The Duality of Theoretical Computer Science: Abstract Logic vs. A Box on Wheels

The top half is what you write in the paper to prove the halting problem is undecidable; the bottom half is the machine you're proving it for, which looks like it couldn't even decide what's for lunch

Only in theoretical CS do you slog through 40 pages of Kripke semantics just to confirm that a cardboard cube sliding left-right on infinite receipt paper can, in principle, run Kubernetes

The duality of theoretical CS: spending months proving a lemma about decidability in higher-order logic, only to realize the entire field rests on a glorified tape recorder that can barely handle a shopping list. Yet somehow that simple box captures all of computation - which is either deeply profound or suggests we've been overthinking this whole thing since 1936

The duality of theoretical CS: spending three pages proving a lemma about modal operators and type systems, then explaining the foundation of all computation with a box on wheels. It's the academic equivalent of deriving quantum mechanics from first principles and then drawing a stick figure to explain gravity - both are technically correct, but one makes you question if your PhD was worth it while the other makes you question if a child drew your textbook diagrams

Only theoretical CS: pages of modal logic followed by a box on an infinite tape; in prod the first question is how to shard the tape and whether the head needs Raft

Only in TCS do you spend 20 pages proving □(A↔B) ⇒ □(F(A)↔F(B)) and then explain computation with a shoebox on a tape - enterprise architecture diagrams usually invert that ratio

Theoretical CS: lemmas proving undecidability. Also: a Turing machine your distributed system wishes it could reduce to