AI and quantum computing: frenemies, not rivals

The Economist argues (July 2026) that AI and quantum computing look more complementary than rivalrous, overlapping in three ways — the problems they tackle, how each is used to build the other, and the resources they compete for — with co-operation the winning strategy in each:

Same problems, different methods. Quantum's most promising use-case is simulating quantum mechanics itself — materials science, chemistry, biology. AI firms such as Isomorphic Labs and CuspAI chase the same targets via pattern-recognition rather than exact simulation. CuspAI co-founder Chad Edwards previously led at quantum startup Quantinuum before the pace of AI progress convinced him to switch sides.

An optimisation loop, not a duel. IBM's Ruchir Puri frames it as symbiotic: quantum computers enable more accurate simulations, which train better AI models, which in turn help identify the promising chemistry at which to point the quantum computers — "the first time that optimisation loop can be closed."

Building each other. AI can tackle quantum's biggest hurdle — qubit error-correction — as with Infleqtion's AI-powered decoder. Conversely, quantum can enhance AI: Silicon Quantum Computing sells quantum "reservoir" chips that cut its investor-and-customer Telstra's AI training times by 90%; founder Michelle Simmons notes they pitch it simply as an "AI accelerator."

Efficiency, not just speed. The appeal has shifted. A quantum machine runs at megawatt rather than gigawatt scale — SQC's rack-sized hardware uses under a tenth of the power of a rack of GPUs. In May 2026 the US government invested $2bn across quantum labs including IBM and D-Wave; IBM boss Arvind Krishna called the quantum-computing industry "right around the corner."

Read more: The Economist