PREDICTING HUMAN BEHAVIOUR IS ONE OF AI’S HARDEST TESTS
LLMs can write sophisticated code, yet struggle to anticipate human decisions that depart from rational expectations.
People act on emotion, habit, social pressure and incomplete information. They contradict their stated intentions, change their minds and sometimes do nothing at all.
Behavioural simulation confronts this ambiguity directly: estimating which actions are likely, under what conditions, and with what uncertainty.
Our research team examines what it takes to build credible models of human action. Read more below:
https://t.co/0sQNGYM1Qb