Episode 6: A Four-Bit WonderThis episode shifts focus from finished machine behavior back down into binary, gates, and physical circuitry. It begins with John R. Pierce's August 1950 Astounding Science Fiction article "How to Build a Thinking Machine", written under the pseudonym of J. J. Coupling. The subtle aspect here is not only the machine idea itself, but Pierce's skepticism about inflated claims around thinking machines. Rather than letting the language run ahead of the machinery, he asks a narrower and better question: what would count, even in rough form, as a machine that learns? His answer takes shape in a maze: a constrained setting in which route choice, bias, memory, reinforcement, and revised behavior can be observed clearly. That gives us a bridge into the later maze-oriented work around David L. Heiserman's Rodney and the existing codebase discussion, not as a proven line of direct influence, but as a logical recurrence in how people think about adaptation in machines over time. From there, the episode turns to the four-bit wonder, a handmade wirewrapped protoboard built as a small test setup for timing, state, and memory. Using switches, lamps, a timing-divider section, and a 2114 static RAM, it provides a concrete place to inspect readback, retention, and even some power-on variability directly that reflects the randomness insistence from Pierce. The point is not to present a grand machine, but to establish a genuine place to begin experimentation with simple tools and modest parts before moving on to durable program placement in ROM. The time invested is so worth it.