Anand Iyer at Canonical Crypto brings a rare combination of deep technical building experience (coding since age 8, computer engineering degree, decade at Microsoft, repeat founder with exits) and crypto education expertise, having personally trained thousands through his DeFi masterclass before becoming an investor. He runs an intentionally small infrastructure-focused fund targeting early rebuilding of developer tools for crypto, while positioning himself uniquely at the AI-crypto intersection with his @ai handle and thesis around user-owned AI agents that will transact on-chain.

• Interest in the rapid evolution of autonomous AI agents and the shift toward automated software engineering. He closely follows how coding is transitioning from a manual craft to an agent-managed workflow, noting instances like Spotify developers shipping features without writing code manually or a researcher's agent that autonomously optimized its own performance to cut costs significantly.
• Focus on the physical infrastructure and custom silicon constraints underlying the AI boom. He tracks the hardware supply chain beyond just Nvidia, highlighting how Western Digital sold out of future hard drive capacity due to AI data needs and observing the trend of major tech companies and startups like Cerebras deploying their own custom chips for training and inference.
• Attention to the rise of open-weight models and the competitive AI landscape in Asia. He frequently notes that high-performance models from Chinese labs are rivaling closed Western systems, citing specific examples where models like DeepSeek or Qwen provide frontier-level intelligence as a low-cost commodity that US developers are eager to use.
• Enthusiasm for model efficiency, quantization, and small language models that run on local devices. He highlights breakthroughs that allow developers to bypass massive compute requirements, such as UnslothAI enabling heavy models to run on consumer-grade video cards, or medical systems that achieve high accuracy using very small models paired with structured data rather than massive computing power.
• A search for applied AI in complex, high-stakes verticals like science, defense, and finance. He looks for AI moving beyond chatbots into rigorous discovery, such as AI deriving novel physics results or defense agencies fast-tracking startup technology, and he specifically requested solutions for automating tedious financial workflows like venture capital fund audits.