Executive Summary
Martin Shkreli disclosed he is building a publicly-traded photonic computing startup targeting NVIDIA's GPU dominance. His core thesis: light performs matrix multiplications (Matmuls) naturally through diffraction, potentially delivering 1000x to 1 million x performance improvements over silicon transistors. Unlike quantum computing, which only excels at integer factorization, photonic computing addresses the fundamental operation powering all AI workloads. Shkreli identifies a $5-10 trillion market opportunity with minimal startup competition—he can name every photonic computing company on one hand, contrasting thousands of AI agent startups. The technical challenges include achieving optical non-linearity and memory, which his team believes they've solved. On pharmaceuticals, Shkreli maintains high conviction that peptide therapeutics represent dangerous DIY medicine with no clinical validation, while defending GLP-1 drugs as genuinely transformative despite inevitable patent cliffs. He advocates focusing pharmaceutical innovation on rare diseases and severe cancers where million-dollar-per-patient value propositions justify development costs. Regarding Big Tech, he views Meta as exceptionally positioned due to distribution advantages, while expressing skepticism about Apple's innovation trajectory and Google's AI execution. The convergence of AI compute bottlenecks, Moore's Law limitations, and energy constraints creates an opening for fundamental hardware disruption.
Key Insights
what Martin Shkreli said“Light does Matmul's for free. So you have a Matmul machine in light. God kind of made light to do Matmul's whenever it diffracts. So you get the speed of light Matmul.”
This is a preview. Log in to see the full analysis including investment opportunities, risks, catalysts, and detailed insights.