Chapter IV: The Endgame Scenarios — 2025–2035

The following three scenarios are grounded in specific technical assumptions and bounded by the trendlines documented above. Each includes falsifiable near-term predictions.

Methodology note: The probability weights assigned below are the author's subjective estimates based on the trendline data and expert commentary surveyed for this report. They are not derived from a formal model, expert elicitation process, or historical base rate analysis. Readers should note that the combined 70% probability assigned to favorable outcomes (Scenarios A and B) is arguably optimistic relative to historical base rates for deep-tech commercialization timelines—analogous technologies (autonomous vehicles, nuclear fusion, gene therapy) have typically taken 2-3× longer than initial projections suggested, and experienced extended "winters" that the 30% plateau scenario may underweight.

Scenario A: "The Error Correction Decade" (Base Case, ~50% Probability)

Technical assumptions: Error correction overhead continues to fall through a combination of improved gate fidelities (reaching ~99.99% by 2028 in at least one modality), better codes (qLDPC, code concatenation), and scaling of demonstrated approaches. No single dramatic breakthrough, but steady accumulation of improvements.

2026–2027: First credible demonstrations of quantum advantage on practical problems—most likely in quantum simulation of molecular systems (drug candidates, catalyst design) or certain optimization problems. These advantages will be narrow: specific problems, specific instances, modest speedups over the best classical methods. IBM targets quantum advantage by end of 2026. Classical counterarguments will be vigorous. Quantinuum Sol (192 physical qubits) deployed.

2028–2029: Fault-tolerant machines with 100-200 logical qubits become available (Quantinuum Apollo-class, IBM Starling). These machines can run algorithms of genuine scientific interest—simulating small drug molecules, modeling materials with strong quantum correlations, certain cryptographic applications. Revenue begins to grow beyond cloud access fees. DARPA QBI Stage C evaluations underway.

2030–2032: 500-1,000 logical qubit machines operating at error rates below 10⁻⁶. First commercially valuable quantum simulations—drug candidates that could not have been identified classically, materials with novel properties. Quantum computing is a recognizable (if small) industry generating $5-10 billion annually. Multiple companies achieve profitability on quantum-as-a-service.

2033–2035: 1,000+ logical qubit machines are standard tools in pharmaceutical R&D and materials science. Total value creation reaches $10-50 billion per year. The cryptographic threat from quantum computing remains 10+ years away for RSA-2048 breaking.

Near-term confirmations (2026–2027)

Two-qubit gate fidelities reach 99.95%+ in at least two modalities. Quantinuum Sol ships on time with 192+ qubits. IBM's Kookaburra demonstrates modular logical qubit storage. At least one peer-reviewed demonstration of quantum advantage on a commercially relevant problem survives classical challenge for 12+ months.

Scenario B: "The Breakthrough Accelerates" (~20% Probability)

Technical assumptions: A hardware or algorithmic surprise dramatically reduces error correction overhead beyond current projections. Candidates include: topological qubits proving out (Microsoft’s approach validated by independent groups), QuEra-style 100× overhead reduction^[1] techniques working at scale, a novel qubit modality emerging from a national lab, or a fundamentally new error correction code that changes the math.

2027–2028: 500+ logical qubit machines available, three to five years ahead of the base case. Revenue cycle kicks in faster as pharmaceutical and financial companies begin using quantum simulation for genuine competitive advantage. Government involvement intensifies, with national security agencies accelerating quantum programs.

2028–2030: 1,000+ logical qubit machines. The materials-discovery feedback loop begins to close: quantum computers identify better qubit materials, which improve quantum computers. Quantum + AI convergence accelerates. The cryptographic threat timeline compresses significantly; PQC migration becomes urgent for all organizations.

Near-term confirmations (2026–2027)

Microsoft's topological qubit claims validated by independent replication, OR QuEra's overhead reduction demonstrated on a production system, OR a surprise announcement of a novel high-fidelity qubit modality with inherent error suppression.

Scenario C: "The Long Plateau" (~30% Probability; historical base rates for deep-tech commercialization may support 40-50%)

Technical assumptions: Gate fidelities plateau at approximately 99.95%—good but insufficient for cost-effective error correction at scale. The overhead wall proves harder to break than hoped. qLDPC codes turn out to require connectivity that is impractical at scale. Code concatenation approaches hit diminishing returns. Classical algorithms continue to improve and erode quantum advantage claims.

2027–2029: Error-corrected machines remain at 50-100 logical qubits. Quantum advantage claims continue to be contested. The revenue cycle fails to materialize at scale. Some quantum startups fail or consolidate. Investment cools but does not collapse.

2030–2035: Useful logical qubit counts grow slowly—100 usable logical qubits not achieved until 2033+. Quantum computing becomes a niche $5-10 billion industry by 2035, valuable for specific applications (quantum sensing, quantum communications, quantum annealing for optimization) but not the transformative general-purpose computing platform some predicted.

Near-term disconfirmations (2026–2027)

Two-qubit gate fidelities remain below 99.95% across all modalities by end of 2027. No credible quantum advantage demonstration survives classical challenge. DARPA QBI downgrades multiple companies from Stage B. Major quantum companies miss roadmap milestones by 12+ months.