Why subscribe to Model, Machine, Scale?

AI becomes a different technology when its predictions become physical actions.

A chatbot can generate another answer. A machine operating in the physical world must contend with uncertainty, latency, energy, hardware, safety and the cost of failure.

Model, Machine, Scale is an independent publication about how AI learns the physical world, becomes a dependable machine and scales into an economically viable system.

It is written for builders, operators, technical leaders, founders and investors who want to understand physical AI beyond research headlines, benchmark results and demonstration videos.

Who is writing this?

I’m Zeki Tekin.

My background spans large-scale technology infrastructure, infrastructure economics and AI systems. That combination shapes how I approach physical AI. A capable model is only one component of a much larger system. The important questions begin when that model must operate through sensors, compute, hardware, control systems and safety mechanisms—and when someone must determine whether the complete system can create sustainable value.

What you’ll get

Each substantial essay will examine one or more of three connected areas:

Model

World models, spatial intelligence, simulation and the methods machines use to represent, predict and reason about physical environments.

Machine

The complete physical-AI stack: sensors, data, models, planning, control, edge infrastructure, safety systems and operational feedback loops.

Scale

The economics of embodiment: compute, physical data, hardware, energy, maintenance, human intervention, utilization, capital requirements and unit economics.

The central question is always the same:

What must happen for an impressive model to become a dependable machine—and for that machine to scale?

Subscribe to receive new work directly, participate in the discussion and follow the development of physical AI from model to machine to scale.