Microsoft $MSFT pushes its "long game" in quantum: it announced a new chip Majorana 2, which was designed with the help of AI and on which it aims to have a commercially usable quantum computer by 2029. The chip uses, instead of common aluminum, unconventional lead-based superconducting materials that Microsoft selected using its own AI tools for materials research and with which it boasts roughly a thousandfold improvement in some parameters compared to the previous generation — at the same time it had to solve the technical problem of how not to "dissolve" water-soluble lead during manufacturing.

Strategically, it's important that Microsoft thereby aligns its timeline with IBM, which is also aiming for a real quantum business by the end of the decade and is massively investing in its own chip manufacturing. The difference is in the approach: Microsoft continues to bet on the topological direction based on quasiparticles called Majoranas — according to the company, the key to more stable, better-scalable qubits — but some physicists remain skeptical because they haven't yet seen enough open data that would unequivocally confirm those observations. Therefore, Majorana 2 is mainly a signal today: Microsoft is showing that it takes quantum as seriously as AI and that it wants to combine the two areas (AI for materials design, quantum as an additional computational layer), but the final verdict on whether the Majorana play was a brilliant shortcut or a dead end will come only when someone demonstrates a truly reliable machine in regular operation.