singularity concept raises questions about the controllability of advanced AI, the possibility of machine consciousness, and the long-term future of humanity. Whether the singularity is a realistic prospect, a fantasy, or a useful thought experiment remains hotly debated.

The politics of technology have become increasingly central to discussions of the digital revolution. The early internet was governed by a technocratic ethos: technical decisions were made by engineers and academics through organizations like the Internet Engineering Task Force (IETF), and the guiding principle was "rough consensus and running code." As the internet became economically and politically important, governance became more contested. Questions about net neutrality, content moderation, data privacy, algorithmic bias, and the power of technology platforms are now central to political discourse.

The concept of "digital colonialism" — the extension of economic and cultural power through control of digital infrastructure and platforms — has been applied to the dominance of American technology companies in global markets. The fact that Google controls search, Facebook controls social networking, and Amazon controls cloud computing in most of the world gives these companies enormous influence over information flow, public discourse, and economic activity. The question of how to govern global digital infrastructure in the interest of all people, not just shareholders, is one of the central political challenges of the 21st century.

The open-source movement, which advocates for making software source code freely available for anyone to use, modify, and distribute, represents an alternative model for technology development. Linux, Apache, Firefox, and many other essential technologies are open-source. The movement demonstrates that collaborative, non-proprietary development can produce software of the highest quality. The philosophical foundations of open-source — shared by the broader "free software" movement — connect to questions about property rights, the commons, and the social organization of creative work.

The philosophy of information

The concept of information as a fundamental quantity raises philosophical questions that go beyond engineering. If information can be measured and transmitted, is it a physical quantity like mass or energy? Some physicists have argued that information is more fundamental than matter — that the universe is, at bottom, a computational process. This "it from bit" hypothesis, proposed by John Archibald Wheeler, suggests that physical reality emerges from information processing rather than the reverse. The black hole information paradox, which asks whether information that falls into a black hole is destroyed (violating quantum mechanics) or preserved (as most physicists now believe), illustrates the deep connection between information theory and fundamental physics.

The concept of the "simulation hypothesis" — popularized by philosopher Nick Bostrom in 2003 — argues that if it is possible to create a simulation of a universe indistinguishable from reality, then statistically, we are more likely to be living in such a simulation than in the "base" reality. While this argument is more thought experiment than serious claim, it illustrates how the computational revolution has transformed philosophical speculation. The idea that reality might be computational was impossible before Turing's work, because there was no framework for understanding what "computation" meant in a rigorous sense.

Privacy, surveillance, and the public sphere

The digital revolution has transformed the relationship between individuals and institutions in ways that raise fundamental questions about privacy, autonomy, and democratic governance. The collection and analysis of personal data on a massive scale — by corporations for commercial purposes and by governments for security purposes — has created what some scholars call the "surveillance society." Shoshana Zuboff has termed this phenomenon "surveillance capitalism": the transformation of human experience into data that can be commodified, sold, and used to predict and influence behavior.

The philosophical foundations of privacy have been debated since the publication of Warren and Brandeis's influential 1890 Harvard Law Review article "The Right to Privacy," which defined privacy as "the right to be let alone." In the digital age, privacy has acquired new dimensions: the right to control one's personal data, the right to be free from algorithmic profiling, and the right to participate in public discourse without surveillance. The European Union's General Data Protection Regulation (GDPR), enacted in 2018, represents one attempt to codify these rights in law, but enforcement remains challenging in a globalized digital environment.

The impact of social media on the public sphere has been a particular focus of concern. Platforms like Facebook, Twitter (now X), and YouTube have become the primary sources of news and information for billions of people. Their algorithmic recommendation systems, designed to maximize engagement, have been shown to promote emotionally charged and polarizing content, contributing to political polarization, the spread of misinformation, and the erosion of shared factual understanding. The concept of "epistemic fragmentation" — the dissolution of a common information environment into filter bubbles and echo chambers — represents a novel challenge to democratic governance that has no clear historical precedent.

Bibliography Master

Primary sources:

• Turing, A. M. "On Computable Numbers, with an Application to the Entscheidungsproblem." Proceedings of the London Mathematical Society 42, 1936.

• Shannon, C. E. "A Mathematical Theory of Communication." Bell System Technical Journal 27, 1948.

• von Neumann, J. "First Draft of a Report on the EDVAC." 1945. Reprinted in IEEE Annals of the History of Computing 15(4), 1993.

• Cerf, V. and Kahn, R. "A Protocol for Packet Network Interconnection." IEEE Transactions on Communications 22(5), 1974.

• Berners-Lee, T. "Information Management: A Proposal." CERN, 1989.

• Engelbart, D. C. "Augmenting Human Intellect: A Conceptual Framework." SRI Summary Report AFOSR-3223, 1962.

Secondary works:

• Isaacson, W. The Innovators: How a Group of Hackers, Geniuses, and Geeks Created the Digital Revolution. New York: Simon and Schuster, 2014.

• Ceruzzi, P. E. A History of Modern Computing. 2nd ed. Cambridge, MA: MIT Press, 2003.

• Abbate, J. Inventing the Internet. Cambridge, MA: MIT Press, 1999.

• Castells, M. The Rise of the Network Society. Oxford: Blackwell, 1996.

• Mazzucato, M. The Entrepreneurial State: Debunking Public vs. Private Sector Myths. Revised ed. London: Penguin, 2018.

• Seaver, N. Computing the Folk: Algorithmic Recommendation and the Production of Culture. PhD dissertation, UC Irvine, 2015.

• Russell, S. and Norvig, P. Artificial Intelligence: A Modern Approach. 4th ed. Hoboken: Pearson, 2021.