IBM Just Made a Very Expensive Bet on Not Knowing the Answer
This week, IBM announced plans to acquire HRL Laboratories, a storied research institution jointly owned by Boeing and General Motors.
The headlines quickly reached for the familiar language reserved for major quantum announcements: breakthrough. Game-changer. A new front in the race.
And the deal is significant. HRL has one of the world’s leading teams working on silicon-spin qubits—a fundamentally different way of building quantum computers from the superconducting technology IBM has spent years developing. Once the acquisition closes, IBM will be pursuing both approaches in parallel.
But the most interesting thing about the acquisition is not that IBM has identified the technology that will win the quantum race.
It is that IBM appears increasingly unwilling to bet its entire future on there being a single winner at all.
IBM did not buy the answer.
IBM already builds quantum computers using superconducting qubits. HRL is one of the strongest teams developing a different approach: silicon-spin qubits.
IBM is not abandoning its current technology. It is acquiring a credible second path—and the people, manufacturing knowledge and engineering capabilities needed to pursue it seriously.
IBM did not buy a finished solution. It bought another promising way to build one.
That may be the more important signal.
the “explain it to me like I’m five” version
A quantum computer needs qubits.
Qubits are the quantum version of the bits inside an ordinary computer—but scientists have not agreed on the best physical thing to use as a qubit.
Imagine that everyone agrees we need to build a new kind of car, but nobody agrees whether the engine should run on electricity, hydrogen, solar power or something that has not been perfected yet.
IBM has spent years building one kind of quantum engine.
HRL Laboratories has been developing another.
IBM is not throwing away the engine it already has. It is buying one of the best teams working on a different option, so it does not have to depend entirely on its first choice being the only right answer.
That is the simplest version of why the acquisition matters.
IBM did not buy a finished quantum computer that suddenly solves everything.
It bought another promising way to build one.
Now, the slightly more technical version
A qubit is a unit of quantum information, just as a bit is a unit of information in an ordinary computer. But the word “qubit” does not tell you what the qubit is physically made from.
Unlike the standardized bits in conventional computers, qubits can be built from several different physical systems—including superconducting circuits, trapped atoms, particles of light and individual electrons. Each approach offers different advantages and different engineering problems.
IBM’s current quantum computers use superconducting qubits.
These are tiny, specially designed electrical circuits that behave quantum mechanically when they are cooled to temperatures near absolute zero. IBM has spent years learning how to manufacture them, control them, connect them and correct the errors they produce.
HRL works with silicon-spin qubits.
These use a quantum property of an individual electron, known as its “spin,” to store information. You can loosely imagine spin as an electron having two possible orientations that can represent quantum information—although the actual physics is stranger than a tiny object literally pointing up or down.
The important difference is size.
Superconducting qubits are relatively large. Spin qubits can potentially be made far smaller and packed together using manufacturing methods related to those already used to make semiconductor chips.
That could matter enormously because a useful, error-corrected quantum computer will require many physical qubits to create each reliable logical qubit—along with the wiring, cooling systems and control equipment needed to operate them.
At that scale, size stops being a detail.
Manufacturing stops being something that happens after the scientific breakthrough. Manufacturing becomes the breakthrough.
Silicon-spin qubits are appealing because they may offer a path toward fitting many more quantum components into a smaller physical space. But “may” is doing significant work there. Spin qubits come with their own hard problems, including controlling individual electrons, making devices consistent, connecting qubits and correcting errors.
IBM has not acquired a finished solution.
It has acquired another credible possibility.
Why HRL?
If IBM wanted to explore spin qubits, it could have started a new internal research program, partnered with an outside laboratory or invested in one of several companies pursuing the technology.
So why acquire HRL?
IBM already had a platform. HRL adds another path.
IBM already had
- Deep superconducting-qubit expertise
- A mature quantum-computing roadmap
- Advanced fabrication infrastructure
- Experience building and operating increasingly complex systems
HRL adds
- World-class silicon-spin-qubit expertise
- Advanced materials and cryogenic electronics
- Packaging, device control and prototype manufacturing
- Decades of aerospace, defense and industrial engineering
The acquisition matters because these capabilities may reinforce one another—not because one approach instantly replaces the other.
Over the last five years, through our work with Quantum World Congress, we have gotten to know IBM’s quantum team—including IBM Quantum Director Jay Gambetta—well. They are among the most rigorous, thoughtful and genuinely brilliant people we have encountered in the field. That does not mean every strategic bet will prove correct. But it does mean that when Gambetta describes HRL’s spin-qubit team as the strongest in the world, it is worth taking seriously.
The answer appears to be that HRL offers something unusually difficult to assemble: a promising technology, one of the field’s strongest teams and the wider engineering capabilities needed to turn an experimental qubit into a manufacturable system.
Just as importantly, IBM’s Quantum Lead, Jay Gambetta, has said IBM would not pursue a second technical path unless it rested on a foundation that could eventually be integrated with the company’s existing work.
That helps explain why HRL is more than a convenient hedge.
Its researchers are not working on an entirely disconnected version of quantum computing. Silicon-spin and superconducting qubits can both draw on advanced semiconductor-fabrication equipment, and HRL’s team is expected to begin producing chips at IBM’s facility in New York. The technologies are different, but the manufacturing knowledge and supporting infrastructure may reinforce one another.
HRL also brings much more than a qubit design.
Its work spans advanced materials, cryogenic electronics, device control, packaging, sensing, communications and prototype manufacturing. Those capabilities matter because scaling quantum computing will require far more than proving that an individual qubit works. The qubits must be fabricated consistently, connected, controlled, cooled and packaged as part of a reliable machine.
And HRL is not a conventional startup organized around one promising technical result.
The laboratory grew from the research operation of Hughes Aircraft and has spent decades developing advanced technologies for aerospace, automotive, defense and government applications. It is accustomed to the slow, difficult work of moving ideas from scientific possibility toward engineered systems.
Boeing and General Motors are also expected to continue partnering with IBM after the acquisition. IBM is therefore gaining not only researchers and intellectual property, but an institution with established industrial relationships and experience applying advanced science to operational problems.
IBM did not simply identify an interesting alternative qubit.
It identified a team, a body of knowledge and a manufacturing culture that could make pursuing that alternative credible.
The seductive myth of the winning qubit
Most emerging industries are narrated as races.
Someone is ahead. Someone is behind. One technology is gaining momentum. Another is about to be rendered obsolete.
This is understandable. Races are easy to explain. They give investors, journalists, policymakers and the public a clean storyline:
There are several competing approaches.
Eventually, one will prove superior.
The market will consolidate around it.
A winner will emerge.
But technological progress is rarely that clean.
The modern computing system did not emerge because one perfect component defeated all others. It is a stack of specialized technologies: processors, memory, networking equipment, storage systems, cooling infrastructure and software layers, each solving a different part of the problem.
There is no particular reason to assume quantum computing will develop differently.
A superconducting system may offer the clearest path to one generation of fault-tolerant computing. Spin qubits may ultimately make much larger systems practical. Photonic technologies may play an important role in connecting machines. Other modalities may prove best suited to sensing, communications, memory or specialized applications.
The future quantum computer may not be one machine built around one miraculous qubit.
It may be an assemblage.
A system.
A stack.
IBM’s acquisition begins to look especially important through that lens.
The company is not announcing that its superconducting approach has failed. It is continuing to invest heavily in it. Nor is it declaring that HRL’s spin qubits have won.
IBM is positioning itself to benefit from either outcome—and perhaps from an eventual combination of the two.
Gambetta has said that the future could be based on superconducting qubits, electron-spin qubits or some mixture of both. HRL’s researchers are also expected to use IBM’s fabrication facilities in New York, connecting a new qubit architecture to IBM’s growing manufacturing operation.
That is not a concession.
It is a strategy.
IBM acquired optionality
Organizations like certainty.
Certainty is easier to fund, easier to market and much easier to put on a roadmap.
We selected the right technology.
We know where this is going.
We are executing against the future.
But in a field as young and technically unsettled as quantum computing, excessive certainty can become a liability.
A company can spend years building expertise, infrastructure, intellectual property and organizational identity around one particular answer. That focus can create extraordinary progress.
It can also make changing direction almost impossible.
The real value of the HRL acquisition may therefore be less about replacing IBM’s current approach than about expanding the range of futures IBM can survive.
IBM is acquiring not only a different qubit modality, but expertise in materials, cryogenics, sensing, packaging and control electronics—the broader infrastructure quantum systems will require.
More importantly, it is acquiring the ability to learn faster if the field changes.
That is optionality—not in the vague corporate sense of “keeping our options open,” but as a serious institutional capability.
IBM can continue pushing superconducting systems toward nearer-term fault tolerance while testing whether spin qubits offer a better path to enormous scale—and absorb new evidence without rebuilding its quantum operation from scratch.
IBM has not eliminated uncertainty.
It has made uncertainty more manageable.
This is what a field beginning to mature looks like
Early conversations about frontier technologies tend to revolve around scientific possibility.
Can this work at all?
Can someone demonstrate it in a laboratory?
Can the number get bigger?
Can the error rate get smaller?
As a field matures, the questions become less cinematic.
Can the components be manufactured consistently?
Can thousands—or millions—of them be controlled, connected and maintained?
Can the system be produced reliably more than once?
Those questions may sound less exciting than a breakthrough experiment, but they are the questions that separate an extraordinary scientific achievement from an industry.
IBM’s move reinforces something the quantum sector is gradually discovering: its defining battles will not be fought exclusively in physics journals.
They will also be fought in fabrication facilities, materials science, packaging and the unglamorous engineering required to make a delicate phenomenon repeatable.
IBM is not only buying an idea.
It is buying an institution built to turn unresolved science into sustained engineering.
What other technology leaders should learn from this
There is a communications lesson here, too.
Frontier companies are often encouraged to present uncertainty as weakness. Investors want conviction. Customers want confidence. Employees want to believe the roadmap leads somewhere definite. So leaders are tempted to describe one technical approach as inevitable long before the evidence supports that conclusion.
That may work in the short term.
It can also trap the organization inside its own story.
Once a company has built its identity around possessing the answer, every adjustment begins to look like retreat. A new partnership looks like a hedge. A competing technology looks like a threat. Learning becomes harder because the narrative leaves no room for revision.
IBM is framing this differently.
It is not saying its existing approach was wrong. It is saying the problem is large enough to justify more than one serious path. That allows the company to expand its strategy without narrating the move as failure.
Other technology leaders should pay attention.
A strong narrative does not require pretending uncertainty has disappeared. It requires explaining what the organization believes today, what evidence could change that belief, which capabilities will remain valuable across multiple outcomes and why pursuing more than one path is a sign of discipline rather than doubt.
That is a particularly important distinction for startups.
Startups still need a point of view. Keeping every option open is not a strategy. But conviction does not have to mean rigidity.
Confidence does not require pretending uncertainty is gone.
A strong frontier-tech narrative should explain what the organization believes today, what remains unresolved and which capabilities will remain valuable across more than one possible future.
The goal is not to sound unsure. It is to show that the company can make a serious bet without trapping itself inside a story that leaves no room to learn.
The strongest narrative may be the one built to survive learning something new.
A company can make a serious technical bet without claiming every competing path is doomed. It can build its story around the problem it is uniquely equipped to solve—not the assumption that it has already predicted the industry’s entire future.
A narrative that leaves room for learning can feel more confident than one that insists uncertainty no longer exists.
The best frontier-tech narratives do not promise that the company has perfectly predicted the future.
They show that the company is built to remain relevant as the future becomes clearer.
A different way to recognize leadership
There is a persistent tendency in emerging technology to mistake conviction for vision.
The leader is the company with the boldest prediction.
The clearest roadmap.
The executive most willing to say that their architecture is the architecture.
But there is another kind of leadership: building an organization capable of being wrong.
Not directionless or endlessly hedging. Capable of making a serious bet, learning from it and still possessing the people, infrastructure and imagination required to change.
IBM’s acquisition does not prove that silicon-spin qubits will succeed. It does not prove that superconducting systems will fall short. It does not move us overnight into an era of useful, universal quantum computers.
Calling it a game-changer may therefore be premature.
But it does change the shape of IBM’s game.
The company is no longer pursuing only one physical answer to quantum computing’s central engineering problem. It is assembling enough scientific and manufacturing capacity to pursue several possibilities—and perhaps eventually combine them.
The quantum race may still produce winners.
But the advantage may belong not to the company that tells the most confident story about the future, but to the one whose strategy—and narrative—can survive learning something new.
