Crypto's Quantum Challenges & Optical as the True Quantum-Class Winner – Martin Shkreli

What It Covers

Martin Shkreli joins James Altucher to analyze Bitcoin's quantum computing vulnerability, the case for stablecoins as censorship-resistant money, and why optical/photonic computing—not quantum—represents the most viable path to next-generation AI infrastructure, potentially reducing energy consumption by up to one million times versus current GPU architectures.

Key Questions Answered

• •Bitcoin's Encryption Risk: Bitcoin relies on elliptic curve cryptography, which quantum computing will eventually break. Patching it every five to ten years undermines the "hardest money" narrative promoted by advocates like Michael Saylor. Investors should treat Bitcoin as a speculative vehicle for excess capital—a "gasket" for surplus wealth—rather than a foundational store of value.
• •Stablecoins vs. Legacy Banking: Stablecoins solve a real problem that PayPal and traditional banks cannot: censorship-resistant, borderline-instant global transfers. A $100 PayPal transfer to Africa can trigger fraud flags and account cancellation. Stablecoins eliminate that gatekeeping. Regulatory opposition, framed as consumer protection, primarily serves incumbent financial institutions earning 1% savings rates while blocking 3–5% staking yields.
• •Optical Computing's Energy Advantage: Photonic chips perform matrix multiplication using light rather than electricity, potentially reducing AI inference energy consumption by up to one million times compared to current NVIDIA GPUs. For data centers already hitting grid capacity limits, this efficiency gain matters more than raw speed—a photonic chip running at equivalent speed still wins if the electrical grid is rationed.
• •AI's Perfect Match with Optical Architecture: Neural networks tolerate imprecise calculations—reducing floating-point precision from 32-bit to 16-bit, 8-bit, or even 1-bit produces negligible output degradation, as Microsoft's research demonstrated. This tolerance for inexact math is exactly what optical computing delivers naturally. Unlike cryptography, which requires perfect digit accuracy, AI workloads are structurally compatible with photonic imprecision.
• •Jevons Paradox and AI Energy Demand: Efficiency gains in AI compute do not reduce total energy consumption—they expand usage. DeepSeek's rumored 10–40x efficiency improvement will likely trigger proportionally greater demand for video generation, robotics, and enterprise reasoning tasks. AI currently consumes electricity comparable to France; unchecked scaling could approach continental-scale consumption, making energy-efficient optical alternatives a structural necessity rather than an optional upgrade.