More Than a Machine

More Than a Machine is a technical blog focused on AI systems and the engineering work that surrounds them. Rather than covering product launches or industry headlines, it examines how models actually behave in practice, where their limits show up, and what it takes to build reliable systems on top of them. The site is edited and curated by Nishaanth Reddy and collects writing across a small set of recurring concepts.

The content is organized by concept instead of by publication date. The agents section covers routing, orchestration, and protocol boundaries, with posts on coding agents, MCP layering, and running agents in sandboxed environments. The inference engineering section deals with serving, memory pressure, and decode performance, including explainers on KV cache, speculative decoding, and wafer-scale compute.

Each concept mixes different post formats. Field reports document real experiments, such as running models on a DGX Spark or tracking Reddit stock chatter over 100 days. Explainers build background knowledge on inference topics, while opinion and review pieces add perspective on tools and courses. This blend lets readers move from a concrete trial to the underlying theory.

The site is aimed at engineers, researchers, and technically literate readers who want more depth than general AI news provides. Some familiarity with machine learning and systems concepts helps, since the writing assumes readers can follow technical detail without extensive setup. Those ne