One of the biggest problems with measuring AI progress is the ambiguity of measuring intelligence itself. AGI is treated as a milestone we have yet to cross, but there is no central definition of AGI. Depending on who you ask, AGI is achieved when a system: * can fool humans into