Book
The Sciences of the Artificial
by Herbert A. Simon · 1969 · 2 reading cards
2 cards
The Sciences of the Artificial · 1969
People and computers settle for “good enough” solutions because of bounded rationality
Simon called this bounded rationality: nobody can analyze absolutely every possible option before deciding. When choosing dinner, you do not compare every recipe in the world; you settle on the first reasonable option found quickly. Computer algorithms use the same strategy, searching for good solutions rather than perfect ones.
The Sciences of the Artificial · 1969
The environment the agent sees explains more of its behaviour than the model does.
"Context engineering" is the 2025–2026 name for what the model sees on each turn: system instructions, tool descriptions, retrieved documents, memory, the trajectory so far. A mediocre model in a well-shaped environment beats an excellent model drowning in a messy one. The levers are concrete: short tool descriptions with one example; retrieve only what is needed; compact the trajectory; put the constraint next to the decision; use the file system as memory instead of the context window. The typical failure is "context rot": stale tool output piling up until the model can no longer see the task.
“An ant, viewed as a behaving system, is quite simple. The apparent complexity of its behavior over time is largely a reflection of the complexity of the environment in which it finds itself.”