Learning in a Brave New World of AI

Opportunities and Threats

Technology will not replace great teachers but technology in the hands of great teachers can be transformational.

— George Couros

 

Abstract

Artificial Intelligence — embodied in machines that learn from experience rather than fixed rules — has become the most consequential technology of our era, reshaping all facets of life today.  With inputs from the app Claude AI, this article explains, in plain language, what AI is and how it works, before turning to its central concern: the teaching-learning process in schools. AI's promise for education is real and substantial — personalized tutoring that addresses Bloom's long-intractable two-sigma problem, instant adaptive feedback, relief for overburdened teachers, and the democratization of quality instruction across economic and geographical divides. But the risks are equally real: the erosion of academic integrity, the atrophy of independent thinking when cognitive effort is routinely outsourced, and the impoverishment of the human community that school uniquely provides. The article concludes that AI literacy must become a core curricular competence, and that the technology's value in education ultimately depends on preserving the teacher's irreplaceable role — as mentor, moral model, and the one intelligence that really matters.

The Thinking Machine

Something extraordinary has happened in our lifetime. Machines have begun to do things we once considered the exclusive property of the human mind — composing poetry, doing math, diagnosing cancer, holding conversations, teaching children. The technology behind this upheaval is called Artificial Intelligence. Here is an attempt to explain it honestly, in plain language, from the ground up.

What Is Artificial Intelligence?

Begin with a simple observation: when a child sees a cat for the first time, she doesn't need to be given a textbook definition. She watches. She notices the fur, the ears, the particular way it moves. The next time she sees a cat — any cat, of any color — she recognizes it immediately. She has learned from experience, without being explicitly programmed.

Artificial Intelligence, at its most essential, is the attempt to give machines this same ability — to learn from experience, to recognize patterns, to make judgements, and to act usefully in situations they have not been explicitly programmed to handle.

The term itself was coined in 1956, when the American mathematician John McCarthy (see picture below) gathered a group of visionaries at Dartmouth College and proposed that "every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it." It was an audacious wager. Seven decades on, the bet is paying off — in ways that astonish even its makers.

Intelligence, Artificial and Otherwise

Human intelligence is not a single thing. It encompasses memory and reasoning, language and creativity, perception and emotion, intuition and the painstaking logic of a chess grandmaster. When we speak of AI, we are not — at least not yet — talking about a machine that possesses all of these simultaneously. What we have today is what researchers call narrow AI: systems that are extraordinarily capable within a defined domain but helpless outside it. The AI that defeats world champions at chess cannot recognise a cat. The AI that transcribes speech cannot drive a car.

What has changed dramatically in the last decade is the breadth and fluency of what narrow AI can do. The most advanced systems today — the large language models and multimodal AI that power tools like ChatGPT, Gemini, and Claude — can carry on sophisticated conversations, write code, analyze images, summarize legal documents, and explain quantum physics, all within the same conversation. This is still narrow AI, but its narrowness has become very wide indeed.

"Artificial Intelligence is not magic. It is statistics applied at breathtaking scale, trained on the accumulated text of human civilization."

The goal of general AI — a machine that reasons across all domains as flexibly as a human being — remains an open research frontier. Some researchers believe it is decades away. Others think it may never arrive in quite the form we imagine. But the debate is no longer merely philosophical: it is engineering.

How Is AI Actually Built?

To understand modern AI, you need to know three things: data, neural networks, and training. These three ideas, taken together, explain almost everything that has happened in the field since roughly 2012.

The Neural Network: A Brain Made of Numbers

The human brain contains roughly 86 billion neurons. Each neuron is connected to thousands of others. When you think — when you recognize a face, recall a name, decide to stand up — electrical signals travel through these connections in patterns. The strength of the connections changes with experience. That is, in a rough biological sense, what memory and learning are.

An artificial neural network mimics this architecture, very loosely. It consists of layers of mathematical "nodes" — the artificial analogue of neurons — connected by numerical weights. Input arrives at the first layer (an image, a sentence, a sound recording). Each node in that layer processes the input and passes a transformed signal to the next layer. This continues through many layers — sometimes hundreds — until the final layer produces an output: "this is a cat," or "the answer to your question is," or "the next word in this sentence is likely."

The "intelligence" of the network lives entirely in those numerical weights — the strengths of the connections between nodes. The key question is: how do you set them to the right values?

Training: Learning by Getting Things Wrong

You set the weights through training, and training works by making mistakes and correcting them — millions or billions of times.

Imagine you are teaching the network to identify cats. You show it a photograph of a cat and ask it to classify the image. At first, the weights are random, and the network guesses wildly — "this is a bicycle." You tell it this is wrong. A mathematical procedure called backpropagation traces the error back through all the layers and nudges every weight very slightly in a direction that would have produced a better answer. Then you show the network another image. And another. And another — until you have shown it millions of images of cats and non-cats. After enough iterations, the weights settle into values that let the network identify cats it has never seen before, with remarkable accuracy.

An analogy that helps: Think of training a neural network like sculpting with water erosion. You pour water (data) over rock (the network). Each trickle carves the rock very slightly. After an enormous number of trickles — millions, billions — the rock has taken on a shape that efficiently channels water in the right direction. Nobody sculpted it deliberately. The final form emerged from the accumulated pressure of the data itself.

Deep Learning: Going Deeper

The breakthrough that changed everything was the realization that networks with many layers — deep neural networks — could learn hierarchical features automatically. A deep network looking at a photograph does not need to be told what edges, textures, shapes, or faces are. It discovers these concepts itself, layer by layer, from raw pixels alone. This idea, called deep learning, was championed by researchers Geoffrey Hinton*, Yann LeCun, and Yoshua Bengio (who shared the Turing Award — computing's Nobel Prize — in 2018), and it is the engine behind virtually every AI achievement of the past decade.

[*Hinton was also the joint recipient of the 2024 Nobel Prize in Physics, signaling the acceptance of AI as a branch of Physics]

Language Models: Teaching Machines to Read and Write

The most consequential application of deep learning in recent years is the large language model (LLM). These are neural networks trained not on images but on text — vast libraries of books, websites, academic papers, and conversations. The training task seems almost too simple: predict the next word. Given "The cat sat on the," predict "mat." Do this billions of times across hundreds of billions of words, and something remarkable happens. The network, in learning to predict text, is forced to develop internal representations of grammar, facts, reasoning, analogy, and context. It absorbs a compressed model of much of what humanity has written down.

The architecture that made this possible is called the Transformer, introduced by Google researchers in 2017. Its key innovation — the "attention mechanism" — allows the model to consider the entire context of a sentence or paragraph at once, rather than reading word by word. This is why modern LLMs can maintain coherent conversations across long exchanges and grasp subtle references and nuances.

The result — when you scale these models to hundreds of billions of parameters and train them on much of the written record of human civilization — is a system that can discuss philosophy, write poetry, debug code, explain medical symptoms, and tutor a child in mathematics. Not because it understands in the way a human does, but because it has distilled, in its enormous weight matrices, the statistical shape of human knowledge and expression.

AI and the World: A Revolution in Human Affairs

No technology since electricity has the potential to touch every domain of human activity simultaneously. AI is already doing this. Let us take a rapid survey.

Medicine and Health

AI systems can now read medical scans — X-rays, MRIs, retinal photographs — with an accuracy that matches or exceeds specialist physicians. Google's DeepMind built a system that predicted the three-dimensional structure of nearly every known protein — a problem that had stumped biochemists for fifty years — and released the results freely to science. AI is accelerating drug discovery, personalizing cancer treatment, predicting sepsis in hospital patients hours before symptoms appear, and helping radiologists catch tumors they might otherwise miss. In low-resource settings with few doctors, AI-assisted diagnostics may prove to be one of the most life-saving technologies ever deployed.

Science and Research

AI is becoming a co-author of scientific discovery. It designs experiments, sifts through petabytes of astronomical data, models climate systems, and generates hypotheses for human researchers to test. In materials science, AI has proposed thousands of new compounds with potentially useful properties. In mathematics, AI systems have collaborated with human mathematicians to find new proofs. The pace of scientific literature has outrun any human's ability to read it; AI can synthesize thousands of papers and surface connections that would otherwise remain buried.

Work and the Economy

This is where AI's arrival feels most disruptive and most personal. Automation has always replaced physical labor — the loom, the tractor, the assembly robot. AI is different because it encroaches on cognitive labor: the work of lawyers, accountants, writers, programmers, customer service agents, financial analysts. Tasks that once required years of professional training can now be done, in draft form, in seconds.

The economic consequences are real and unequal. Some jobs will disappear. Many will be transformed. New jobs — AI trainers, prompt engineers, AI auditors, human-AI collaboration specialists — will emerge. History suggests that technological revolutions ultimately create more work than they destroy, but that cold comfort is small consolation to those whose skills become suddenly redundant. The challenge of managing this transition — through retraining, social support, and new regulatory frameworks — is one of the defining policy challenges of our era.

Creativity and Culture

AI can now generate images from text descriptions, compose music in the style of any composer, write screenplays, and produce video. This raises profound questions about authorship, originality, and the nature of creativity itself. When an AI produces a painting, who owns it? When a film studio uses AI to generate backgrounds, what becomes of the artist who used to paint them? These are not hypothetical dilemmas — they are live legal battles in courts around the world. At the same time, AI is giving ordinary people creative superpowers they never had: the amateur musician who can now orchestrate her melody, the writer who can draft ten versions of a paragraph and choose the best, the designer who can iterate visually in real time.

Governance, Ethics, and Risk

AI inherits the biases of its training data. A hiring algorithm trained on historical data may learn to discriminate. A facial recognition system trained mostly on lighter-skinned faces may fail dangerously on darker-skinned ones. Deepfake technology can make a political leader appear to say anything. Autonomous weapons raise questions about accountability in warfare that existing law is not equipped to answer.

There are also longer-term risks that serious researchers take seriously: as AI systems become more capable, ensuring they remain aligned with human values — that they do what we actually want, not just what we literally instruct — becomes both more important and harder. The field of "AI safety" exists precisely to grapple with these questions before the systems become too powerful to easily correct.

 

AI in the Classroom: Teaching, Learning, and the School

Perhaps no arena is simultaneously more promising and more fraught with risk than education. The school is where society reproduces itself — where the young are inducted into knowledge, values, skills, and the habits of mind needed for a life well-lived. AI's arrival in education is not a minor administrative convenience. It is a structural challenge to what school is for.

The Traditional Model and Its Limits

For over a century, mass schooling has operated on a factory model: one teacher, thirty students, one curriculum, one pace. The teacher delivers a lesson; the students receive it. Some thrive. Many struggle. A few are bored rigid because they understood the concept five minutes in. The constraints are real — teachers are human beings with limited time and energy — but the mismatch between what the model delivers and what children need is profound.

Benjamin Bloom, the American educational psychologist, demonstrated this in a celebrated 1984 study now known as the "two-sigma problem." Students who received one-to-one tutoring performed two standard deviations better than those in conventional classrooms. That is the difference between an average student and one at the 98th percentile. Bloom called it the "two-sigma problem" because society could not afford one private tutor for every child. AI may be about to change that arithmetic.

The AI Tutor: Personalized Learning at Scale

The most immediately transformative application of AI in school education is the intelligent tutoring system — an AI that works with a student one-on-one, adapting to their pace, identifying their misconceptions, and explaining concepts in multiple ways until the student genuinely understands.

These systems have existed in rudimentary form since the 1970s. What is new is their fluency. Modern AI tutors can converse naturally, adjust their explanations based on a student's responses, ask Socratic questions that prompt thinking rather than simply delivering answers, detect frustration or confusion from the pattern of a student's responses, and maintain a persistent model of what the student knows and doesn't know. Khan Academy's Khanmigo and several other platforms are already deploying these capabilities with real students.

Adaptive Pacing

AI systems can identify exactly where a student's understanding breaks down and address that specific gap — not the generic "chapter 5 difficulty" but this student's particular confusion about this concept at this moment.

Infinite Patience

An AI tutor can explain the same concept forty different ways without fatigue, irritation, or judgement. Students who feel embarrassed asking a teacher to repeat themselves will ask a machine as many times as they need.

Immediate Feedback

In traditional schooling, students submit work and wait days for feedback. AI provides instant, specific, actionable feedback — the kind that cognitive science tells us is far more effective for learning.

Learning Data

AI systems generate granular data on student progress — not a single exam score but a continuous record of what is known, what is emerging, and what needs work. Teachers can use this data to intervene precisely.

The Teacher's New Role

It must be said clearly: AI does not replace teachers. It changes what teachers do. The teacher's role has always contained elements that no machine can perform — mentorship, moral guidance, the modelling of intellectual passion, the recognition of a child's fragile self-esteem, the sense of belonging that a good classroom community creates. These are irreducibly human. What AI can take off a teacher's plate is the most mechanical, least creative, and most time-consuming part of their work.

Lesson planning that once took hours can be done in minutes, with the teacher's role shifting from creation to curation and critique. Grading routine exercises — spelling, arithmetic, short-answer comprehension — can be automated, freeing teachers to spend their limited attention on the work that genuinely requires it: the essay that reflects a struggling child's tentative first encounter with abstract thought, the math proof that shows a student reaching beyond what they were taught. AI handles the routine; the teacher handles the human.

An important caveat: The quality of AI-assisted teaching depends entirely on the teacher's capacity to use these tools wisely. A bad teacher with AI is still a bad teacher. A great teacher with AI becomes something closer to a great teacher with a superb support staff. Investment in teacher development is, if anything, more important in the AI era, not less.

Democratizing Access: The Equity Argument

One of the most compelling promises of AI in education is its potential to reduce the enormous inequalities in educational opportunity that currently exist. The child in a rural village with a single overworked teacher and no library now has, in principle, access to the same intelligent tutoring system as a child at an elite private school. The student whose parents cannot afford after-school coaching can get patient, adaptive support from an AI at no cost. The child who speaks a minority language can be tutored in that language.

This remains, for now, more aspiration than reality — access to devices and reliable internet remains deeply unequal, especially in the developing world. But the direction of AI's economics is encouraging: as computing costs fall and connectivity spreads, the marginal cost of providing an excellent AI tutor approaches zero. No previous educational technology has had this property.

Dangers and Distortions: The Risks for School Education

Honesty requires that we also look at what could go wrong. And much can indeed go wrong.

The Cheating Problem

The most immediately visible challenge in schools is academic dishonesty. When a student can instruct an AI to write their essay in seconds, what is the point of assigning essays? Educators are right to be disturbed. The essay is not just a product — it is a process. The struggle to organize one's thoughts, to find the right word, to discard the weak argument — these are where learning happens. When AI performs that struggle on behalf of the student, the product exists but the learning does not.

The response cannot simply be prohibition — AI is too ubiquitous to ban, and children who learn to use it thoughtfully will be better prepared for a world saturated with it. The response must be pedagogical: design assessments that cannot be faked — in-person discussions, oral examinations, iterative projects done in class under observation, portfolios of process rather than just product. The examination system as a whole will need to evolve.

The Thinking Problem

There is a subtler and deeper risk: that students who outsource their thinking to AI will never develop the cognitive muscles that sustained, difficult intellectual work builds. Reading a hard book is frustrating. Working through a mathematical proof is exhausting. Sitting with a complex ethical question until something clarifies is uncomfortable. These discomforts are not obstacles to learning — they are the learning. If AI provides the comfortable path around every cognitive difficulty, the mind may remain permanently weak. This is the educational equivalent of never walking because cars exist.

The Surveillance Risk

AI-powered educational platforms generate torrents of data about children — their errors, their learning speeds, their emotional states, their interests. In the wrong hands or with inadequate regulation, this data can be exploited commercially, used to categorize children in ways that follow them for life, or handed to governments with authoritarian inclinations. The surveillance architecture of an AI-powered school could be, if carelessly designed, profoundly hostile to the freedom and privacy that healthy development requires.

The Human Connection Risk

School is not only a place where children learn mathematics and history. It is where they learn to be with each other — to navigate disagreement, to build friendship, to develop empathy, to exist in community. An over-reliance on AI-mediated learning, where each child sits alone with their device, could impoverish this social dimension of education in ways whose consequences we would not fully appreciate until a generation later.

"The goal of education is not a well-filled bucket but a well-lit mind. AI can help pour in knowledge. Only human community can light the fire."

Teaching Critical Thinking About AI Itself

Perhaps the most important thing schools can now do is teach children to think critically about AI itself. AI systems make mistakes — sometimes subtly, sometimes spectacularly. They can produce fluent, confident, and entirely false information. They reflect biases in their training data. They optimize for the appearance of helpfulness rather than truth. A child who cannot evaluate whether an AI's output is reliable is not educated; they are merely a sophisticated consumer of machine-generated text.

Media literacy in the twenty-first century must include AI literacy: understanding what these systems are, how they work (in broad outline), what they are good at, where they fail, and how to use them as tools without becoming dependent on them as authorities.

Looking Forward: The Shape of Things to Come

It is worth stepping back and asking: what do we actually want from our schools? What is education for? The answers have always included knowledge transmission, skill formation, socialization, character development, and the cultivation of the capacity to think, question, and imagine. AI is relevant to some of these goals. It is irrelevant to others. And the danger is that, in the excitement of the technology, we allow the things AI can automate to crowd out the things only humans can do.

The school of the future — the school worth building — uses AI to free teachers from drudgery and to ensure that no child is left to flounder alone with a concept they haven't grasped. It uses AI to bring the world's knowledge to a child in a village as readily as to a child in a city. It uses AI to identify the struggling students before they fall irretrievably behind. But it insists, as fiercely as ever, on the irreplaceable importance of the teacher as mentor and model; on the classroom as community; on the essay, the debate, and the experiment as disciplines that build minds, not just fill them.

What AI cannot do is want anything for a child. It cannot look at a twelve-year-old who is bright but bored and stubborn, and recognize — as a gifted teacher can — that the stubbornness is not obstruction but unfulfilled ambition waiting for the right challenge. It cannot sit with that child after class and say: I think you are capable of something no rubric has yet measured.

That recognition — that patient, particular, irreducibly human act of seeing a child whole — is what education has always been, at its best. The technology changes. That does not.

 

 

 

Black Hole Formation

and a Supermassive Black Hole at the Core of our Galaxy

Nobel Prizes in Astrophysics & Cosmology - Part 11

(A Twelve Part Series)

Penrose, Genzel and Ghez

 The black holes of nature are the most perfect macroscopic objects there are in the universe: the only elements in their construction are our concepts of space and time.

-         S Chandrasekhar 

 

The Nobel Prize is equated with the pinnacle of human achievement in both popular perception and professional esteem.  Since it was first awarded in 1901, the annual Nobel Prize for Physics has gone to major contributions in Astrophysics and Cosmology related fields only on eleven occasions. The first ten awards (1967, 1974, 1978, 1983, 1993, 2002, 2006, 2011, 2017 and 2019) were the subjects of earlier articles (see here 1,2,3,4,5,6,7,8,9,10). The next and the last one was in 2020, one half to British mathematician and theoretical physicist Roger Penrose “for work showing general relativity predicts black holes”, and the other half jointly to Reinhard Genzel and Andrea Ghez “for the discovery of a supermassive black hole at the center of our galaxy.”  

Roger Penrose and the Mathematics of Black Hole Formation

A Prize Half a Century in the Making

In October 2020, the Royal Swedish Academy of Sciences announced that the Nobel Prize in Physics would go, in part, to Roger Penrose "for the discovery that black hole formation is a robust prediction of the general theory of relativity." Penrose was 89 years old. The work being honoured had been published in 1965 — fifty-five years earlier — in a single dense paper in Physical Review Letters. It is not an exaggeration to say that this paper changed, permanently and irrevocably, how physicists think about space, time, gravity, and the ultimate fate of matter. To understand why it was so momentous, we need to begin not with Penrose, but with Einstein, and with a question that haunted theoretical physics for the better part of four decades: can gravity, if pushed hard enough, truly crush matter out of existence?

Einstein's Masterpiece and an Uncomfortable Solution

In November 1915, Albert Einstein came up with the completed field equations of General Relativity — ten coupled, nonlinear partial differential equations relating the curvature of spacetime to the distribution of matter and energy within it. The equations were beautiful, profound, and, for most purposes, essentially unsolvable in closed form. The geometry of spacetime, Einstein was saying, is not a fixed stage on which physics plays out. It is the physics. Mass and energy warp the fabric of spacetime; that warping is what we call gravity; and objects, from planets to photons, simply follow the straightest possible paths — geodesics — through that curved geometry.

Within weeks, the German physicist Karl Schwarzschild (see picture below) found an exact solution to Einstein's equations for the case of a perfectly spherical, non-rotating mass sitting alone in otherwise empty space. The Schwarzschild solution is elegant and exact. It describes how spacetime curves around any spherical body — a star, a planet — and for practical purposes it works beautifully.

But the Schwarzschild solution contained something deeply disturbing.

At a specific radius — now called the Schwarzschild radius, or the event horizon — the mathematics seemed to go haywire. The equations produced what mathematicians call a singularity: quantities that diverge to infinity. For the Sun, this radius is about three kilometres — far beneath the solar surface, so of no physical relevance. But the question was: what if you could compress a star down to that radius? What would happen then? And lurking at the very centre, at radius zero, the equations suggested something even worse — a point where spacetime curvature itself becomes infinite.

For decades, most physicists, Einstein included, assumed these singularities were mathematical embarrassments, artefacts of the idealized symmetry of the model. Real stars, they reasoned, are not perfectly spherical. They rotate, they pulsate, they have irregularities. Surely, if you modelled a real gravitational collapse — the death of a massive star — these infinities would somehow smooth themselves out. Surely nature would find a way to avoid the catastrophe.

This intuition turned out to be spectacularly wrong.

The Landscape Before Penrose: Chandrasekhar, Oppenheimer, and Frozen Stars

To appreciate Penrose's achievement, we must sketch the troubled intellectual landscape he entered.

In the 1930s, Subrahmanyan Chandrasekhar showed (see here) that white dwarf stars — the dense remnants left when stars like our Sun exhaust their fuel — have an upper mass limit, beyond which electron pressure cannot support them against gravity. This was the Chandrasekhar limit, roughly 1.4 times the mass of the Sun. Above it, a stellar remnant must collapse further. Chandrasekhar was ridiculed for this result by Arthur Eddington, the most eminent British astrophysicist of the time, who found the conclusion — that gravity might overwhelm all resistance — physically unacceptable. Chandrasekhar was right; Eddington was wrong; the Chandrasekhar limit is now a cornerstone of stellar astrophysics.

In 1939, J Robert Oppenheimer (see picture below) and his student Hartland Snyder took the analysis further. They solved Einstein's equations for the collapse of a perfectly spherical, pressureless ball of dust — an idealized model of a massive star whose fuel has been exhausted. Their conclusion was stark: from the perspective of a distant observer, the collapsing star's surface would appear to slow asymptotically as it approached the Schwarzschild radius, its light shifting ever more to the red, growing ever dimmer, asymptotically "freezing" — but never quite reaching it in finite external time. To a hypothetical observer riding on the stellar surface, however, the collapse would proceed through the event horizon and to the central singularity in a finite and rather short proper time.

The object Oppenheimer and Snyder had described — what we now call a black hole — seemed to be a genuine prediction of the theory. But almost no one believed it described physical reality, for two reasons.

First, the model was absurdly idealized — perfectly spherical, perfectly pressureless. Real stellar collapse is messy, asymmetric, turbulent.

Second, the great Soviet physicist Lev Landau, and later the formidable team of Evgeny Lifshitz and Isaac Khalatnikov, published analyses in the early 1960s purporting to show that a general, non-symmetric collapse — one without the artificial perfect spherical symmetry — would not produce a singularity. The infalling matter, they argued, would swirl past itself, the compression would partially reverse, and the singularity would be avoided. This was the dominant view as late as 1964.

Enter Roger Penrose.

The Man and His Mind: Penrose as Geometer

Roger Penrose was born in 1931 in Colchester, England, into an extraordinarily gifted family — his father Lionel was a distinguished geneticist and psychiatrist, his brother Oliver became a noted mathematician, and his brother Jonathan a grandmaster of chess. Mathematics was, from his earliest years, as natural to Penrose as breathing.

What made Penrose unusual, even among mathematicians, was his profoundly visual and geometric mode of thinking. Where many physicists and mathematicians worked through equations and symbols, Penrose thought in pictures — in diagrams, in spatial relationships, in the global shape of mathematical objects. He was the kind of mathematician who could look at a complicated geometric situation and immediately perceive structural features that others would take pages of computation to uncover.

A New Language for Spacetime: Causal Structure and Penrose Diagrams

Before tackling singularities directly, Penrose made a fundamental conceptual contribution: he brought to General Relativity the full power of the mathematical theory of topology and global differential geometry — the study of the large-scale, overall shape of mathematical spaces.

Previous analyses of General Relativity had been largely local. You wrote down the metric — the mathematical object encoding spacetime curvature — at a particular point, and you solved equations in the neighborhood of that point. Penrose asked a different kind of question: what can we say about the global causal structure of spacetime? What does the large-scale architecture of spacetime look like, not just near a star, but across the entire universe?

He introduced — and this is one of his most enduring contributions to the toolkit of theoretical physics — what we now call Penrose diagrams (or conformal diagrams). The idea is ingenious. Spacetime extends infinitely in space and in time. By applying a mathematical transformation called a conformal rescaling, Penrose showed how to map the entire infinite spacetime — all of space for all of time — into a finite diagram, while preserving the causal relationships between points. The structure of light cones, which determine what can causally influence what, is faithfully preserved. You can see, in a single compact picture, the entire life history of a spacetime.

Penrose diagram of an infinite Minkowski universe, horizontal axis u, vertical axis v

These diagrams revealed immediately that the event horizon of the Schwarzschild black hole was not a true singularity at all — it was a coordinate singularity, an artefact of the coordinate system being used, like the apparent problem at the North Pole in certain map projections of the Earth. The actual spacetime was smooth at the event horizon; only the coordinate description broke down there. The true singularity was at the centre, at r=0.

But Penrose's topological thinking went much further than diagrammatic clarity.

The Key Idea: Trapped Surfaces

The central and revolutionary concept in Penrose's 1965 paper is the notion of a trapped surface. To understand this, think first about how light behaves in ordinary, flat, empty space.

Imagine a flash of light emitted from a small sphere. The light radiates outward in an expanding spherical shell, growing larger and larger as it travels. Now imagine a second flash, directed inward — toward the center of the sphere. That converges inward, shrinks to a point, then re-expands outward. In flat space, outward-directed light always expands, and inward-directed light initially converges then expands. This is obvious and familiar.

Now imagine you are deep inside a region where gravity is enormously strong — so strong that spacetime itself is severely curved. Consider a closed surface — a sphere — inside this region. Emit light from this sphere in both directions: outward and inward. In normal circumstances, the outward flash expands and the inward flash converges. But in a sufficiently strong gravitational field, something extraordinary happens: both the outward and the inward flashes converge. Even the light you direct outward gets pulled back inward by the enormous curvature of spacetime. Every ray of light emitted from the surface, in any direction, is heading toward smaller and smaller regions.

This is a trapped surface (see diagram below from the original publication).

Once a trapped surface forms, Penrose realised, there is no escape — not for light, not for matter, not for anything. The geometry of spacetime itself has closed off all outward-pointing paths. The formation of a trapped surface is the geometric signature that the point of no return has been passed. Inside a trapped surface, the future is, in a deep topological sense, bounded and closed.

This insight — clean, visual, geometric — is the foundation of everything that follows.

The Singularity Theorem: The 1965 Paper

In his 1965 paper (see above), Penrose proved the following theorem, in a form that stunned the physics community:

If spacetime contains a trapped surface, and if the energy conditions on matter are satisfied, and if the spacetime is globally well-behaved in a certain topological sense, then the spacetime must be geodesically incomplete — that is, there must exist at least one light ray or material particle path that comes to an abrupt end after a finite proper time.

Let us unpack this carefully, because every element is important.

·      Geodesic incompleteness is the mathematical way of saying that something catastrophic happens. A geodesic is a path through spacetime that a freely falling particle or a ray of light follows. If a geodesic is "incomplete," it means the path simply terminates — it cannot be extended any further. There is no more spacetime "ahead." The particle, or the light ray, runs into the edge of the world. This is what we mean, mathematically, by a singularity.

·   The energy conditions Penrose invoked are physically very reasonable assumptions about matter — essentially, that matter and energy have positive density and that gravity is attractive. No exotic matter that "gravitationally repels" is allowed. This was important, because one potential escape hatch from singularities was the idea that quantum effects or exotic matter might generate a kind of pressure that halts the collapse. Penrose's theorem shows that, within classical General Relativity, as long as matter obeys these natural conditions, singularities are inevitable.

·    The trapped surface condition is the trigger. The theorem says: once a trapped surface forms in your spacetime, singularity is guaranteed, full stop. There is no escape, no matter how asymmetric or turbulent or non-spherical the collapse is. The Lifshitz-Khalatnikov claim that asymmetric collapse avoids singularities was wrong — and Penrose's theorem proved it with complete mathematical rigor.

The argument Penrose constructed to prove this was not a calculation in the traditional sense. There was no solving of differential equations, no laborious perturbation theory. It was a topological argument — a proof about the global shape and connectivity of spacetime. He showed that if a trapped surface exists and the energy conditions hold, then the set of all future-directed light rays from the trapped surface must be compact — it cannot keep expanding indefinitely. And then, using a beautiful result from the theory of smooth manifolds, he showed this compactness forces at least one light ray to terminate. The elegance of the argument is such that it takes less than eight pages to present, yet demolishes four decades of doubt.

The Lifshitz-Khalatnikov result, it turned out, had a subtle error — they had proven that a special class of singular solutions was unstable, but they had not proven, as they thought they had, that generic solutions were singularity-free. Khalatnikov himself, to his enormous credit, acknowledged the error after Penrose's paper appeared.

What the Theorem Actually Tells Us — and What It Doesn't

It is worth being precise about what Penrose proved, because popular accounts sometimes blur important nuances.

Penrose proved that geodesic incompleteness — the termination of at least one path through spacetime — is inevitable once a trapped surface forms. He did not prove, in this 1965 paper, that a singularity of diverging curvature must occur at the end of that path. In principle, the incompleteness could be of a different mathematical character. However, subsequent work — including collaboration between Penrose and Stephen Hawking, to which we will return — strongly suggests that curvature divergence is the generic outcome.

Penrose also proved that the theorem applies to gravitational collapse — the physical process of a massive star running out of nuclear fuel and being crushed by its own weight. The critical question then became: can a trapped surface actually form in realistic astrophysical collapse? Here the answer, as Penrose and others showed, is yes. For a sufficiently massive collapsing star, once enough mass is concentrated within a small enough region, a trapped surface forms. The exact threshold depends on the mass and the equation of state of the matter, but it is clear that nature, in the collapse of massive stars, routinely creates the conditions Penrose's theorem requires.

This is the "robust" quality singled out by the Nobel Committee. Black holes are not a fragile artefact of unrealistic symmetry. They are a generic, unavoidable consequence of General Relativity whenever sufficient mass collapses, regardless of the messy details.

Hawking, the Cosmological Singularity, and the Grand Synthesis

Penrose's 1965 theorem concerned black holes — the future singularities produced by gravitational collapse. One of the people most immediately galvanised by the paper was a young Cambridge graduate student named Stephen Hawking (see a later picture of the remarkable man below), who had the insight to run Penrose's argument backwards in time.

If collapsing matter inevitably produces a future singularity, Hawking reasoned, then the expanding universe — which, when run backwards, is a collapsing universe — should also lead to a singularity in the past. That singularity is, of course, the Big Bang. Working together between 1966 and 1970, Penrose and Hawking published a series of increasingly powerful singularity theorems, culminating in the celebrated Hawking-Penrose theorem of 1970.

This theorem showed, under very general conditions, that any spacetime that satisfies Einstein's field equations, contains ordinary matter satisfying reasonable energy conditions, and is either collapsing or expanding (as ours demonstrably is), must contain singularities — either in the future, as black holes, or in the past, as a Big Bang. Or both.

The Hawking-Penrose theorems established, definitively, that singularities are not pathological special cases but are generic features of the relativistic universe. Einstein's theory carries, encoded in its geometry, the seeds of its own incompleteness. Where curvature diverges, where geodesics terminate, General Relativity breaks down and a deeper theory — perhaps quantum gravity — must take over. This insight has shaped the research agenda of theoretical physics for half a century.

The Physical Picture: What Happens When a Star Dies

Let us now translate all this mathematics into a physical narrative.

A massive star — say, twenty times the mass of the Sun — spends its life in a battle between two titanic forces. The outward pressure generated by nuclear fusion in its core, and the inward pull of its own self-gravity. For millions of years, these forces are in balance: the star shines, it burns, it lives.

But nuclear fuel is finite. When the core has converted all its available material into iron — the endpoint of stellar nucleosynthesis, beyond which no further fusion releases energy — the outward pressure fails. In less than a second, the core collapses catastrophically. If the remaining core mass exceeds a certain threshold (roughly two to three solar masses), neither electron pressure nor neutron pressure can halt the collapse. The infalling matter crosses the Schwarzschild radius. A trapped surface forms.

From this moment, Penrose's theorem takes hold. The geometry of spacetime inside the event horizon is such that all future-directed paths — for matter, for light, for information — lead toward decreasing radius. There is no future-pointing direction in spacetime that leads outward. The singularity at the centre is not a point in space — it is, in a profound sense, a moment in time, lying in the inevitable future of everything inside the horizon. The matter that formed the star, compressed beyond all physical limits, reaches the singularity in a finite proper time — roughly microseconds after horizon crossing for a stellar-mass black hole.

What happens at the singularity itself, we do not know. Classical General Relativity describes the approach to it but cannot describe what occurs there, because the theory breaks down when curvature becomes infinite. This is widely understood as a signal that quantum effects — the quantisation of spacetime itself — must become important, and that a theory of quantum gravity is required to describe the singularity. This remains, as of today, one of the deepest unsolved problems in all of physics.

The Event Horizon and the Outside World

To an observer far away from the black hole — watching the star collapse through a telescope — the story looks different. As the stellar surface falls inward and approaches the event horizon, the light it emits becomes increasingly gravitationally redshifted. The star appears to grow dimmer and dimmer, its surface appearing to freeze asymptotically at the horizon, never quite crossing it in the time of the distant observer. In practice, the star's visible light fades exponentially and becomes undetectable within a fraction of a second.

The black hole that remains is invisible — no light escapes it. It is detectable only through its gravitational effects on surrounding matter, through the accretion of infalling gas that heats up and radiates X-rays as it spirals inward, and through the gravitational waves emitted when two black holes collide and merge — the phenomenon spectacularly detected by LIGO in September 2015, in the event known as GW150914 (see here).

The Information Paradox and Quantum Horizons

Penrose's work also planted the seed of one of the most contentious and productive puzzles in modern theoretical physics: the black hole information paradox.

In 1974, Hawking made a stunning discovery by combining quantum field theory with curved spacetime. An event horizon, he showed, is not perfectly black in the quantum mechanical sense. Due to quantum fluctuations near the horizon, a black hole slowly emits thermal radiation — Hawking radiation — and, over astronomically long times, gradually evaporates. If a black hole can evaporate completely, what happens to the information about all the matter that fell into it? Does it disappear from the universe? That would violate a fundamental principle of quantum mechanics. Or is it somehow encoded in the emitted radiation? This question — whether black holes destroy information — has animated theoretical physics for fifty years and has not yet been fully resolved. The tension between General Relativity and quantum mechanics, first exposed by Penrose's singularity theorems, continues to generate some of the most profound physics of our era.

What Made Penrose's Contribution Revolutionary

Looking back, the genius of Penrose's 1965 paper lies not just in what it proved but in how it proved it.

Previous work on singularities had been computational — you solved the equations in a specific model, found a singularity, and worried about how special your model was. Penrose bypassed all that. He proved a theorem — a result that holds for any spacetime satisfying the stated conditions, regardless of symmetry, regardless of the details of the matter distribution, regardless of the messiness of real astrophysics. This was a shift in methodology as much as a specific result: General Relativity could be analysed by the global, topological methods of modern differential geometry, yielding inescapable conclusions from minimal assumptions.

This opened an entirely new field — global methods in General Relativity — which Penrose himself developed further and which Hawking, Robert Geroch, Brandon Carter, Werner Israel, and many others elaborated over the following decades. Results like the area theorem (the total area of event horizons can never decrease, a result strikingly analogous to the second law of thermodynamics), the uniqueness theorems for black holes ("black holes have no hair" — they are characterised entirely by mass, charge, and spin), and the cosmic censorship conjecture (Penrose's own conjecture that naked singularities — singularities not hidden behind event horizons — are generically forbidden by nature) all flow from the conceptual framework Penrose established.

Penrose the Visionary: Beyond the Nobel Work

It is worth noting, for completeness, that Penrose's influence on fundamental physics extends far beyond his Nobel work. He has developed, over decades, the twistor programme — an ambitious attempt to reformulate the laws of physics in terms of a complex geometric space he invented called twistor space, in which points of spacetime emerge as derived objects and the fundamental entities are light rays. This programme has had unexpected and profound connections to modern particle physics, including deep links to the mathematical techniques used in calculating particle scattering amplitudes. He has also written extensively, controversially, about the implications of Gödel's incompleteness theorems for the nature of mathematical intuition, and about the possible role of quantum effects in consciousness.

But it is the eight-page paper of 1965, with its trapped surfaces and its topological proof of inevitable incompleteness, that stands as his definitive contribution to physics — the work that established, beyond any reasonable doubt, that the formation of black holes is not a curiosity of idealised toy models but an inescapable consequence of the geometry of reality itself.

Nobel Prize, 2020

When the Nobel Committee called Penrose at his home in Oxford in October 2020 to inform him of the prize, he was 89 years old. He had been doing mathematics every day, as he had throughout his adult life. The other half of the 2020 Nobel Prize in Physics went to Reinhard Genzel and Andrea Ghez, who led the teams that discovered the supermassive black hole at the centre of the Milky Way galaxy, Sagittarius A* — an object about four million times the mass of the Sun, confirmed through decades of painstaking observation to be exactly the kind of object Penrose's mathematics demanded must exist.

It was a perfect pairing: the theorist who proved black holes must form, and the observers who found one. The mathematics of 1965 and the astronomy of four million light-years came together, in Stockholm, in the autumn of a pandemic year, to honour a truth about the universe that had taken fifty years to fully appreciate.

That truth, stated simply, is this: when enough matter collapses under its own gravity, the geometry of spacetime itself conspires to create a region from which nothing — not matter, not light, not information, not time itself, in any meaningful sense — can escape. And this is not a special case, not an artefact of idealisation. It is written into the equations of General Relativity as surely as anything in all of physics. Roger Penrose read that writing in the geometry, fifty-five years before the rest of the world fully caught up.

* * * * *

 

Discovery of Supermassive Black Hole

The Mystery at the Centre of the Galaxy: A Story of Stars, Patience, and Gravity

For much of the twentieth century, the centre of the Milky Way was a place of rumour and inference. Optical astronomers could not even see it — the galactic core lies some 26,000 light-years away, buried behind immense clouds of dust and gas that swallow visible light almost entirely. What little evidence existed came from radio waves. In 1974, Bruce Balick and Robert Brown, working with the Green Bank radio telescope in West Virginia, detected a compact, anomalously bright radio source at the precise dynamical centre of the Galaxy. They eventually named it Sagittarius A*— the asterisk denoting its exceptional brightness, in the notation of radio astronomers. It was intriguing, but its nature was deeply uncertain.

The question was stark: was there a supermassive black hole lurking there, as some theorists were beginning to suspect existed at the hearts of galaxies? Or could other explanations — a dense cluster of massive stars, for instance — account for what was observed? To answer it, you needed to measure the gravitational field at the galactic centre with precision. And to do that, you needed to watch stars move.

Two Teams, Two Telescopes, One Question

From the early 1990s, two research groups on opposite sides of the planet took up this challenge, independently and in parallel, in what became one of the most sustained observational campaigns in the history of astronomy.

Reinhard Genzel — a German astrophysicist at the Max Planck Institute for Extraterrestrial Physics in Garching and also at UC Berkeley — led the European team. They worked primarily with the Very Large Telescope (VLT) of the European Southern Observatory on Cerro Paranal in Chile, one of the most powerful ground-based observatories ever built (see picture below).

Andrea Ghez — an American astronomer at UCLA — led the American team, using the Keck Observatory on Mauna Kea in Hawaii (see picture below), whose twin 10-metre mirrors were among the largest on Earth.

Both groups faced the same formidable obstacle: the galactic center is not just obscured by dust, it is turbulent. Earth's atmosphere distorts the incoming light, smearing stellar images into blobs. The key technology that made the whole enterprise possible was adaptive optics (see diagram below) — a system in which a deformable mirror, adjusting its shape hundreds of times per second based on the twinkling of a reference star, corrects for atmospheric distortion in real time, recovering the sharpest possible image. Combined with near-infrared detectors (infrared light penetrates the dust far better than visible light), adaptive optics gave both teams the ability to resolve individual stars deep within the galactic center.

The Orbits that Changed Everything

What they saw, over years and then decades of patient observation, was extraordinary.

Stars near Sagittarius A* were not drifting gently. They were racing — moving at thousands of kilometres per second, tracing tight, curved paths around an invisible point. As the data accumulated through the 1990s and 2000s, these paths resolved into complete orbits, with periods as short as fifteen or sixteen years. One star in particular, designated S2 by Genzel's group (and S0-2 by Ghez's), became the centrepiece of the story. It sweeps around the galactic centre in an ellipse so elongated that at closest approach — its periapsis — it comes within about 120 AU of Sagittarius A*, travelling at nearly three percent of the speed of light.

The implications were inescapable. By applying Kepler's third law to these stellar orbits — just as Newton used the Moon's orbit to weigh the Earth — both teams could calculate the mass of whatever lay at the focus of these ellipses. The answer converged on something in the range of four million solar masses, all confined within a region smaller than our Solar System (see illustration below).

No known astrophysical object other than a black hole could pack that much mass into so small a volume. A cluster of dark stellar remnants would itself have collapsed into a black hole on astronomical timescales. The case for a supermassive black hole — Sagittarius A* — became overwhelming.

Genzel's team published key orbital results through the late 1990s and early 2000s, and Ghez's team did the same. There was a productive, occasionally competitive, mutual watching between the two groups. Their results were consistent and their methods independent, making the conclusion all the stronger.

The General Relativistic Signature

The story acquired a further layer of depth in 2018–2019. With S2 completing another close passage around Sagittarius A*, both teams could now test general relativity in the strong-field regime near the black hole.

Ghez's team measured the gravitational redshift of S2's light during its 2018 periapsis passage — the stretching of light wavelengths as photons climb out of the black hole's deep gravitational well — precisely as Einstein's theory predicts. Genzel's team confirmed this and additionally detected the Schwarzschild precession of S2's orbit: the slow rotation of the orbital ellipse around the focus, the same effect (though vastly larger in scale) as the precession of Mercury's perihelion that was one of general relativity's earliest triumphs. Newtonian gravity alone cannot account for these effects. The galactic center had become a laboratory for Einstein's theory.

The Nobel Prize of 2020

The 2020 Nobel Prize in Physics was divided into two parts.

One half went to Roger Penrose, the British mathematical physicist, for his landmark 1965 proof that the formation of black holes is a robust, inevitable consequence of general relativity — not a mathematical curiosity dependent on perfect symmetry, as many had believed, but something that must happen whenever enough mass collapses within a critical surface.

The other half was shared equally between Reinhard Genzel and Andrea Ghez — the first time the physics prize had gone to a woman since Maria Goeppert Mayer in 1963 — "for the discovery of a supermassive compact object at the centre of our galaxy."

The Nobel Committee's pairing was elegant and deliberate. Penrose had shown, theoretically, that black holes must exist. Genzel and Ghez had shown, observationally, that one of them does exist — right at the heart of the galaxy we inhabit. Theory and observation, separated by five decades, rewarded together.

Reinhard Genzel (1952 - ) – A Biographical Sketch

Reinhard Genzel (born 1952 in Bad Homburg, Germany) is a German astrophysicist renowned for his pioneering work in infrared and radio astronomy. He studied physics at the Universities of Freiburg and Bonn, earning his PhD in 1978, and later worked at institutions such as the Harvard-Smithsonian Center for Astrophysics and the University of California, Berkeley. He has long been associated with the Max Planck Institute for Extraterrestrial Physics, where he served as director.

Genzel led a research group that, beginning in the 1990s, developed high-resolution observational techniques (including adaptive optics) to track the motion of stars near the centre of the Milky Way. By carefully measuring their orbits, his team demonstrated that these stars are bound to an extremely massive and compact object—now identified as a supermassive black hole (Sagittarius A*).

For this discovery, he shared the 2020 Nobel Prize in Physics with Andrea Ghez (and Roger Penrose, who received the other half), establishing direct observational evidence for a black hole at our galaxy’s center.

Andrea Ghez (1965 - ) – A Biographical Sketch

Andrea Mia Ghez (born 1965, New York City, USA) is an American astrophysicist and professor at the University of California, Los Angeles (UCLA). She completed her undergraduate studies at MIT and obtained her PhD from Caltech in 1993.

Ghez independently led a competing research team that used the Keck Observatory in Hawaii and advanced imaging techniques such as speckle imaging and adaptive optics to study stars orbiting the Milky Way’s center. Her precise measurements of stellar motions provided compelling evidence that a supermassive black hole governs the dynamics of this region.

She became only the fourth woman* to win the Nobel Prize in Physics, and her work has significantly advanced high-resolution astronomical imaging. Like Genzel, she was awarded the 2020 Nobel Prize for demonstrating that an invisible, extremely massive object lies at the heart of our galaxy.

[*The other three were: Marie Curie (1903), Maria Goepper Mayer (1963) and Donna Strickland (2018). Anne L’Huillier (2023) joined this august group later.]

 

Epilogue: The Shadow of Sagittarius A*

The story did not end with the Nobel announcement in 2020. In May 2022, the Event Horizon Telescope collaboration — a planet-spanning network of radio observatories functioning as a single Earth-sized dish — released the first direct image of Sagittarius A*: a glowing ring of superheated plasma surrounding a dark central shadow, the black hole's event horizon silhouetted against the accretion glow. It was the visual confirmation of everything the stellar orbits had implied. The invisible object that Genzel and Ghez had spent thirty years weighing through the movements of stars finally had a face.

What began as a faint anomaly on a radio map in 1974 had, by the early twenty-first century, become one of the most scrutinised objects in the cosmos — and one of the most compelling demonstrations that general relativity, a theory born in Einstein's imagination in 1915, describes gravity all the way down to the most extreme environments the universe can produce.

 

Postscript

This concludes my narratives of twenty-six recipients of the Nobel prize in Physics on eleven occasions to contributions specifically in Astrophysics and Cosmology, from Hans Bethe in 1967 to Andrea Ghez in 2020. It has been an exciting voyage of discovery, with more to come.

Incidentally, today (April 24) marks the 36th anniversary of the launch of the Hubble Space Telescope (see picture below) which has been instrumental in contributing enormously to many of the discoveries in Astrophysics and Cosmology documented in this series of articles.