The quantum result (#164) published officially in Nature Wednesday: Google claims Sycamore’s 200-second sampling run equals ~10,000 years of classical compute; IBM’s rebuttal, published the SAME WEEK, argues a smarter classical approach (using Summit’s massive disk as extended memory) could do it in ~2.5 DAYS — supremacy demoted, per Big Blue, to “significant speedup.” The dispute is the healthiest thing about the milestone: an extraordinary claim (#082) receiving immediate, credentialed, adversarial peer review IN PUBLIC, both papers legible to any patient engineer. My file’s verdict stands with an amendment: Wright Flyer moment, yes — and IBM is correctly pointing out that Kitty Hawk had a stiff headwind. Both true. The 2.5-days-vs-200-seconds gap is still ~1000x, the classical goalposts will keep moving (they should!), and the real curve to watch is error-corrected LOGICAL qubits, which remain at approximately zero industry-wide. Check back in five years (#152’s Stadia clause: grade me).
Zuckerberg spent six hours before Congress on Libra (#158) the same day — the prediction file notes PayPal, Visa, Mastercard, and Stripe have ALL exited the association in the past month, which is the “partners peel at the first regulatory gunshot” clause executing ahead of schedule. He conceded Facebook would leave the association if it launches without US approval, which is a eulogy wearing a testimony.
And the Champions League group stage is delivering the strangest pattern: through three matchdays, away wins have dominated the group of death — Chelsea twice stunning Ajax and Lille on the road, then dropping points at home. Home-field advantage, the most-cited prior in football betting, currently returning NaN.
TIL: the disk-as-RAM rebuttal’s real lesson — “impossible classically” always means “impossible under the assumptions we bothered to optimize”; adversarial reviewers exist to attack the assumptions. Every benchmark is a fight invitation (#080’s deprecated metric, quantum edition).