AINEWS
Trump Sets a 2028 Quantum Deadline. Microsoft Says 2029. A Nature Paper Calls It Complete Codswallop.
Three days after Microsoft unveiled a quantum chip and Trump signed an executive order setting a 2028 deadline, a University of St.
Three days after Microsoft unveiled a quantum chip and Trump signed an executive order setting a 2028 deadline, a University of St.
On June 22, 2026, the White House and Microsoft delivered two of the loudest quantum computing announcements of the year, just hours apart. Within 72 hours, the University of St. Andrews physicist Henry Legg had put a peer-reviewed rebuttal on the record in Nature. The three events do not contradict each other. They expose the same structural gap that has defined quantum computing for a decade: the distance between what is being promised and what has actually been demonstrated.
The Trump administration used the date to release an executive order calling for a quantum computer "powerful enough for scientific discovery" by 2028, framed as part of a race with China. Microsoft used the same day to unveil Majorana 2, a chip built around a controversial hypothetical particle that the company says brings a "scalable, practical" quantum machine within reach by 2029.
Then came Legg's "complete codswallop." The University of St. Andrews physicist called the Microsoft announcement "complete codswallop" in comments to The Verge, and published a peer-reviewed paper in Nature on June 24 walking through why Microsoft's underlying physics claims remain contested rather than treating them as settled engineering.
So what is a quantum computer actually built for? Specialized math. Quantum machines exploit subtle quantum behaviors: their basic units, called qubits, can hold combinations of values rather than just 0 or 1, and they can be linked so that operations on one reach across the system. That gives a narrow set of problems, including simulating molecules, optimizing routes, and factoring large numbers, a fundamentally different attack angle than classical chips. For most everyday computation, from streaming a video to training a language model, classical hardware wins easily and cheaply. The promise of quantum is narrow but real: certain problems should eventually become tractable that no amount of additional silicon will solve the classical way.
The catch is that no quantum computer has yet performed a single commercially useful task. Existing machines are still small and error-prone, typically housing a few hundred to a few thousand qubits. Running a useful algorithm requires aggressive error correction, which itself eats most of the available qubits. Headline milestones show the hardware is improving, but they have not yet produced a result that beats classical hardware on a commercially or scientifically meaningful task.
Google, IBM, Amazon, Microsoft, and national quantum programs in the U.S. and China have collectively poured billions of dollars into quantum computing development over the last decade. The promise of the technology is that it could eventually lead to discoveries in medicine, advances in materials science, and improvements in machine learning. That is the contest Legg's paper is pushing back against: not the underlying science, but the cadence of corporate and political claims.