Functional magnetic resonance imaging studies demonstrate that subjects performing two steps of strategic thinking in the two-thirds of the average game exhibit elevated mentalizing circuit activity

Overview Debates Related claims Topics

A study by Stanford University and BetterUp found that 40% of United States desk workers received sloppy digital work monthly

A NeurIPS research paper demonstrated that artificial intelligence models developed by competing companies converge on identical answers

Stanford researchers identified typicality bias as human post-training evaluators rewarding artificial intelligence responses that feel familiar

As of 2026, Tutor Intelligence operates a dedicated fleet of 100 Sunny robots in Boston to collect research data

A player performing one step of strategic reasoning in the two-thirds of the average game chooses 33 based on the assumption that others choose 50 on average

In large-scale iterations of the two-thirds of the average game, choices of 33 and 22 appear far more frequently than choices of 1 or 0

Time discounting experiments evaluate intertemporal preferences by offering subjects choices between smaller immediate monetary rewards and larger delayed rewards

Strategic reasoning depth in cognitive hierarchy games correlates positively with individual working memory capacity

Level-zero players in cognitive hierarchy models make non-strategic decisions, such as picking random, salient, or lucky numbers

In the 11-20 game, players choose an integer between 11 and 20, winning their chosen amount plus a bonus of 20 if their choice is one below their opponent's

Amos Tversky and Daniel Kahneman introduced the concept of framing effects, where logically equivalent representations of an event alter decision-making

In a study by Eric Johnson, asking individuals their likelihood of surviving to age 75 produced probability estimates that did not sum to 100% with mortality likelihood estimates

The neural autopilot model posits that habit mode relies on retrieving the previous action and its absolute reward prediction error rather than re-evaluating options

In 2003, the California Institute of Technology acquired brain scanners dedicated to studying non-human primates, rodents, and humans

Optogenetic upregulation of reward prediction error neurons in rodents conditions them to consume aversive saline solution by artificially signaling positive reward

In the two-thirds of the average game, participants pick numbers from 0 to 100, and the winner is whoever is closest to two-thirds of the average chosen number

Action slips occur when an individual automatically executes habitual routines despite environmental changes requiring conscious intervention

Anthropomorphism is the tendency to attribute humanlike emotional and cognitive characteristics to nonhuman entities, including animals and artificial intelligence

The cognitive reflection test includes brain teasers, such as the algae pond problem and the bat-and-ball problem, that test impulse override against deliberate reflection

Auditory cues played during specific non-conscious sleep stages increase waking valuations of corresponding snack foods by approximately 10%