Are We Computing Quantum in the Wrong Base? with Ivan Deutsch

Are We Computing Quantum in the Wrong Base? with Ivan DeutschIvan Deutsch is Distinguished Regents' Professor of Physics and Astronomy at the University of New Mexico and the founding director of CQuIC, the Center for Quantum Information and Control. Along with his longtime collaborator Poul Jessen, Ivan helped lay the theoretical foundations for neutral-atom quantum computing in the 1990s: trapping individual atoms in optical lattices, cooling them to near absolute zero, and shuttling them in parallel to perform quantum logic. The companies commercializing those ideas today — QuEra, Pasqal, Atom Computing, Infleqtion, and the newly announced Aurora out of Caltech — are building on architectural concepts that trace directly to his group's early papers. His 9,600+ citations across quantum information, atomic physics, and quantum control place him among the most-cited theorists in the field.The reason to talk to Ivan now is that he has been making a quietly heterodox argument: every one of those commercial platforms encodes information in two energy levels of an atom that has ten or sixteen, and Ivan thinks the field should be asking whether that's the right choice — not for information density, which is only a logarithmic gain, but for fault tolerance. This conversation goes deep on qudits, spin cat codes, and the co-design philosophy that has shaped Ivan's career at the interface between theory and experiment, ions and neutral atoms, and academia and industry. If you are following neutral-atom hardware, fault-tolerant quantum error correction, or the emergence of regional quantum ecosystems, this episode is essential.What You'll LearnWhy neutral atoms were the "underdog cousins" of trapped ions — and the precise trade-off at the heart of a 30-year rivalry: ions are great and terrible because they're charged; neutral atoms are great and terrible because they're neutralWhat the original neutral-atom quantum computing paper actually got right: the parallel atom-movement architecture now central to QuEra, Atom Computing, and Infleqtion's roadmaps was already there — even if the Rydberg blockade's full power wasn't appreciated until laterWhat qudits are and why fault tolerance, not information density, is the compelling argument: the information gain from base-2 to base-10 is only logarithmic, but co-designing error-correcting codes with the physical structure of the hardware may be transformativeHow spin cat codes work: using the extra energy levels inside a single atom for error redundancy, directly analogous to bosonic cat codes in microwave cavities, with fault-tolerant thresholds that may surpass standard qubit surface codesWhy biased error correction matters: real physical errors in neutral atoms aren't arbitrary, and codes designed around the dominant error channels — including leakage and erasure — can dramatically outperform worst-case generic schemesHow leakage becomes an asset: when population escapes the qubit subspace into other levels, detecting that escape converts it from an unknown error into an erasure error, which is far easier to correctWhy working at interfaces is where the creative work happens: Ivan's career has been built at the boundary between theory and experiment, between ion-trap and neutral-atom communities, and now between research and industryHow New Mexico became a quantum hub: the founding of QNM-I, the partnership with Colorado, and the Elevate Quantum Tech Hub — turning decades of national-lab and university strength into an actual industrial ecosystemResources &amp; LinksGuest LinksIvan Deutsch — CQuIC Faculty Page — Research profile and publication list at the Center for Quantum Information and Control at UNMGoogle Scholar Profile — 9,600+ citations across quantum information, atomic physics, quantum optics, and quantum controlNSF Q-SEnSE Research Profile — Ivan's role in the NSF quantum sensing and engineering centerKey PapersQuantum optimal control of ten-level nuclear spin qudits in Sr-87 (LANL/CQuIC) — The theoretical demonstration of arbitrary SU(10) maps in strontium-87 with average fidelity ~0.9992; the core technical result behind the qudit computing program discussed in the episode<a href="https://www.researchgate.net/profile/