So Microsoft put out a press release last week saying they've built a working quantum chip called Majorana 1, and it's got eight topological qubits on it, and the roadmap slide (there's always a roadmap slide) says they'll scale that to a million qubits on a single chip eventually. It made the rounds Wednesday, splashed across every tech outlet with the same headline about a "new state of matter," and I sat there reading it with a very specific feeling in my stomach. Deja vu, basically. Not the good kind.
Here's why. Anyone who's followed this space for more than about five minutes remembers that Microsoft already claimed to have found Majorana zero modes back in 2018. That was a big Nature paper, a lot of triumphant quotes, the whole thing. Then in 2021 it got retracted after other physicists dug into the underlying data and found the peaks in the measurements had been selectively presented — cleaned up, basically, in a way that made a messy signal look like clean evidence of something that wasn't actually there. That's not a small correction. That's the kind of thing that makes a field suspicious of you for a while.
And now, four years later, they're back with a chip, a name (Majorana 1 sounds like a Bond villain's yacht), and another Nature paper, and the initial reaction from a chunk of the condensed matter physics community was basically "okay, show us the data this time." Some researchers pointed out the paper doesn't actually contain a direct topological qubit measurement, just supporting evidence that the ingredients are there. Microsoft's own framing on the announcement leaned pretty hard into "trust us, it's coming," which, given the history, is not exactly the reassurance you'd want.
I don't say any of this to be a jerk about it. Topological qubits are a genuinely interesting idea, the pitch is that they're protected from noise at a hardware level instead of needing the mountain of error correction that superconducting qubits (Google, IBM) or trapped ion qubits (IonQ, Quantinuum) require. If it works, it's a real shortcut to a useful quantum computer instead of the current approach of bolting a thousand physical qubits together to get one good logical one. It's a legitimately elegant idea. I just have zero ability to take an announcement like this at face value anymore, and I don't think that's an unreasonable position after 2018.
This is also just a pattern I've watched play out for basically the whole time I've been writing this blog. D-Wave was doing this in 2013, showing off "quantum annealers" and slide decks with exponential curves on them, and it took years of independent benchmarking for the field to mostly conclude their early machines weren't demonstrating quantum speedup on anything that mattered. IBM's had multiple "we'll have a useful quantum computer by [year]" roadmaps that quietly slid. Google's 2019 "quantum supremacy" claim got immediately contested by IBM, who said a classical supercomputer could do the same task in days, not the ten thousand years Google claimed, given the right optimizations. None of this means quantum computing is fake or a scam. It means the PR cycle for this field runs way ahead of the actual science, every single time, and the gap between "we published a paper" and "you can rent time on this to do something useful" has consistently been measured in many years, not months.
What'll actually tell us something is whether independent groups can replicate the topological protection claim on this specific device, and whether Microsoft publishes the raw data instead of curated figures this time around. That's not a fast process. Papers get written, other labs request the datasets, people argue about methodology in journal comments sections that nobody outside the field ever reads. I'm bookmarking this one and setting a reminder to come back to it in a year, because that's about how long it took last time for the cracks to show. Maybe this time it holds up. I kind of hope it does, honestly, the physics is cool even if the marketing gives me a headache. I'm just not popping any champagne over a press release with a roadmap slide in it.