Quantum Observations from Deep Tech Week - Seattle
The Ecosystem Entanglement™ Series - Post 4: A practitioner’s view on commercializing deep tech in hyperbolic markets
This week, 1,100 leaders from 60 countries are gathered in London for The Economist Enterprise’s 5th Annual Commercialising Quantum Global 2026. The UK Minister for Science is on stage. NVIDIA’s VP of Quantum is on stage. The Deputy Chief Science Officer of the US Space Force is on stage. The heads of quantum at Barclays, Lloyds, Sanofi, and E.ON are in the room.
Last week I spoke at #Deep Tech WeekSeattle in #QForce — a panel on building the deep tech and quantum workforce. This week I am part of a #WomenInCloud
panel in the #WICxMarketMaker Summit 2026 talking about the physics of
commercializing deep tech.
Here are five observations around quantum from these Seattle discussions that I suspect are common to the dialogue in London this week:
Quantum is not one market.
Quantum is three markets on three different timelines. The harvest-now-decrypt-later risk is a time-critical security problem that is in the process of being addressed today by Post Quantum Cryptography (PQC). Quantum Sensing is delivering commercial advantage now. Q-CTRL’s navigation system outperformed classical GPS backups by 100X in real-world flight tests in 2025. Quantum Computing is 3-7 years from broad commercial deployment.
Talking about quantum as a single market, just like talking about AI as a catch-all, is misleading. It overlooks very different product, problem, and market attributes across three distinct technology domains.
Problem-Solution fit is unusually difficult in quantum.
The wrong question is “how can quantum improve our current processes?” That imports classical computing assumptions into a fundamentally non-classical problem space. Not every problem benefits from quantum.
The right question is: where does our most important problem hit an exponential scaling wall that quantum can solve? Does quantum offer orders-of-magnitude advantage there, or just incremental improvement?
That question requires technical depth most commercial teams don’t yet have. It’s also why the translation layer matters so much. The translators aren’t just explaining the technology. They’re working within their teams to identify the use cases, many of which are some of the most natively complex and mathematical use cases that exist. Discerning which quantum computing use cases will bring the greatest commercial value will require deep industry expertise married with category creation muscle.
The quantum tech stack requires architectural re-engineering.
Quantum is foundationally different from Cloud and AI because of this point, and the GTM implications naturally follow.
Cloud infrastructure can be plugged into existing architecture. Workloads can be moved to the cloud one use case at a time without touching the underlying architecture - same applications, same data, same workflows, just running on someone else’s servers. The integration surface is relatively shallow: APIs, connectors, lift-and-shift migrations. Architecture could be revisited during a cloud migration but it is an option, not a necessity.
AI integration goes deeper but still largely augments existing systems. You add a model to your existing data pipeline. You embed a copilot into your existing application. You call an API. The underlying architecture — the databases, the compute, the software stack — remained largely intact. AI sits on top.
Quantum requires something more fundamental: a re-engineering of the computational workflow itself. Quantum processors operate on qubits — superposition, entanglement, interference — that have no direct equivalent in classical computing’s binary logic. To extract value from a quantum processor you need what researchers call a hybrid classical-quantum architecture — where classical computers handle data preparation, pre-processing, orchestration, and post-processing, and quantum processors handle only the specific computational core where quantum advantage exists.
This is not an integration. It is a re-engineering of the computational workflow from the ground up — identifying which part of the problem benefits from quantum, designing the handoff between classical and quantum systems, managing the error correction and noise mitigation that quantum hardware requires, and then connecting the output back into classical infrastructure in a way that makes it usable.
You can’t interface a quantum processor with a classical system the way you integrate an API or migrate a workload. You have to design a hybrid classical-quantum architecture from the ground up — identifying which specific part of the problem benefits from quantum, redesigning the handoff between two fundamentally different computational paradigms, managing error correction at the hardware level, and connecting the output back into classical infrastructure in a usable form. The middleware to do this reliably at scale is still being built. The skills to design it are scarce, and the investment required is architectural, not incremental.
As a result, the Quantum ROI model is more surgical.
Cloud ROI was horizontal and broad. AI adoption has been characterized by early, rapid growth with the full ROI still to be determined. Both Cloud and AI are not comparable to quantum as their market adoption is virtually ubiquitous and the tech stacks and not transformational.
Quantum ROI follows directly from the first principles question. If you’ve identified the right problem, the investment becomes deliberate and the outcome measurable. If you haven’t, you’re spreading budget across vague exploration and getting nothing back from any of it. You measure in option value — what is the cost of not having this capability when the threshold is crossed? — not in quarterly returns. The ROI is built use case by use case and not peanut-buttered across an organization or an infrastructure layer.
Category Creation then becomes a key asset.
Even when organizations identify the right use case — where quantum provides genuine, outsized advantage — that insight has to be converted into market category creation and buyer education. In most technology markets that work happens quickly. In quantum it adds material time to an already extended GTM timeline.
You are not just selling a product. You are building the buyer, defining the category, writing the playbook, and educating a market that doesn’t yet know what questions to ask. That is not a sales problem. It is a market formation problem. And it is one of the most underestimated dimensions of quantum commercialization.
The translation layer is a key bottleneck.
The people who can sit between a quantum research team and an industry problem owner, fluent in both “languages,” are the scarcest resources in the ecosystem today. But this is more than a talent gap. The translators aren’t just bridging science and industry. They’re building a market that doesn’t yet exist. Category creation is a fundamentally different job than driving GTM into an established category.
The Economist holding this event for the fifth consecutive year is itself a signal. When the world’s most credible business publication dedicates its platform to quantum commercialization — and fills the room with 1,100 leaders from 60 countries — the question stops being “is quantum real?” and starts being “are you positioned for it?”
More in my Ecosystem Entanglement™ series — link in comments.
#QuantumComputing #EcosystemEntanglement #DeepTech #CommercializingQuantum #QuantumAdvantage #QForce