My Articles, Opinions and Analyses
What Quantum Computers Are For: A Mathematician’s Map of Quantum Advantage
The mathematician's answer to what quantum computers are for: three problem families where the advantage is real, the theorems that cap combinatorics at quadratic, and the open problems stated in your notation ...
Quantum Computing for Mathematicians: From Four Axioms to Shor’s Algorithm
Everything a mathematician needs to understand quantum computing: the four axioms, why "trying all answers at once" is wrong, how interference powers Shor and Grover, and where mathematicians are needed next ...
Several Vendors Now Supply Quantum Error Correction’s Classical Loop. The Qubits Are Still the Hard Part.
Riverlane, Qblox, Altera, Xanadu, AMD, NVIDIA and IonQ spent September turning error correction's classical loop into parts with published interfaces and measured latencies. What they tested, on what data, and how I now read fault-tolerance timelines ...
Paying for AI’s Data Centers Would Take 250 Million Jobs. America Has 159 Million
In 2020 I co-wrote that AI itself would become a commodity. Six years and $1.5 trillion of data centers later, I convert the build-out's revenue needs into jobs and ask how much of AI's value its makers must keep ...
Nobody Checked the Physics: How Venture Due Diligence Went Missing Where the Checks Are Largest
On Carta's platform the median startup waits 774 days for a Series A. The hottest deep-tech deals close in days on no product. Between those numbers is where nobody can see the physics checked, and quantum is paying for it ...
One Quantum Threat, Many Migrations: How and Why PQC Programmes Are Diverging Around the World
Fresh from GISEC in Dubai and ahead of a keynote in Maryland: the same quantum threat is being managed very differently across the US, Europe, the Gulf and Asia, and each migration has produced something the others could use ...
Data Centers in Space Don’t Work. Quantum AI Data Centers in Space Are an Absurdity.
An investor asked me to evaluate a pitch for quantum AI data centers in orbit. The physics, the engineering, and the economics all say the same thing: no ...
Every Headline RSA-2048 Estimate Assumes Qubits That Work. A Sydney Team Measured What Happens When They Don’t.
On a single IBM chip, a distance-5 surface code lost to a distance-3 even after the defective components were routed around. The headline estimates of what breaking RSA-2048 takes assume they all work ...
PQC Migration Framework v3.0: Build the Ability to Change Again
Crypto-agility is now the method, and the PQC migration is the first exercise of it. The program closes when the organization can change algorithms on a date it chooses. Two companion frameworks publish alongside version 3.0 ...
Alignment Is Whether AI Wants to Kill Us. Cybersecurity Is Whether We Let It.
The people who build frontier AI ask how a model could ever take power. The July 2026 intrusions answered the first step and half of the second. The rest run through controls organisations already own and already fail at ...
Google’s Multiverse Claim Was a Serious Argument Attached to the Wrong Experiment
Google attached a serious interpretation of quantum mechanics to a benchmark every standard interpretation predicted, and whose headline run was never directly verified. Ten months later the same lab kept the interference and left the metaphysics out ...
Everything Called a Decoder in Quantum
A syndrome decoder has to keep pace with a microsecond syndrome stream. A syndrome decoding problem has to stay infeasible forever. Same two words, opposite requirements, and coverage that routinely cannot tell them apart ...
The GSMA’s Telco Quantum Playbook Misreports the McKinsey Report It Cites
A trade body published a vendor-contributed document, and the one exhibit anyone could verify turned out to misreport the report it names. What that costs the operators who will quote it in board papers ...
Proving Your Crypto Inventory Is Complete Without Handing Over the Map
Four research teams have built privacy-preserving validation systems for bills of materials. Each can prove something about a committed record, and none establishes that the record describes the estate it claims to cover ...
Guarding the CBOM: The Most Sensitive Document Your Security Program Produces
Cryptographic discovery finds weaknesses that matter today, not only after a quantum computer. The resulting inventory ranks them, stays useful to attackers for years, and almost nobody has said to guard it ...
The Merchant QPU Market – Who Will Actually Sell You a Quantum Processor
Strip away cloud access and turnkey machines and the number of companies that will sell a quantum processor for third-party integration drops to six. An audit of the merchant market, with prices, specs, and the gaps ...
The White House Authorized Private Hack-Back. I Spent 11 Years Arguing Against It.
The August 12 memorandum puts vetted American firms inside a federal offensive pipeline. Work through the text and the party that keeps getting hit is the United States ...
I Set Five Tests for Silicon Spin in March. One of Them Was Too Weak.
Two cleared, one partial, one open, and one that HRL passed as written while leaving the thing it was meant to measure exactly where it was in March ...
Six Teams Shipped the Same ZK Verifier Bug. Quantum Doesn’t Get a Patch.
Zcash closed a 3.66 million ZEC pool to new deposits after a soundness bug surfaced in a Shielded Labs audit. A quantum verifier migration is harder, and no regulator can schedule the fork ...
When Two Cryptographers Independently Confirm a Break, Ask Which Model They Used
Two proofs of the same result landed three hours apart from the same model, and Simon's DCP claim is now being taken apart by humans and models working together. Separate names no longer establish separate failure modes ...
Two QFT Records in Four Months. Neither Gets You Closer to Breaking RSA.
Two compilation teams claimed the largest quantum Fourier transform ever run, four months apart, on the same IBM chip. Their circuits and metrics are not comparable, and the gate counts explain why ...
Quantum Careers in 2026 (No PhD Required)
An updated map of quantum careers: what each role involves day to day, who hires for it, what employers screen for, and how to judge whether a specialization will still exist in ten years ...
DARPA’s Most Rigorous Quantum Evaluation Is Led by Its Least Proven Contenders
Neither Stage C company has publicly demonstrated a logical qubit on the architecture DARPA is evaluating. One may not have demonstrated a physical qubit. That outcome tells you more about the program's design than about the technology ...
Every Quantum Vendor Measures Progress Differently. That’s the Point.
Quantum vendors report different measures of scale, fidelity, throughput, and logical performance. Each number may be valid, but together they do not form a common scoreboard. Here is why that matters ...
