Argument

Quantum Physics Reverses Entropy Objection Defeater

Intro

The objection runs like this. The Argument from Thermodynamics says the universe cannot be infinitely old, because entropy, the spread-out, run-down state of energy, always increases, and an infinitely old universe would already be run down. But quantum physics shows entropy can go backwards. Scientists have made heat flow from cold to hot and even "reversed time" on a quantum computer. So entropy is not a one-way street, and the universe could be eternal after all.

The experiments are real, and they are fascinating. Here is what they actually show.

In a tiny, carefully prepared system, a few atoms or a few qubits, the usual direction can be pushed backwards for a moment. But every time, something pays for it: energy pumped in by the experimenters, information they use and then have to erase, or special links between particles (correlations) that they set up beforehand and that get used up. Count everything, including what was spent, and total entropy still goes up. The scientists who ran the experiments say this themselves.

A refrigerator is the everyday version. The inside gets colder, which looks backwards, but the back of the fridge dumps more heat into your kitchen. Quantum experiments are very refined refrigerators.

And the size matters. Moving a few atoms backwards for a split second, at great effort, says nothing about whether a whole universe could keep winding itself back up forever with nobody doing the work.

◎ See this page's graph and discover linked topics →

In full

The objection argues: (1) the thermodynamic argument for a finite past assumes entropy never decreases; (2) quantum experiments show entropy can decrease; (3) so the argument fails and a past-eternal universe remains open.

The defeater runs in five moves:

• Grant the results (P1). Fluctuation theorems, the 2019 heat-reversal experiment and the 2019 quantum-computer "time reversal" are genuine, peer-reviewed findings.
• Every reversal is paid for (P2). In each case a resource is consumed: external work, information that must later be erased (Landauer), or prepared correlations. With the resource counted, total entropy increases, and the experimenters say the generalized second law holds.
• Local is not global (P3). Entropy decreases in small, open, prepared systems are exported or paid for elsewhere. The universe as a whole has nowhere to export to.
• The arrow comes from the beginning, not from the laws (P4). The underlying quantum laws are time-symmetric, which is exactly why physicists trace the arrow of time to an extraordinarily low-entropy initial state. That makes the beginning more special, not less.
• At cosmic scale the escape fails (P5). Cyclic and fluctuation models of an eternal universe either carry entropy forward from cycle to cycle (Tolman) or predict a universe of random momentary observers rather than ours (the Boltzmann-brain problem).

Argument structure

Form

Defensive. The objection is a factual claim with a real basis, so the reply concedes the experiments and contests what they mean. The key move is resource accounting: ask what was spent to produce the reversal.

Cheatsheet

The 30-second reply (read aloud). Those experiments are real, and I love them. But look at what they actually did. Scientists took a few atoms, or a two-qubit computer, set them up very carefully, and pushed them backwards for a moment. Every time, they paid for it, with energy they put in, or with special links between the particles that they built first and then used up. Count what they spent and total entropy still went up. The researchers say so in their own papers. It is the same as your fridge: cold inside, but hotter in your kitchen. And a few atoms going backwards in a lab, with scientists doing all the work, tells us nothing about a whole universe winding itself back up forever with nobody doing anything.

The 5 fast facts:

1. Heat ran from cold to hot, using up correlations. Micadei and colleagues (Nature Communications, 2019) used the nuclear spins of carbon and hydrogen atoms in a chloroform molecule. When the spins started out correlated, heat flowed from the colder to the hotter one. The correlations were consumed, and the authors show the result is consistent with a second law that counts correlations.
2. The "time reversal" was a program. Lesovik and colleagues (Scientific Reports, 2019) returned a two- or three-qubit state toward its earlier condition on an IBM quantum computer by running a deliberately designed operation. They also estimated that the same thing happening on its own in nature, even for a single electron, is vanishingly unlikely, and their success rate fell as qubits were added.
3. Fluctuation theorems say small reversals are exponentially rare. For tiny systems over short times, entropy-decreasing paths occur, but they get exponentially less likely as the system grows or time lengthens (Evans and Searles; tested by Wang et al., Physical Review Letters, 2002).
4. Information has an entropy price. Landauer (1961) showed that erasing one bit of information releases a minimum amount of heat, and Bérut and colleagues measured it (Nature, 2012). A "demon" that sorts molecules to lower entropy has to erase its memory eventually, and that pays the bill.
5. Quantum events usually drive the arrow, not reverse it. Decoherence, the process by which quantum states spread into their environment, is itself entropy-increasing. Most quantum events in nature push the usual direction.

The 3 strongest counter-moves:

• "What did they spend?" For each experiment, ask what resource was used to produce the reversal. There is always an answer.
• "How big, and for how long?" A few atoms for milliseconds. Ask how that scales to 10^80 particles for an infinite past.
• "Who is running the experiment for the universe?" Every lab reversal needed someone to prepare it and pay for it. An eternal universe would need the same, forever, with no one there.

Reciprocal concessions (grant the small point, then collect a bigger one):

• Grant: entropy can decrease locally, even at the quantum level. Now collect: it always does so by drawing on a resource and raising entropy elsewhere. The universe has no "elsewhere," so the local exception is not a cosmic one.
• Grant: the basic laws of quantum mechanics run the same forwards and backwards. Now collect: then the one-way arrow we see everywhere has to come from the starting conditions, and Roger Penrose calculated how special those were, roughly 1 in 10^(10^123). The time-symmetry of the laws makes the beginning more remarkable, not less.
• Grant: nobody yet has a complete theory of quantum gravity. Now collect: then "quantum physics might allow an eternal universe" is a hope about physics we do not have, while the second law is the best-tested principle we do have.

The closing line. Scientists can run a few atoms backwards for a moment if they pay for it. A universe that has run down for an infinite past with no one paying would need to do that forever, everywhere, for free.

P1, Grant the experiments

Affirmative case

1. Fluctuation theorems. Denis Evans, Debra Searles and colleagues showed in the 1990s that in small systems observed over short times, trajectories that decrease entropy occur, with a probability that falls off exponentially relative to entropy-increasing ones. Wang and colleagues confirmed it experimentally with a microscopic bead in an optical trap (Physical Review Letters 89, 2002).
2. Heat flowing from cold to hot. Micadei and colleagues (Nature Communications 10, 2019) prepared two nuclear spins in a chloroform molecule with initial quantum correlations. Heat then flowed from the colder spin to the hotter one.
3. Quantum-computer "time reversal." Lesovik and colleagues (Scientific Reports 9, 2019) evolved a simple qubit state forward and then applied a designed operation that returned it close to its starting state.
4. These are good science. None of this is fringe. The objection deserves a straight answer, not a dismissal.

Anticipated objections

1. "So you admit the second law is broken."

Rebuttals

1. No, and neither do the authors. The second law in its statistical form already predicts small, brief fluctuations; the fluctuation theorems quantify them. The heat-reversal paper presents its result as consistent with a generalized second law that accounts for correlations. "Entropy can locally decrease when paid for" is a statement of the second law, not a refutation of it.

P2, Every reversal is paid for

Affirmative case

1. Correlations are a resource. In the Micadei experiment, the reversed heat flow lasted only as long as the initial correlations, and it used them up. Preparing those correlations required work. Counting the correlations as part of the entropy budget, the total did not decrease.
2. The time reversal was driven from outside. In the Lesovik experiment, the reversal was an operation the experimenters programmed and ran. The quantum computer, its cooling system and its control electronics generated far more entropy than the qubits lost.
3. Information has a price. A Maxwell's-demon arrangement that lowers entropy by sorting particles must record what it sees. Landauer's principle (1961) says erasing that record releases at least a fixed minimum of heat per bit, which Bérut and colleagues measured (Nature 483, 2012). The demon's savings are repaid when its memory is cleared.
4. So the pattern is constant. Local reversal, external resource, net increase. The experiments are demonstrations of how precisely the accounting works.

Anticipated objections

1. "The universe could have its own correlations or resources that let it run backwards."

Rebuttals

1. Then name them and show they regenerate. A finite stock of correlations is used up as it is spent, which is the point of the experiment. To keep an eternal universe from running down, the resource would have to renew itself forever without an entropy cost, which is the very thing the second law forbids.

P3, Local is not global

Affirmative case

1. Every lab system is open. It exchanges energy and information with apparatus, cooling and experimenters.
2. The universe is not. By definition it has no outside to dump heat into or to draw work from. The Argument from Thermodynamics is about that total system.
3. The refrigerator is the right picture. Cold inside, hotter outside, net increase. Quantum experiments are refrigerators run on a very small scale with very fine control.

Anticipated objections

1. "Maybe the universe is part of a larger system it exports entropy to."

Rebuttals

1. Then the larger system is the universe in the relevant sense, and the same question applies to it. Moving the boundary does not remove the problem; it relocates it. See Borde-Guth-Vilenkin Theorem for why expanding spacetimes, including those embedded in larger ones, are not past-eternal.

P4, The arrow comes from the beginning

Affirmative case

1. The laws are time-symmetric. The fundamental equations of quantum mechanics, like those of classical mechanics, run the same in both directions. This is the old puzzle Josef Loschmidt raised against Boltzmann in 1876.
2. So the one-way arrow needs a special start. Physicists explain the observed arrow by an extremely low-entropy initial state of the universe, sometimes called the "past hypothesis." Roger Penrose estimated the improbability of that initial state at roughly 1 in 10^(10^123). See Second Law of Thermodynamics.
3. That cuts against the objection. If entropy increase is not written into the laws but into the starting conditions, then the question "why was the beginning so ordered?" becomes sharper, and a past-eternal universe has no beginning to put that order into.

Anticipated objections

1. "Time-symmetric laws mean the universe could run backwards."

Rebuttals

1. In principle, and with vanishing probability in practice. Reversing a macroscopic system requires reversing the motion of every particle exactly. Lesovik and colleagues estimated that even a single electron spontaneously retracing its spread is effectively never observed in nature. For a universe the odds are beyond any meaningful number.

P5, At cosmic scale the escape fails

Affirmative case

1. Cyclic models carry entropy forward. Richard Tolman showed in 1934 that in an oscillating universe, entropy built up in one cycle carries into the next, making each cycle larger and longer. Run backwards, the cycles shrink toward a first one.
2. Fluctuation models predict the wrong universe. If an eternal universe produced our low-entropy state by a random fluctuation, it would be vastly more likely to produce the smallest fluctuation compatible with an observer, a lone momentary brain, than a whole ordered cosmos 13.8 billion years old. Physicists including Sean Carroll (From Eternity to Here, 2010) treat this "Boltzmann brain" result as a reason to reject such models. See Boltzmann Brain.
3. Quantum vacuum proposals are not "nothing" and not eternal by default. See Quantum Fluctuations Are Uncaused Objection Defeater.

Anticipated objections

1. "Penrose's conformal cyclic cosmology resets entropy every cycle."

Rebuttals

1. It does so by a contested assumption, not by quantum reversal. Penrose's model has to treat the far future of one cycle as physically equivalent to the beginning of the next, which requires information and mass to effectively disappear. Whatever its merits, it is not a case of quantum events reversing entropy, and it is not the standard model of cosmology.

Master objections to the defeater

MO1: "You keep saying 'count everything.' That's an unfalsifiable bookkeeping trick."

• It is falsifiable: show a closed system in which total entropy, including correlations and erased information, goes down and stays down. No experiment has done so, and the experiments the objection cites were designed and reported within that accounting.

MO2: "Quantum mechanics is weird, so you can't say what's impossible."

• Quantum mechanics is the most precisely tested theory in physics, and its predictions are definite. "It's weird" is not a result. The specific claim in dispute has been tested, and it came out the way the second law says.

Honest limits

• The relationship between quantum mechanics and thermodynamics is an active research field, and some foundational questions (for example, how to define entropy for quantum states far from equilibrium) are open.
• The thermodynamic argument for a beginning is stronger in combination with the Borde-Guth-Vilenkin Theorem and the Kalam Cosmological Argument than on its own.
• Penrose's figure is an estimate built on assumptions about how to count states; it signals scale, not a precise value.

Tactical opening / closing

• Opening. "Those experiments are real, and they're brilliant. Let's look at what they actually spent to do it."
• Closing. "A few atoms ran backwards because scientists paid for it. Your eternal universe needs that to happen everywhere, forever, for free."

Connection to Scripture

• "They shall perish, but thou shalt endure; Yea, all of them shall wax old like a garment" (Psalm 102:26, ASV). The psalm treats the created order as something that wears out, against an unchanging Creator.
• "The creation itself also shall be delivered from the bondage of corruption into the liberty of the glory of the children of God" (Romans 8:21, ASV). Paul describes creation as subject to decay and awaiting renewal from outside itself, not by its own winding back.
• These are not claims about physics, and they should not be read as predictions of the second law. They do describe a cosmos that runs down and depends on God for renewal, which is the picture the thermodynamic evidence fits.

Live-cite kit

Scientific:

• Micadei et al., "Reversing the direction of heat flow using quantum correlations," Nature Communications 10 (2019)
• Lesovik et al., "Arrow of time and its reversal on the IBM quantum computer," Scientific Reports 9 (2019)
• Wang et al., experimental demonstration of the fluctuation theorem, Physical Review Letters 89 (2002)
• Bérut et al., experimental verification of Landauer's principle, Nature 483 (2012)
• Richard Tolman, Relativity, Thermodynamics and Cosmology (1934)
• Sean Carroll, From Eternity to Here (2010)

Aphorism:

• "Entropy can go backwards in the lab the same way your fridge gets cold: somebody pays the bill."
• "The laws run both ways. The universe runs one way. That means the beginning was special."

Scholarship

Published academic works this page draws on, each matched to a publisher, index or library catalog record.

• Sean M. Carroll, From eternity to here (Dutton, 2009). Open Library
• G. B. Lesovik et al., "Arrow of time and its reversal on the IBM quantum computer," Scientific Reports 9 (2019). doi:10.1038/s41598-019-40765-6 PubMed 30867496 open access
• Kaonan Micadei et al., "Reversing the direction of heat flow using quantum correlations," Nature Communications 10 (2019). doi:10.1038/s41467-019-10333-7 PubMed 31165732 open access
• Richard Chace Tolman, Relativity, thermodynamics and cosmology (1934). Open Library

See also

• Argument from Thermodynamics, the argument this objection targets
• Second Law of Thermodynamics, the law itself and Penrose's low-entropy beginning
• Boltzmann Brain, why fluctuation models predict the wrong universe
• Quantum Fluctuations Are Uncaused Objection Defeater, the companion quantum objection
• Borde-Guth-Vilenkin Theorem, why expanding universes are not past-eternal
• Kalam Cosmological Argument, the broader case for a beginning
• Big Bang, the observed beginning
• Sean Carroll, cited for the Boltzmann-brain argument
• Atheism, master hub