LHCb was designed to exploit the enormous beauty and charm cross-sections at the LHC. It has made several unique discoveries and has uncovered exciting anomalies in the current data.
Since its inception the LHCb trigger has faced a unique challenge: How to produce a meaningful data reduction from 30MHz of proton-proton collisions when a significant fraction of those collisions contain signals of interest to researchers. The trigger has evolved over time to meet the requirements of an ever expanding physics programme, and for the upcoming Run 3 has adopted three approaches that are unique at a collider experiment: A triggerless readout operating at 30MHz, a fully software trigger and reconstruction, and a fully GPU based first level trigger and reconstruction. In this seminar I will describe this evolution and some of the recent exciting physics results that it enables.