Quantum computing has spent much of its public life in a kind of limbo: always “almost there,” frequently impressive in demonstrations, and yet stubbornly reluctant to deliver the one thing that would silence most scepticism—clear, repeatable evidence that quantum machines can outperform the best classical computers on problems that matter to real businesses.
Now, that argument is shifting. Over the past year, more companies have begun treating quantum not as a speculative research programme but as an investment category with timelines, milestones and procurement plans. The result is a noticeable migration from lab meetings to boardroom conversations, and from “what if” to “when.” In pharmaceuticals, finance and crypto, the pitch is familiar: quantum could accelerate molecular modelling, improve optimisation and risk analysis, and eventually force a rethink of cryptography. But the new twist is that the conversation is increasingly about implementation details—error correction, system reliability, software workflows and integration with existing infrastructure—rather than only about theoretical advantage.
At the same time, sceptics are pushing back harder than ever. Their concern isn’t that quantum computing will never be useful; it’s that the industry’s communication often outruns its engineering. They point to a recurring pattern: ambitious claims based on idealised models, followed by years of incremental progress that don’t map neatly onto business expectations. In other words, the debate is no longer simply “is quantum real?” It’s “how real, how soon, and for which workloads?”
To understand why the momentum feels different this time, it helps to separate three layers that are often blended together in headlines: hardware capability, algorithmic promise, and operational usefulness. Quantum computing can be “real” at the hardware level—meaning qubits behave quantum mechanically and experiments run—while still being far from operationally useful for the kinds of tasks companies care about. Conversely, even when hardware is limited, software and hybrid approaches can sometimes extract value earlier than full fault-tolerant quantum computers would allow. The current wave of corporate interest appears to be driven by a belief that these layers are converging faster than before, even if the endgame remains uncertain.
Pharmaceuticals: from chemistry fantasies to workflow engineering
In pharmaceuticals, the most compelling quantum narrative is also the most intuitive: molecules are quantum systems, so simulating them should be a natural fit for quantum computers. But the path from intuition to drug discovery is not straightforward. Drug development is not a single computation; it’s a pipeline of modelling, screening, refinement, and experimental validation. Even if quantum algorithms can represent certain aspects of molecular behaviour more efficiently than classical methods, the practical question becomes: can they do so within the constraints of today’s noisy devices, and can they plug into existing computational chemistry workflows?
That’s where the corporate shift is happening. Rather than betting solely on a future quantum computer that can run large-scale simulations, many teams are exploring near-term use cases that focus on specific subproblems: estimating energies of candidate molecules, exploring reaction pathways, or improving parts of the search process used in screening. The emphasis is increasingly on hybrid strategies—quantum circuits embedded inside classical optimisation loops—because fully quantum end-to-end workflows are still beyond reach for most organisations.
However, hybrid does not mean trivial. Hybrid quantum-classical systems introduce their own bottlenecks: the number of circuit evaluations required can be enormous, noise can distort results, and the optimisation landscape can be difficult to navigate. Companies are therefore paying attention to what might sound mundane but is actually decisive: how many shots (repeated measurements) are needed, how sensitive outputs are to calibration drift, and whether error mitigation techniques can recover useful signal without erasing the computational advantage.
A unique angle emerging in pharma discussions is the move toward “quantum readiness” rather than “quantum replacement.” Instead of asking whether quantum will replace classical chemistry, teams are asking whether quantum can become another instrument in the toolbox—one that is particularly valuable for certain classes of problems or for accelerating specific steps. This reframing changes the success criteria. A quantum system doesn’t need to solve every chemistry problem to justify investment; it needs to demonstrate measurable improvements in targeted workflows, such as reducing time-to-candidate or improving the quality of intermediate approximations.
There is also a growing recognition that data and model integration matter as much as the quantum computation itself. Drug discovery teams already rely on machine learning models trained on large datasets. The question becomes how quantum outputs—whether exact, approximate, or probabilistic—can be used to inform those models. In practice, this means building interfaces between quantum solvers and classical pipelines, defining how uncertainty is represented, and ensuring that quantum-derived features are stable enough to be useful downstream.
Sceptics argue that this is still too early for broad claims. They note that many quantum chemistry proposals assume error rates and scaling that are not yet available. But proponents counter that the industry is learning to treat quantum as a component in a larger system, not as a magic oracle. The most credible near-term efforts are those that define narrow, testable hypotheses: for example, whether a quantum-assisted method improves a particular energy estimation task under realistic noise conditions, or whether it reduces the number of classical iterations required to reach a target accuracy.
Finance: optimisation, risk, and the hard reality of advantage
In financial services, the quantum pitch tends to revolve around optimisation and sampling: portfolio construction, risk analysis, constraint satisfaction, and certain forms of Monte Carlo-like estimation. These are areas where classical algorithms are already highly sophisticated, and where the bar for “quantum advantage” is extremely high. A quantum approach must either outperform classical methods on speed, accuracy, or cost—or it must offer a qualitatively different capability, such as handling constraints or distributions in ways that are difficult to replicate classically.
The reason corporate interest is increasing anyway is that finance has a strong culture of experimentation and benchmarking. When quantum vendors and research groups provide tools, financial firms can test them quickly against internal baselines. That creates a feedback loop: even if quantum doesn’t win outright, it can reveal where hybrid methods might help, or where certain problem formulations are more amenable to quantum representations.
But the sceptical critique remains sharp. Many optimisation problems can be mapped to quantum circuits, yet the mapping is not free. Circuit depth, connectivity constraints, and noise can erase any theoretical benefit. Moreover, finance problems often come with messy real-world constraints—transaction costs, liquidity effects, regulatory requirements—that complicate the clean mathematical formulations used in academic papers.
So the most interesting developments in finance are not necessarily about dramatic quantum speedups. They’re about problem reformulation and workflow integration. Teams are experimenting with how to encode constraints efficiently, how to choose ansätze (the parameterised circuit structures) that are expressive enough but not too deep, and how to design classical optimisers that work well with noisy quantum measurements. In other words, the “quantum advantage” story is increasingly entangled with software engineering.
Another factor is the growing attention to error mitigation and robustness. In finance, decisions can be sensitive to estimation errors. If a quantum routine produces outputs with high variance or systematic bias, it may be unusable even if it is fast. Therefore, firms are focusing on whether quantum outputs can be made reliable enough for decision-making, not just whether they can produce a plausible answer.
There is also a strategic dimension. Even if quantum computers are not yet capable of solving major finance workloads, the industry is preparing for the long-term implications of cryptographic change and for the possibility that quantum-enhanced methods could emerge sooner than expected. That preparation includes inventorying cryptographic dependencies, evaluating post-quantum cryptography options, and building governance around migration timelines. In that sense, quantum investment in finance is partly about resilience planning, not only about immediate computational gains.
Crypto: the security clock is real, but the timeline is nuanced
Cryptocurrency and broader digital security communities have been among the most vocal about quantum computing’s potential impact, largely because of cryptography. Public-key cryptosystems used widely today—such as those based on factoring and discrete logarithms—are vulnerable to sufficiently powerful quantum computers running algorithms like Shor’s. That vulnerability is not hypothetical in the abstract; it is a known consequence of quantum computing’s ability to exploit periodicity and perform certain computations more efficiently than classical machines.
Yet the practical question is timing and scale. “Sufficiently powerful” is doing a lot of work. Quantum computers capable of breaking modern public-key cryptography would require not just more qubits, but fault-tolerant operation with error correction at a level that is still under active development. That means the crypto community’s challenge is to plan for a threat that is real but uncertain in when it becomes actionable.
This is why the most credible quantum-related activity in crypto is often less about building quantum computers and more about cryptographic migration. Post-quantum cryptography (PQC) aims to replace vulnerable schemes with alternatives designed to resist quantum attacks. The work involves selecting algorithms, assessing performance trade-offs, and updating protocols across ecosystems. For blockchain networks, this can be especially complex because upgrades must maintain compatibility, security guarantees, and decentralised governance.
The unique tension in crypto is that hype can spread quickly. Some narratives imply that quantum will “break everything” overnight, while others dismiss the threat entirely until a quantum computer exists that meets specific engineering thresholds. The more useful middle ground is to treat PQC migration as a long project with staged milestones: identify where vulnerable cryptography is used, evaluate PQC candidates, test implementations, and plan for phased rollout.
Meanwhile, there is also a separate line of speculation: whether quantum capabilities could enable new cryptographic primitives or improve certain aspects of blockchain security. While that’s possible in the long term, the near-term focus remains on protecting existing systems. In practice, the quantum revolution in crypto is less about quantum computation replacing current cryptography and more about ensuring that the cryptographic foundations survive the transition.
The sceptics’ core argument: timelines, error correction, and the gap between demos and deployments
Across all three sectors, sceptics converge on a few themes.
First is the timeline problem. Quantum