In December 1945, Alexander Fleming stood in Stockholm to accept a Nobel Prize for the discovery already saving lives. Penicillin was the closest thing the 20th century had to a miracle. And at the height of that moment, Fleming spent part of his lecture warning the world about the drug for which he earned the honor. He described how easily he could breed resistant microbes by exposing them to doses too small to kill.
Then he made a prediction. “The time may come when penicillin can be bought by anyone in the shops,” he said, and careless underdosing would teach the bacteria to survive.
He was right. Resistance emerged quickly, and antimicrobial resistance is now a permanent feature of modern medicine.
But Fleming understood the threat came from the flip side of the very mechanism that made it work. Every time the drug kills the susceptible bacteria, it clears the field for the survivors. You can’t separate the cure from the pressure that erodes it.
Quantum computing has that same structure. A conventional computer stores information in bits, each either a 0 or a 1. A quantum computer uses qubits, which can hold a blend of both at once and can be linked so that the state of one depends on the state of another.
The machine can work through an enormous number of possible answers simultaneously rather than one at a time. Quantum matters for a narrow yet valuable class of problems—modeling how molecules behave, optimizing systems with too many moving parts to brute-force, and breaking certain kinds of encryption.
The industries quantum reaches first rely on a deeper understanding of chemistry, materials, or logistics: pharmaceuticals, energy, and finance. But what those industries develop with quantum will show up everywhere else—in the batteries, semiconductors, and supply chains the rest of the economy runs on.
Quantum’s promises and risks
Fragile qubits pick up errors from heat, stray electromagnetic noise, and the act of measuring them. Today’s quantum systems accumulate errors faster than they can finish a long calculation.
A “fault-tolerant” machine is a quantum computer capable of correcting those errors while it runs, spreading each unit of information across many physical qubits so the calculation survives the failure of any single qubit.
The computational leap that lets a fault-tolerant machine simulate a catalyst, battery chemistry, or a drug binding pocket can also factor the large numbers protecting today’s data.
The upside and the exposure come from the same place, and the timeline has been moving. For years, “fault tolerance” was a someday phrase. IBM has published a detailed path to a fault-tolerant machine by 2029, with roughly 200 logical qubits, and expects demonstrable quantum advantage sooner, in 2026. Quantinuum has laid out an accelerated roadmap to universal, fully fault-tolerant computing by 2030. My company, QuEra, has published a roadmap that expects a fault-tolerant machine in 2028.
Different groups are approaching quantum differently—from individual atoms held in place by laser beams, electrically charged atoms suspended in electromagnetic traps, or supercooled circuits etched onto chips much like the ones in a laptop.
These approaches share almost no hardware, and yet they converge on the same destination, pulling their dates inward.
Caveats and stewardship
The machine that breaks today’s encryption isn’t yet here. RSA is the standard security algorithm that protects most of what you do online, including bank transfers, medical records, and private messages. RSA rests on an asymmetry: Multiplying two enormous primes is easy, and reversing it is effectively impossible for a classical computer.
A large enough fault-tolerant quantum computer would make that backward step practical. However, such a machine is likely still a few years away.
But that doesn’t buy you time. Encrypted data can be harvested today and decrypted later, once the hardware arrives. For anything that must stay secret for a decade, the exposure clock started years ago.
So the stewardship response resembles health systems’ disciplined use, surveillance, and a pipeline of new drugs, built on the assumption that frontline antibiotic treatment will eventually fail and need replacing.
Cryptography is being asked to do the same thing. In August 2024, the U.S. National Institute of Standards and Technology, which most of the world follows for encryption standards, published its first approved replacements for the algorithms a quantum computer could break. The changeover has three parts:
- Inventory: Establish what data you protect and how long it must stay protected.
- Replacement: Put the new quantum-resistant algorithms in wherever the old ones are doing the work.
- Flexibility: Build systems so that if a new algorithm is later found to be weak, you can pull it out and drop in another without rebuilding everything around it.
We’ve been here before
This has all happened before, at the infrastructure level. When wireless radio arrived, communication became a signal anyone in range could hear. That shift from private to open by default turned cryptography from a niche craft into a serious discipline.
Quantum forces the next upgrade of radio’s lesson to grow defenses to match the new capability. Antibiotics cured infections and made entire fields of medicine far safer—routine surgery, organ transplants, cancer chemotherapy—by giving clinicians a way to prevent and treat the infections those procedures invite.
The cure’s real value was everything it made possible downstream. Similarly, quantum can “simulate a molecule” or “solve an optimization,” but the prize is the adjacent problem-solving in materials, chemistry, logistics, and models that classical machines simply can’t reach.
So the useful posture is both “quantum is a threat” and “quantum is a miracle.” I believe quantum belongs on a company’s roadmap the way infection control belongs in a hospital. Invest in the upside for returns and manage the exposure continuously.
Fleming argued for using penicillin well. Eighty years later, the same discipline applies to one of the most powerful computing capabilities on the horizon. Treat quantum like the wonder drug it is. Take the opportunity seriously, and respect the risk enough to stay ahead of it.
Yuval Boger is the chief commercial officer of QuEra.