AI summary by TwinMind
Surprise Progress in Quantum Computing
Sabine Hossenfelder
6m 59s
IBM Quantum Computing Breakthrough
- IBM unveiled new plan for commercially viable quantum computer using superconducting circuits
- Developed new error correcting algorithm reducing physical qubits needed for error correction
- Improvement cuts required overhead by approximately 90%
- Goal: deliver 200 logical (fully error corrected) qubits by 2029
- Later target: 2,000 logical qubits
- Commercial relevance threshold: ~150 logical qubits
- Current challenge: scaling from few hundred physical qubits to few hundred thousand
- Press release lacks details on scaling methodology
Alternative Quantum Approaches
- Sai Quantum building commercially interesting quantum computer
- Uses photonic qubits (less studied/understood than IBM's superconducting approach)
RSA Encryption Breaking Timeline Acceleration
- New preprint by Greg Gidney (Google Quantum AI, Santa Barbara) on archive
- Claims quantum computers will break RSA encryption earlier than anticipated
- Key finding: quantum computers with under 1 million noisy (not fully error corrected) qubits could factor 2048-bit RSA in roughly one week
- Significant reduction from 2019 estimates by same researcher:
- 2019: 8 hours with 20 million physical qubits
- 2024: 1 week with under 1 million noisy qubits
- Improvement due to algorithmic advances, not accounting for IBM error correction improvements
- Prediction: RSA broken before 2029 if estimates correct
Security Implications
- RSA encryption used for decades to encrypt digital data
- Includes confidential high-stakes information
- World transitioning to post-quantum cryptography (quantum-resistant protocols)
- Intelligence organizations likely possess sensitive encrypted data currently unreadable
- Data becomes readable with sufficient quantum computer capability
- Potential dramatic political consequences
Oxford Qubit Quality Record
- University of Oxford researchers achieved new world record in qubit quality
- Record: one error in 6.7 million operations
- Technology: trapped ion qubits (different from IBM and Sai Quantum approaches)
- Demonstrates rapid technology development across multiple platforms
Industry News
- IronQ CEO sold all company stocks despite months of public praise for company
Expert Commentary
- Author remains skeptical about useful quantum computers in coming decade
- Concern: more qubits may push systems into chaotic, difficult-to-control ranges
- Paradox: need quantum computer to determine when quantum computers become useful
