Why quantum has been "10 years away" for 30 years | E2316
Summary
This episode examines why quantum computing has remained perpetually "10 years away" for decades and what has changed recently to make progress more credible. The discussion focuses on the fact that quantum computers are not general-purpose replacements for CPUs or GPUs, but specialized machines for problems like optimization, chemistry, and certain cryptographic tasks. The guest explains that advances have come from both better algorithms and more realistic hardware assumptions, dramatically reducing the scale required for some useful applications. A major theme is reliability: quantum hardware is extremely error-prone, making quantum error correction essential for any scalable system. The episode also explores Yakumo's neutral-atom approach, the emerging international quantum supply chain, and how near-term commercialization will likely involve quantum accelerators working alongside classical data centers. There is also a cautionary discussion about the long-term cryptographic impact of fault-tolerant quantum machines, including risks to RSA and potentially Bitcoin-related security.
Key Takeaways
- 1Quantum computing is best understood as a specialized accelerator, not a replacement for classical computers.
- 2Progress has come from both improved algorithms and better hardware assumptions, which has lowered the practical qubit threshold.
- 3Error rates remain the central bottleneck, making quantum error correction (QEC) indispensable.
- 4Yakumo is pursuing a neutral-atom architecture and an integrated company strategy rather than a single-component business.
- 5The commercialization path for quantum is likely hybrid, with data centers using quantum and classical systems together.
- 6Fault-tolerant quantum computing could have major cybersecurity consequences, including threats to RSA and stored secrets.
Notable Quotes
""Quantum computing has been five to ten years away for the past 30 years.""
""We believe that we leads to like two of them each others close to two thousand tiny for example, in next four to five years.""
""So once we do quantum microclashes, they made a lot of error. For example, every thousand times, every hundred times, they make it.""
""AI for quantum and vice versa, quantum for AI.""
Episode questions
Why has quantum computing taken so long to commercialize?
The speaker says the main reasons are hardware difficulty, high error rates, and the need for error correction. He also notes that progress depends on both better hardware and better software/algorithms.
What kinds of problems are quantum computers actually good for?
They are best for specific classes of problems, not general-purpose computing. Examples mentioned include cryptography-related problems like RSA and complex scientific tasks such as chemistry and fluid dynamics.
How does AI help quantum computing today?
AI is already being used for quantum error correction and for estimating what kinds of errors occurred during computation. The speaker specifically references using transformers and classical models to support QEC.
What is Yakumo's business model?
Yakumo expects to sell quantum computers and later quantum-computing services, especially to data centers and academia. The company also aims to build integrated systems rather than just isolated components.