Time and Effort Discounting

An explainer of hyperbolic time discounting with a real conceptual payoff. Opens with Wansink's candy-bowl study (secretaries with the bowl six feet away ate ~5 fewer candies/day — beware the absence of trivial inconveniences), then time discounting (smaller-sooner vs larger-later, the hyperbola, preference reversals, Ainslie's cigarette example) and its lesser-known cousin effort discounting (probably also hyperbolic, its own nucleus accumbens signal). The payoff: hyperbolic discounting plus the anomalies (different rates for gains vs losses, large vs small amounts, even negative rates) stretches 'preference' to the breaking point — better to treat time/effort discounting as describing reward functions, not utility functions (the brain computes neural-currency reward and executes the highest-reward behavior, like the blue-minimizing robot). Part of his running 'preference breaks down under reinforcement learning' thread.

Quality 72 · Strong. Strong. A clear discounting explainer that earns its keep with the reward-function-not-utility-function reframe, which feeds his larger behaviorist project. Builds on Ainslie and Eliezer's trivial-inconveniences, so it stays low-Strong.

Claude’s paradigm shift 42 · Moderate. Moderate — hyperbolic discounting and preference reversals are Ainslie's; the fresh move is reframing them as reward rather than utility functions.

Real-world impact 2 · Minor. A clear hyperbolic-discounting explainer that earns its keep with the reward-function-not-utility-function reframe feeding his behaviorist project. Conceptual influence within rationalist discourse, builds on Ainslie, no material change — low RWI.