When Science Faced Limits (Again): The New Impossibilities
February 2023. Physicist Sabine Hossenfelder publishes a provocative essay: "Particle Physics Is in Trouble."
Her argument: Fundamental physics has made no major breakthroughs in 50 years.
Standard Model completed 1970s. Higgs boson (2012) confirmed what was predicted in the 1960s.
Since then? Nothing.
No new particles beyond the Standard Model. No quantum gravity. No dark matter detection. No supersymmetry. No extra dimensions.
Fifty years. Billions of dollars. Thousands of physicists.
No fundamental discoveries.
Not because physicists are worse. Not because they're not trying.
Because they're hitting limits.
Physical limits: Need particle accelerators larger than Earth to test some theories.
Observational limits: Can't see beyond cosmic horizon. Can't access Planck scale. Can't observe inside black holes.
Economic limits: Next-generation collider costs $100+ billion. For uncertain returns.
Theoretical limits: String theory makes no testable predictions. Multiverse is unfalsifiable.
Science is encountering impossibilities again.
Not temporary obstacles—fundamental limits on what can be known, tested, observed.
This isn't the first time.
Aristotle couldn't see atoms (no microscopes). Medieval alchemists couldn't see nuclear structure. Victorian physicists couldn't access quantum scales.
Each time, technology eventually overcame limits.
But this time might be different.
Because we're hitting final limits—cosmic horizons, quantum uncertainty, computational irreducibility, thermodynamic constraints.
Some things can't be known. Not "can't yet"—can't ever.
Let's examine what limits science is hitting, why this time is different, where science still has room to grow, and what science becomes when the frontier closes.
THE HISTORICAL PATTERN: Limits, Then Breakthroughs
PAST LIMITS OVERCOME
LIMIT 1: INVISIBLE SMALL (Pre-1600s) ┌─────────────────────────────────────────┐ │ Limit: Can't see cells, bacteria │ │ ↓ │ │ Microscope (1600s) overcomes it │ │ ↓ │ │ Opens: Microbiology, cell theory │ └─────────────────────────────────────────┘
LIMIT 2: INVISIBLE FAR (Pre-1600s) ┌─────────────────────────────────────────┐ │ Limit: Can't see planets' moons, distant│ │ stars │ │ ↓ │ │ Telescope (1600s) overcomes it │ │ ↓ │ │ Opens: Astronomy, cosmology │ └─────────────────────────────────────────┘
LIMIT 3: ATOMIC STRUCTURE (Pre-1900s) ┌─────────────────────────────────────────┐ │ Limit: Can't see inside atoms │ │ ↓ │ │ Particle detectors, quantum theory │ │ overcome it (1900s) │ │ ↓ │ │ Opens: Nuclear physics, quantum │ │ mechanics │ └─────────────────────────────────────────┘
LIMIT 4: SPEED OF COMPUTATION (Pre-1950s) ┌─────────────────────────────────────────┐ │ Limit: Can't calculate complex models │ │ ↓ │ │ Computers (1950s+) overcome it │ │ ↓ │ │ Opens: Climate modeling, genomics, AI │ └─────────────────────────────────────────┘
THE PATTERN: ┌─────────────────────────────────────────┐ │ Each "impossible" limit eventually │ │ overcome by technology │ │ ↓ │ │ Optimistic view: Current limits will │ │ also be overcome │ │ ↓ │ │ But: What if some limits are final? │ └─────────────────────────────────────────┘
Past suggests: Every limit is temporary.
But is that true this time?
THE NEW LIMITS: Why This Time Is Different
FUNDAMENTAL LIMITS (Possibly Final)
LIMIT 1: ENERGY SCALE ┌─────────────────────────────────────────┐ │ PROBLEM: │ │ Testing grand unified theories requires │ │ Planck-scale energies │ │ ↓ │ │ Planck scale: 10^19 GeV │ │ LHC: 10^4 GeV │ │ ↓ │ │ Gap: 15 orders of magnitude │ │ ↓ │ │ REQUIRED COLLIDER: │ │ • Size: Larger than solar system │ │ • Cost: More than global GDP │ │ ↓ │ │ CONCLUSION: Physically impossible │ └─────────────────────────────────────────┘
LIMIT 2: COSMIC HORIZON ┌─────────────────────────────────────────┐ │ PROBLEM: │ │ Can't observe beyond ~46 billion │ │ light-years │ │ ↓ │ │ Universe beyond = causally disconnected │ │ ↓ │ │ Even with perfect instruments, can't see│ │ ↓ │ │ MULTIVERSE THEORIES: │ │ If true, other universes beyond horizon │ │ ↓ │ │ Fundamentally unobservable │ │ ↓ │ │ CONCLUSION: Some cosmology unfalsifiable│ └─────────────────────────────────────────┘
LIMIT 3: QUANTUM MEASUREMENT ┌─────────────────────────────────────────┐ │ PROBLEM: │ │ Heisenberg uncertainty principle │ │ ↓ │ │ Can't measure position and momentum │ │ simultaneously │ │ ↓ │ │ Not technical limitation—fundamental law│ │ ↓ │ │ Some questions literally unanswerable │ │ ↓ │ │ CONCLUSION: Quantum reality partially │ │ unknowable │ └─────────────────────────────────────────┘
LIMIT 4: COMPUTATIONAL IRREDUCIBILITY ┌─────────────────────────────────────────┐ │ PROBLEM (Wolfram): │ │ Some systems can't be predicted without │ │ simulating step-by-step │ │ ↓ │ │ Example: Complex systems, chaotic │ │ dynamics │ │ ↓ │ │ No shortcut to prediction │ │ ↓ │ │ CONCLUSION: Some systems fundamentally │ │ unpredictable │ └─────────────────────────────────────────┘
LIMIT 5: THERMODYNAMICS ┌─────────────────────────────────────────┐ │ PROBLEM: │ │ Information processing requires energy │ │ ↓ │ │ Landauer's principle: Erasing 1 bit │ │ generates heat │ │ ↓ │ │ Limit on computation density │ │ ↓ │ │ CONCLUSION: Physical limit on │ │ computational power │ └─────────────────────────────────────────┘
These aren't engineering challenges.
They're laws of physics.
No technology can overcome them.
CASE STUDY 1: Particle Physics—The Stall
FUNDAMENTAL PHYSICS STAGNATION
THE STANDARD MODEL (1970s): ┌─────────────────────────────────────────┐ │ Completed by mid-1970s │ │ ↓ │ │ Describes: All known particles, three │ │ fundamental forces │ │ ↓ │ │ Incredibly successful: Predictions │ │ confirmed to 10+ decimal places │ │ ↓ │ │ Problem: Incomplete (no gravity, dark │ │ matter, dark energy) │ └─────────────────────────────────────────┘
50 YEARS OF SEARCHING (1970s-2020s): ┌─────────────────────────────────────────┐ │ Built bigger colliders (LEP, Tevatron, │ │ LHC) │ │ ↓ │ │ Searched for: │ │ • Supersymmetry particles (not found) │ │ • Extra dimensions (not found) │ │ • New forces (not found) │ │ • Dark matter particles (not found) │ │ ↓ │ │ ONLY new particle: Higgs boson (2012) │ │ But: Already predicted in 1964 │ │ ↓ │ │ No surprises. No new physics. │ └─────────────────────────────────────────┘
WHY THE STALL? ┌─────────────────────────────────────────┐ │ THEORY 1: NATURAL SCALE │ │ Standard Model works at accessible │ │ energies │ │ ↓ │ │ New physics at much higher energies │ │ (unreachable) │ │ ↓ │ │ THEORY 2: FINE-TUNING │ │ Nature is "accidentally" at scales we │ │ can probe │ │ ↓ │ │ No deeper structure, just fundamental │ │ constants │ │ ↓ │ │ THEORY 3: WRONG APPROACH │ │ Need new theoretical framework, not just│ │ bigger colliders │ └─────────────────────────────────────────┘
THE DILEMMA: ┌─────────────────────────────────────────┐ │ Build next collider? │ │ • Cost: $100B+ (Future Circular Collider│ │ at CERN) │ │ • Guaranteed discovery? No │ │ ↓ │ │ Might find nothing (like LHC for │ │ supersymmetry) │ │ ↓ │ │ Or: Redirect resources to other science │ │ ↓ │ │ But: Abandon frontier of physics │ └─────────────────────────────────────────┘
Particle physics may have hit THE end, not just AN end.
CASE STUDY 2: Cosmology—Beyond the Horizon
COSMOLOGICAL LIMITS
WHAT WE CAN SEE: ┌─────────────────────────────────────────┐ │ Observable universe: ~93 billion light- │ │ years across │ │ ↓ │ │ Contains: ~2 trillion galaxies │ │ ↓ │ │ But: This is tiny fraction of total │ │ universe │ └─────────────────────────────────────────┘
WHAT WE CAN'T SEE: ┌─────────────────────────────────────────┐ │ Beyond cosmic horizon: Light hasn't had │ │ time to reach us │ │ ↓ │ │ Expanding faster than light can travel │ │ ↓ │ │ Fundamentally unobservable │ │ ↓ │ │ No information can cross horizon │ └─────────────────────────────────────────┘
THE MULTIVERSE PROBLEM: ┌─────────────────────────────────────────┐ │ Inflation theory (1980s) suggests: │ │ Our universe = one bubble in multiverse │ │ ↓ │ │ Other universes beyond horizon │ │ ↓ │ │ Each with different physical constants │ │ ↓ │ │ Question: How to test? │ │ ↓ │ │ Answer: Can't. Unfalsifiable. │ └─────────────────────────────────────────┘
IS IT STILL SCIENCE? ┌─────────────────────────────────────────┐ │ Multiverse can't be observed or tested │ │ ↓ │ │ But: Follows logically from tested │ │ theories (inflation) │ │ ↓ │ │ Debate: │ │ • PRO: Mathematical consequence of good │ │ theory │ │ • CON: Unfalsifiable = not science │ │ ↓ │ │ Science reaching limits of empiricism │ └─────────────────────────────────────────┘
Some cosmological questions are literally unanswerable.
CASE STUDY 3: Complexity—Irreducible Systems
COMPUTATIONAL LIMITS ON PREDICTION
THE PROBLEM: ┌─────────────────────────────────────────┐ │ Some systems: │ │ • Weather beyond ~10 days │ │ • Ecosystems │ │ • Brains │ │ • Economies │ │ ↓ │ │ Are "computationally irreducible" │ │ (Wolfram) │ └─────────────────────────────────────────┘
WHAT IRREDUCIBILITY MEANS: ┌─────────────────────────────────────────┐ │ Can't shortcut to prediction │ │ ↓ │ │ Only way to know future state: Simulate │ │ every step │ │ ↓ │ │ But: Simulation takes same time as │ │ reality │ │ ↓ │ │ No predictive advantage │ └─────────────────────────────────────────┘
IMPLICATIONS FOR SCIENCE: ┌─────────────────────────────────────────┐ │ Some systems fundamentally unpredictable│ │ ↓ │ │ Not due to insufficient data or │ │ computing power │ │ ↓ │ │ Due to nature of system itself │ │ ↓ │ │ Science's goal (predict) impossible here│ └─────────────────────────────────────────┘
EXAMPLE: WEATHER ┌─────────────────────────────────────────┐ │ Chaotic system │ │ ↓ │ │ Tiny measurement errors → huge forecast │ │ errors │ │ ↓ │ │ 10-day limit is fundamental, not │ │ technological │ │ ↓ │ │ Better computers won't fix │ └─────────────────────────────────────────┘
EXAMPLE: CONSCIOUSNESS ┌─────────────────────────────────────────┐ │ Brain = 86 billion neurons, 100 trillion│ │ connections │ │ ↓ │ │ Simulating requires atom-by-atom model │ │ ↓ │ │ Computational cost: Astronomical │ │ ↓ │ │ May never fully simulate consciousness │ └─────────────────────────────────────────┘
Some scientific questions have no computable answers.
WHERE SCIENCE STILL HAS ROOM
OPEN FRONTIERS (Not Yet Limited)
FRONTIER 1: BIOLOGY ┌─────────────────────────────────────────┐ │ Still discovering: │ │ • New species (millions unknown) │ │ • Genetic mechanisms │ │ • Proteomics, metabolomics │ │ • Brain function │ │ ↓ │ │ No fundamental limit in sight │ │ ↓ │ │ Just immense complexity │ └─────────────────────────────────────────┘
FRONTIER 2: MEDICINE ┌─────────────────────────────────────────┐ │ Still developing: │ │ • Cures for diseases (cancer, Alzheimer)│ │ • Gene therapies │ │ • Personalized medicine │ │ • Aging interventions │ │ ↓ │ │ Decades of work remaining │ └─────────────────────────────────────────┘
FRONTIER 3: MATERIALS SCIENCE ┌─────────────────────────────────────────┐ │ Still inventing: │ │ • New materials (metamaterials, │ │ superconductors) │ │ • AI-designed molecules │ │ • Quantum materials │ │ ↓ │ │ Combinatorial space vast │ └─────────────────────────────────────────┘
FRONTIER 4: EARTH SYSTEMS ┌─────────────────────────────────────────┐ │ Still understanding: │ │ • Climate dynamics │ │ • Biodiversity │ │ • Ocean currents │ │ • Geological processes │ │ ↓ │ │ Our own planet not fully mapped │ └─────────────────────────────────────────┘
FRONTIER 5: APPLIED PHYSICS ┌─────────────────────────────────────────┐ │ Engineering applications: │ │ • Fusion energy │ │ • Quantum computing │ │ • Room-temperature superconductors │ │ ↓ │ │ Uses known physics, but hasn't achieved │ │ goals │ └─────────────────────────────────────────┘
Most science still has frontiers.
Fundamental physics is unusual in hitting limits.
THE PHILOSOPHICAL SHIFT: From Discovery to Engineering
CHANGING NATURE OF SCIENCE
OLD SCIENCE (Discovery): ┌─────────────────────────────────────────┐ │ Find new laws │ │ ↓ │ │ Discover new particles │ │ ↓ │ │ Uncover nature's secrets │ │ ↓ │ │ Goal: Understand reality │ └─────────────────────────────────────────┘
NEW SCIENCE (Engineering): ┌─────────────────────────────────────────┐ │ Apply known laws │ │ ↓ │ │ Design new materials, drugs, systems │ │ ↓ │ │ Optimize existing knowledge │ │ ↓ │ │ Goal: Solve problems │ └─────────────────────────────────────────┘
THE SHIFT: ┌─────────────────────────────────────────┐ │ As fundamental discoveries slow: │ │ ↓ │ │ Science becomes more applied │ │ ↓ │ │ Less "What is reality?" more "How can │ │ we use what we know?" │ │ ↓ │ │ Still valuable, but different │ └─────────────────────────────────────────┘
IS THIS BAD? ┌─────────────────────────────────────────┐ │ PESSIMISTIC VIEW: │ │ Science losing its sense of wonder │ │ ↓ │ │ Becoming mere technology │ │ ↓ │ │ OPTIMISTIC VIEW: │ │ Science maturing │ │ ↓ │ │ Adolescent discovery → adult application│ │ ↓ │ │ Engineering can still be profound │ └─────────────────────────────────────────┘
The age of discovery may be ending.
The age of application continues.
THE ECONOMIC PROBLEM: Diminishing Returns
COST VS. BENEFIT TRENDS
THE RISING COST OF DISCOVERY: ┌─────────────────────────────────────────┐ │ Each breakthrough costs more: │ │ ↓ │ │ Early 1900s: Individual scientists │ │ • Cost: Minimal │ │ • Discoveries: Quantum mechanics, │ │ relativity │ │ ↓ │ │ Mid-1900s: Big Science │ │ • Manhattan Project: $2B (1940s dollars)│ │ • Discoveries: Nuclear physics │ │ ↓ │ │ Late 1900s: Huge Science │ │ • LHC: $10B │ │ • Discoveries: Higgs (predicted) │ │ ↓ │ │ 2020s+: Impossible Science? │ │ • Next collider: $100B+ │ │ • Discoveries: ??? │ └─────────────────────────────────────────┘
BLOOM ET AL. (2020) STUDY: ┌─────────────────────────────────────────┐ │ Research productivity declining │ │ ↓ │ │ Constant number of ideas per year │ │ requires exponentially more researchers │ │ ↓ │ │ "Are ideas getting harder to find?" │ │ ↓ │ │ Yes. Significantly. │ └─────────────────────────────────────────┘
THE DILEMMA: ┌─────────────────────────────────────────┐ │ Society willing to fund science while │ │ returns are high │ │ ↓ │ │ As returns diminish, funding pressured │ │ ↓ │ │ Harder to justify $100B for uncertain │ │ payoff │ │ ↓ │ │ Science may face funding limits even │ │ before physical limits │ └─────────────────────────────────────────┘
We may stop doing science not because it's impossible, but because it's too expensive.
WHAT SCIENCE BECOMES AT THE FRONTIER
POST-DISCOVERY SCIENCE
SCENARIO 1: PERMANENT PLATEAU ┌─────────────────────────────────────────┐ │ Fundamental physics complete │ │ ↓ │ │ Standard Model + gravity = final theory │ │ ↓ │ │ No more deep discoveries │ │ ↓ │ │ Science = applying known laws │ └─────────────────────────────────────────┘
SCENARIO 2: NEW PARADIGM ┌─────────────────────────────────────────┐ │ Current theories wrong │ │ ↓ │ │ Breakthrough requires reconceptualizing │ │ ↓ │ │ Not bigger colliders, but new ideas │ │ ↓ │ │ Historical examples: Relativity, │ │ quantum mechanics came from thought, │ │ not just data │ └─────────────────────────────────────────┘
SCENARIO 3: AI SCIENTISTS ┌─────────────────────────────────────────┐ │ Humans can't go further, but AI can │ │ ↓ │ │ AI finds patterns, makes theories humans│ │ can't comprehend │ │ ↓ │ │ Science continues, but beyond human │ │ understanding │ └─────────────────────────────────────────┘
SCENARIO 4: PRACTICAL FOCUS ┌─────────────────────────────────────────┐ │ Redirect resources to applied problems: │ │ • Climate change │ │ • Diseases │ │ • Energy │ │ ↓ │ │ Fundamental questions take backseat │ │ ↓ │ │ Science = technology development │ └─────────────────────────────────────────┘
The frontier may be closing. Or shifting. Or transcending human reach.
CONCLUSION: The End of Easy Science
Science has always been hard. But for 400 years, there was always somewhere further to go.
Bigger telescopes revealed more distant stars. Stronger microscopes revealed smaller structures. Higher energies revealed new particles.
The universe kept giving up secrets.
Now we're hitting fundamental walls:
Physical walls: Can't build colliders large enough to test grand unified theories.
Observational walls: Can't see beyond cosmic horizon. Multiverse unfalsifiable.
Computational walls: Some systems irreducibly complex. Can't predict without simulating.
Economic walls: Discoveries cost exponentially more. Society may stop funding.
These aren't temporary obstacles. They're final limits.
Not everything is knowable. Not everything is testable. Not everything is computable.
The Enlightenment dream—that science could eventually understand everything—may be false.
But this doesn't mean science is over.
Biology, medicine, materials, Earth systems, engineering—all have vast unexplored territory.
AI might take science beyond human comprehension.
Applied science can still transform civilization.
The shift:
From "What is reality?" to "How can we use what we know?"
From discovery to application.
From frontier expansion to depth exploration.
This is the maturation of science.
The hardening of science created the greatest explosion of knowledge in history.
But every explosion has a boundary.
We may have reached it.
Or we may be on the cusp of a new paradigm that makes our current limits seem quaint.
History suggests: Every previous generation that thought it was near the end was wrong.
But history also suggests: Exponential growth always ends.
And we might be the generation that finds out which.
[Cross-references: For particle physics specifics, see Physics Companion #80-85. For cosmological horizon and multiverse, see Physics Companion #88-90. For complexity and chaos, see Mathematics Companion #142-144. For diminishing research productivity, see "When Funding Shaped Questions: Science as Investment" (Core #43). For AI potentially transcending limits, see "When AI Started Doing Science: Machines as Researchers" (Core #46). For what comes after these limits, see "What Comes After Falsification? New Epistemologies" (Core #48) and "The Future of Hardening" (Core #50).]