Stephen Hawking black hole paradox has been running, unresolved, for exactly fifty years now, and the strangest part of researching this piece wasn’t the physics. It was discovering that the man who created the paradox eventually conceded he’d been wrong about it — publicly, in front of an audience, with a baseball encyclopedia changing hands as a wager payout — and the argument still isn’t actually settled, even after his own concession.
I went into this expecting a tidy story. Brilliant physicist makes a discovery, discovery causes decades of confusion, resolution eventually arrives, everyone moves on. What I found instead was something closer to a fifty-year argument that occasionally produces a headline announcing it’s “finally solved,” only for the solution to generate a new, more specific version of the same disagreement a few years later. As I’m writing this, in 2026, there’s a paper proposing black holes might never fully evaporate at all, provided the universe secretly has seven dimensions instead of four. That’s not a resolution. That’s the argument putting on a new coat.
I find something genuinely disorienting about a scientific dispute this durable, and I think sitting with why it’s lasted this long is more interesting than trying to pretend there’s a clean answer.
The Discovery That Broke Its Own Rules
To really understand the Stephen Hawking black hole paradox, you first need to understand what black holes were supposed to be, before Hawking complicated things.
For most of the 20th century, a black hole was understood as the simplest possible object in the universe — a region where gravity is so intense that nothing, not even light, can escape once it crosses the boundary known as the event horizon. Classically, a black hole could be fully described using just three numbers: its mass, its electric charge, and its rotation. Everything else about whatever fell in — its composition, its structure, its history — was thought to be erased completely, collapsed down into that handful of properties. Physicists half-jokingly called this idea “no hair,” meaning black holes retained no distinguishing details about what they’d consumed.
In the early 1970s, Jacob Bekenstein, building on this picture, proposed something that initially struck many physicists, including Hawking, as slightly absurd: that black holes should have entropy — a measure of disorder — proportional to the surface area of their event horizon. Hawking initially pushed back against this idea. Then, in 1974, working through the implications more carefully, he discovered something that would upend his own skepticism and reshape the field for the next half-century.
Applying quantum theory to the region around a black hole’s event horizon, Hawking found that pairs of particles should constantly be popping in and out of existence there, as quantum fluctuations do throughout empty space. Occasionally, one particle of a pair would fall into the black hole while its partner escaped outward. The result: black holes weren’t actually perfectly black. They emitted a faint, thermal glow — now called Hawking radiation — and in doing so, they slowly lost mass and eventually, given enough time, evaporated away entirely.
This should have been a triumph, and in most respects it was. But it created an entirely new problem, one considerably harder to live with than the one it solved.
The Paradox in Plain Terms
Here’s the core tension, stripped down as much as I can manage.
Quantum mechanics has a rule — technically called unitarity — that information can never actually be destroyed. It can be scrambled beyond any practical hope of reconstruction, hidden in overwhelming complexity, rendered functionally unrecoverable in every way that matters to an observer, but in principle, the underlying information describing a physical system is supposed to persist. Burn a book to ashes, and the information about what was written on its pages hasn’t vanished from the universe — it’s just been scattered into smoke, heat, and soot in a way so complex that reconstructing the original text is beyond any practical technology. The information is still, technically, there.
Hawking’s calculation suggested something different for black holes. If a black hole evaporates completely via Hawking radiation, and that radiation is purely thermal — meaning it depends only on the black hole’s mass, charge, and spin, exactly as the “no hair” picture predicted — then whatever specific, detailed information fell into the black hole over its lifetime seems to be gone. Not scrambled. Not hidden. Actually, fundamentally erased from the universe, once the black hole itself has fully evaporated. That directly violates the “information can’t be destroyed” rule at the foundation of quantum theory.
So physicists were left holding two theories that had each been extraordinarily well-tested in their own domains — general relativity, which produced the picture of black holes evaporating this way, and quantum mechanics, which insisted information loss like this shouldn’t be possible — and those two theories were flatly contradicting each other at exactly the point where a black hole finished evaporating.
I find something almost uncomfortably familiar in this shape of problem, from years of debugging complex systems. Two subsystems, each individually well-tested, each behaving correctly according to its own internal logic, producing a genuine contradiction only when you connect them at a specific boundary condition neither one was really designed to handle gracefully. The bug isn’t in either component. It’s in the interface between them — the exact place where black hole evaporation happens to sit, wedged uncomfortably between the domain where relativity works and the domain where quantum mechanics works, with nobody yet holding a complete theory that covers both simultaneously.
A Bet, a Baseball Encyclopedia, and a Concession
What I didn’t expect, going into this research, was how much of the paradox’s fifty-year history has played out through actual, formal wagers between physicists — and I mean that literally, with signed documents and agreed-upon prizes, not just metaphorical scientific disagreement.
In 1997, Hawking and Caltech physicist Kip Thorne bet fellow Caltech theorist John Preskill on the resolution of the paradox. Hawking and Thorne staked their position that information really is destroyed when a black hole evaporates. Preskill bet the opposite — that information somehow, eventually, gets out, even if physicists couldn’t yet explain the precise mechanism. The formal wager was signed, with actual signatures, on February 6, 1997, and the loser owed the winner an encyclopedia, “from which information can be recovered at will.”
For seven years, the bet sat unresolved. Then, in July 2004, at a conference in Dublin, Hawking stood up and conceded. He’d become convinced that black hole horizons should fluctuate in ways that allow information to leak out gradually, embedded within the details of the Hawking radiation itself, rather than being destroyed outright. He presented Preskill with a baseball encyclopedia as the agreed prize.
I find this detail almost unbearably charming, and I don’t think “charming” is a word I use often about theoretical physics. A dispute about the fundamental fate of information in the universe, resolved with the same formal structure as a friendly bet over a sports game, settled by handing over a book about baseball statistics. Hawking apparently kept his sense of humor about it too — he later joked that comparing black hole information to an encyclopedia’s worth of scrambled data was accurate enough that maybe he should have just handed Preskill the ashes instead.
Here’s the detail that genuinely surprised me, though, and the one I think matters most for understanding why this story doesn’t have a tidy ending: Kip Thorne never conceded. As of the most recent records I could find, more than two decades after Hawking changed his mind, Thorne still hasn’t accepted that he lost. Hawking’s own scientific collaborator and co-bettor looked at the same argument that convinced Hawking himself, and simply wasn’t persuaded.
I sat with that fact for a while, because it tells you something important about the actual epistemic status of this “resolution.” It wasn’t a proof in the mathematical sense, the kind where every competent expert who checks the work has to agree. It was an argument compelling enough to change one brilliant physicist’s mind, while leaving another brilliant physicist, who’d spent decades studying the exact same problem, unmoved. That’s not how a settled scientific question is supposed to behave. It’s how an ongoing, genuinely contested one behaves.
The 2019 Breakthrough That Wasn’t Quite the End
For years after Hawking’s concession, the field searched for something more rigorous than a persuasive argument — an actual calculation demonstrating how information could escape a black hole, using the tools of quantum gravity itself.
That calculation arrived, in significant part, in 2019, through what’s now called the Page curve, developed using a technique called “replica wormholes” by researchers including Geoff Penington and Ahmed Almheiri. Without getting lost in the mathematical machinery, the core achievement was this: physicists finally produced calculations, grounded in the actual physics rather than persuasive argument alone, showing that the entropy of Hawking radiation follows a specific curve over time that’s consistent with information ultimately being preserved rather than destroyed — supporting the side of the bet Preskill had taken all along.
Penington himself described the moment with a specific kind of restraint that I find more convincing than triumphant language would have been: he called it, to some degree, the end of a revolution rather than the beginning of one. Not “we finally understand everything about black holes.” Something more measured — we finally have calculational tools sophisticated enough to demonstrate what many physicists had already suspected was true.
Even this achievement, though, comes with an asterisk that I think gets lost whenever headlines describe the paradox as “solved.” The 2019 work uses simplified models — toy versions of black holes, not the messy, complicated real objects sitting at the centers of actual galaxies. Whether the same mathematical machinery holds up cleanly for genuinely realistic black holes remains, formally, an open question. The framework is now the field’s leading approach. It is not universally regarded as the final word.
What’s Still Being Argued in 2026
This is the part of the story that made me want to write this piece in the first place, because it means the paradox isn’t safely tucked away in physics history — it’s still actively, currently unresolved.
As recently as this year, researchers have proposed that black holes might never fully evaporate at all, leaving behind stable remnants that store the information they once consumed — provided reality includes additional hidden spatial dimensions beyond the four we experience directly. Other very recent work has focused on the mathematical structure connecting different physical states after evaporation, still working out exactly how the underlying mechanism operates even within frameworks that already assume information is preserved. Meanwhile, entirely separate strands of black hole research — how supermassive black holes formed so quickly after the Big Bang, what mechanisms allow black holes to grow beyond expected mass limits — continue to produce new findings on a near-monthly basis, each one interacting in some way with the broader theoretical picture Hawking’s original 1974 calculation forced physicists to construct.
Fifty years. A conceded bet. A baseball encyclopedia. A landmark calculation in 2019 that moved the field forward without fully closing the case. And now, in 2026, entirely new proposed resolutions involving extra dimensions of space that didn’t exist as serious proposals when Hawking first sat down to do the math.
I think about the sheer duration of this more than I probably should, as someone used to problems that get resolved, even painfully, within a project timeline measured in months rather than half-centuries. There’s something almost humbling about a genuinely brilliant community of specialists working continuously on the same core question for fifty straight years without full closure — not because they’re incompetent or the problem is somehow fake, but because the problem sits exactly at the seam between two of the best-tested theories human beings have ever constructed, and that seam turns out to be extraordinarily difficult terrain.
A Thought to Leave You With
What stays with me most, after sitting with this story for a while, isn’t the physics itself, even though the physics is genuinely remarkable. It’s Kip Thorne’s continued refusal to concede.
We tend to want scientific disputes to resolve the way a debugging session resolves — you find the actual bug, you fix it, the tests pass, everyone moves on to the next problem, case closed. Some disputes really do work that way. This one, five decades in, still doesn’t. Two people who worked the same problem for years, who trusted each other’s judgment enough to sign a formal wager together as partners on the same side, ended up in different places once the evidence started shifting. One updated his belief. One didn’t. Both remain, as far as I can tell, serious and respected physicists.
I find something worth sitting with in that, beyond the specific physics of evaporating black holes. Genuine expertise doesn’t guarantee convergence, even among people examining literally the same evidence, even among close collaborators who trust each other enough to shake hands on a bet. Sometimes the honest state of a hard problem, even after fifty years of the best minds available working on it continuously, is a persuasive majority position, a stubborn and equally credentialed holdout, and a new paper every few months proposing yet another way the whole picture might need to be reconsidered.
Hawking spent much of his final years still trying to work out the precise mechanism by which information actually escapes, even after conceding the bet in principle. He never got to see the field settle the way settled scientific questions eventually do. Neither, as of this year, has anyone else. The information paradox outlived the man who discovered it, and it’s still, quietly, unfinished business.
More Stories Like This
← Previous: Columbus’s Map Error: How Bad Math Accidentally Led to the New World in 1492
→ Next: The First Computer Bug Was an Actual Bug: The Moth in Grace Hopper’s Logbook (Coming soon)