Is String Theory Dead? A Deep Reflection from a Fermilab Physicist

Introduction: A Theory That Cannot Be "Killed"

In a conversation on the Lex Fridman podcast, Don Lincoln, a senior physicist at Fermi National Accelerator Laboratory, offered his unique perspective on the controversial question of whether string theory is dead. The discussion touches on the most fundamental dilemma in modern theoretical physics: if an elegant theory cannot be experimentally verified or falsified, what is its scientific value?

Fermi National Accelerator Laboratory (Fermilab), located in Illinois, USA, is one of the world's premier particle physics research institutions. It previously operated the Tevatron, which was the world's most powerful particle collider before the LHC was built. As a senior scientist at the lab, Don Lincoln has long been engaged in experimental particle physics research while also being a well-known science communicator. This gives his assessment of string theory both the pragmatic perspective of an experimental physicist and a deep understanding of the theoretical frontier.

String Theory's Core Dilemma: The Landscape Problem

String theory was initially met with great expectations. Physicists hoped it would uniquely explain our universe—why physical constants have their particular values, why certain particles exist, and why there are four fundamental forces. However, as research progressed, a discouraging fact emerged: string theory has an enormous "landscape," meaning the number of possible universes it permits is astronomically large, estimated at around 10^500.

The severity of this "landscape problem" became particularly clear around 2003. At that time, Kachru, Kallosh, Linde, and Trivedi (collectively known as KKLT) published a landmark paper demonstrating how to construct stable vacuum states with a positive cosmological constant within the string theory framework—but at the cost of discovering that the number of such vacuum states is astronomically vast. Subsequently, Leonard Susskind of Stanford University formally proposed the concept of the "string theory landscape" and boldly connected it to the anthropic principle: perhaps our universe is just one of 10^500 possibilities, and the reason we observe these particular physical constants is simply because only these values allow observers (i.e., us) to exist. This position sparked fierce debate in the physics community, with critics arguing that anthropic reasoning essentially abandons physics' mission to explain "why."

What does this mean? As Lex Fridman pointed out in the conversation: string theory can describe all sorts of universes, so you can selectively tune parameters to describe our universe. This essentially strips the theory of its predictive power—a theory that can explain everything actually explains nothing.

Beyond the landscape problem, string theory also relies on unobserved extra dimensions (typically 6 or 7 compactified spatial dimensions), which further widens the gap between theory and experiment. These extra dimensions are thought to be "compactified" at extremely tiny scales (typically on the order of the Planck length, about 10^-35 meters), so small that we cannot detect them at everyday scales or even with the most powerful particle accelerators. The geometry of these extra dimensions is usually described by so-called Calabi-Yau manifolds—a class of six-dimensional compact spaces with special mathematical properties. The problem is that the number of possible Calabi-Yau manifold shapes is enormously large, and each different geometry corresponds to a different set of low-energy physics laws—this is the geometric root of the landscape problem.

Lincoln's Defense: Measurements Can Save the Theory

Don Lincoln's attitude toward string theory is not one of simple rejection. He raises an important argument: the landscape problem is not a fatal flaw of the theory, but rather the result of a lack of experimental constraints.

He used an elegant analogy: just as the expression "X + 5" can take any number, but once we know "X + 5 = 9," all possibilities except 4 are immediately eliminated. Similarly, if we could connect string theory's predictions to physical measurements, we could dramatically reduce those "alternative universes" that don't match reality, ultimately retaining the unique solution that describes our universe.

Lincoln admits that he personally loves the core concept of "vibrating strings"—different particles corresponding to different vibrational modes of strings, which is mathematically extremely elegant. The core picture of string theory is this: all fundamental particles in the universe—electrons, quarks, photons, gravitons—are not point-like zero-dimensional objects, but rather different vibrational modes of tiny one-dimensional strings (with a length of approximately the Planck length, 10^-35 meters). Just as a violin string can vibrate to produce different notes, the different vibrational frequencies and modes of fundamental strings correspond to different particle masses and quantum numbers. The profound aspect of this picture is that it replaces the dozens of seemingly unrelated fundamental particles in the Standard Model of particle physics with a single unified object (the string), and naturally incorporates gravity into the quantum mechanical framework—something no other theory has been able to achieve.

It's worth noting that despite string theory's difficulties with direct experimental verification, it has already had a profound impact in the field of mathematical physics. The most famous example is the AdS/CFT correspondence (Anti-de Sitter/Conformal Field Theory duality) proposed by Juan Maldacena in 1997. This discovery revealed a stunning equivalence between gravitational theories and quantum field theories, and has been widely applied across multiple domains from black hole physics to condensed matter physics and even quantum information theory. Additionally, some physicists are exploring indirect ways to test string theory, including searching for possible string theory signatures in the cosmic microwave background radiation, studying extra-dimensional effects in gravitational wave signals, and searching for supersymmetric particles at the LHC (supersymmetry being a key mathematical element of string theory, though it has not yet been experimentally discovered).

But he also honestly acknowledges: "Until we can verify it, we can't."

Why String Theory Is "Hard to Kill"

Lincoln raises a profound epistemological point: to truly "kill" a theory, you need it to make clear predictions, and then experiments must prove those predictions wrong. But the problem with string theory is precisely that it currently cannot make clear, experimentally testable predictions.

This creates a paradoxical predicament:

• A good scientific theory should be falsifiable (Popper's criterion)
• String theory currently cannot be falsified
• Therefore it can neither be confirmed nor denied
• It exists in a state of scientific "limbo"

The "Popper's criterion" mentioned here originates from Karl Popper, one of the most influential philosophers of science of the 20th century. In his 1934 work The Logic of Scientific Discovery, Popper proposed that the fundamental distinction between scientific and non-scientific theories lies not in whether a theory can be verified, but in whether it can be falsified (falsifiability). A genuine scientific theory must make predictions that could potentially be overturned by observational facts—if a theory can explain away any observational result, it is not science. This criterion has long been regarded as the "demarcation line" separating science from pseudoscience.

The falsifiability problem of string theory sparked intense public debate at the December 2015 Munich workshop "Why Trust a Theory." Critics represented by Columbia University mathematician Peter Woit and physicist Lee Smolin argued that string theory has strayed from the path of science, degenerating into a purely mathematical game. String theory's supporters—including Nobel laureate David Gross and string theory pioneer John Schwarz—countered that the falsifiability criterion is overly simplistic, and that a scientific theory's value should also consider its internal consistency, mathematical richness, and compatibility with known physics. Some supporters also pointed out that string theory does make certain "in principle testable" predictions (such as the existence of extra dimensions and supersymmetric particles), but current technology cannot yet reach the required energy scales.

The Real Crisis: Brain Drain and Opportunity Cost

In Lincoln's view, the real threat to string theory is not falsification but abandonment. Since the 1970s, physicists have invested approximately 50 years of effort in string theory without achieving a decisive breakthrough.

The development history of string theory itself is a dramatic scientific saga. In 1968, Italian physicist Gabriele Veneziano accidentally discovered a mathematical formula (the Veneziano amplitude) while studying the strong interaction, which was later recognized as explainable by vibrating strings—this was the seed of string theory. In 1984, Michael Green and John Schwarz proved that superstring theory is free of mathematical anomalies under certain conditions (the "First Superstring Revolution"), sparking enormous enthusiasm in the physics community. In 1995, Edward Witten proposed M-theory, unifying the previously seemingly different five superstring theories into a higher-dimensional framework (the "Second Superstring Revolution"). However, in the more than two decades since, despite continual refinement of mathematical tools, progress in connecting string theory to the observable physical world has been extremely limited.

He draws an analogy to the quantum mechanics interpretation problem that began in the 1940s. Lincoln recalls that as a child in the 1970s, he wanted to study the meaning of quantum mechanics, but by graduate school, he realized that people far smarter than him had already spent most of their lives on this problem without achieving decisive progress.

This raises the cruel reality that every young scientist must face: Should I dedicate my life to a direction that may see no progress within my lifetime?

Lincoln believes this is exactly what is currently happening in the field of string theory. More and more young physicists are avoiding string theory when making career choices—not because it has been proven wrong, but because it may not yield answers for decades.

This trend is already reflected in academic job market data. In the 1980s and 1990s, string theory was the hottest research direction in theoretical physics, with faculty positions at top universities flowing heavily toward string theorists. But after entering the 21st century, as the LHC failed to discover supersymmetric particles (the absence of supersymmetry became particularly notable after the 2012 discovery of the Higgs boson), and as string theory itself failed to deliver on its promise of a "theory of everything," more and more young theoretical physicists have turned to other directions—including quantum information and quantum computing, topological phases in condensed matter physics, cosmological data analysis, and competing quantum gravity approaches such as Loop Quantum Gravity. Loop Quantum Gravity, developed by Lee Smolin, Carlo Rovelli, and others, does not require extra dimensions or supersymmetry and directly quantizes spacetime itself. While it faces its own difficulties, its more "conservative" assumptions and more direct physical picture hold appeal for some young researchers.

Deeper Implications: The Boundaries of Scientific Progress

This conversation reveals a broader challenge facing modern fundamental physics. When a theory's energy scale far exceeds human experimental capabilities (string theory's characteristic energy approaches the Planck scale, 15 orders of magnitude higher than the LHC), does the traditional "hypothesis-experiment-verification" scientific methodology still apply?

The numbers here are worth spelling out: the Planck energy is approximately 10^19 GeV (giga-electron volts), the energy scale at which quantum gravitational effects become non-negligible and where string theory predicts new physical phenomena should appear. Meanwhile, humanity's most powerful particle accelerator—CERN's Large Hadron Collider (LHC)—has a center-of-mass collision energy of approximately 1.3×10^4 GeV, or 13 TeV. There is a gap of about 15 orders of magnitude between the two. To directly probe Planck-scale physics, we would need to build an accelerator a quadrillion times more powerful than the LHC—with current technology, such an accelerator would need to extend to the scale of the Milky Way galaxy, clearly far beyond the realm of human engineering capability.

String theory's predicament may not be a problem with string theory itself, but rather a manifestation of the chasm between human cognitive abilities and the ultimate laws of the universe. As Lincoln implies, string theory "might be right," but we may never be able to know.

Scientific history is not without similar precedents. In the mid-19th century, atomic theory (the idea that matter is composed of indivisible atoms) was already widely used in chemistry, but many distinguished physicists—including Ernst Mach and Wilhelm Ostwald—insisted that atoms were merely useful mathematical fictions, because no experiment at the time could directly observe atoms. It wasn't until Einstein's 1905 theoretical explanation of Brownian motion, followed by Jean Perrin's precision experiments, that the real existence of atoms was universally accepted by the scientific community. From Dalton's 1803 proposal of atomic theory to its ultimate confirmation, an entire century elapsed. Is string theory also waiting for its own "Brownian motion moment"? Or does the verification gap it faces differ in nature from any historical precedent? This is one of the most thought-provoking questions in contemporary philosophy of science.

This also holds lessons for the AI and computational science fields: as we build increasingly complex models, how do we ensure they remain verifiable and interpretable? The lesson of string theory reminds us that elegant mathematical structures are not equivalent to physical reality, just as sophisticated algorithmic architectures are not equivalent to practical utility. In the field of deep learning, we similarly face the problem of "over-parameterization"—when a large language model with billions of parameters can fit almost any data pattern, does it truly "understand" anything? This is strikingly similar to the philosophical core of string theory's landscape problem.

Key Takeaways

• String theory's landscape problem allows approximately 10^500 possible universes, causing the theory to lose predictive power
• Don Lincoln argues string theory is hard to "kill" because it cannot make clear, experimentally testable predictions
• The real threat to string theory is brain drain—young scientists are unwilling to invest in a research direction that may yield no results in their lifetime
• Lincoln draws an analogy between string theory and the quantum mechanics interpretation problem, arguing both face the dilemma of "brilliant people spending lifetimes without decisive progress"
• If future work can connect string theory predictions to experimental measurements, the landscape problem could be dramatically reduced through elimination