5 ideas
A decision-maker does not rank every alternative. They set an aspiration level, examine options one after another, and take the first one that meets it. If good options come easily, the aspiration rises. If search keeps failing, it falls, so the stopping rule adapts to experience rather than to a computed optimum.
Much of the apparent rationality in behaviour comes from the structure of the environment, not from the chooser's computing ability. When the needs that matter are few, can be met separately, and have acceptable solutions scattered widely, simple rules with short lookahead are enough to survive. What an organism needs to know depends more on its world's layout than on any general optimisation capacity.
Simon models a creature moving along branching paths on a plane where food heaps are scattered at random. It can see only a few moves ahead and must reach food before its energy store runs out. He calculates that survival depends on food density, sight range and storage capacity, and a modest vision radius proves enough. The creature survives by heading for any visible food and wandering otherwise, with no optimal route-finding at all.
An employment contract differs from a sales contract because the worker does not fix their future tasks. Instead, they agree to obey any instruction that falls within an agreed zone of acceptance. This suits both sides when the employer cannot predict which tasks will be needed and the worker cares little which task within the zone they perform. Authority is therefore a rational trade: flexibility for the employer in exchange for a wage premium and limits on what can be demanded.
A published election forecast changes voter behaviour through bandwagon or underdog effects, which seems to make public prediction impossible. Simon shows this is not so. If the reaction function is continuous and known, a fixed-point argument guarantees at least one forecast that, once published, produces exactly the outcome it predicted. Reflexivity complicates social prediction but does not make it impossible in principle.