I'm studying the historical cycles of AI development and want to better understand the so-called "AI winters." When did these major slowdowns actually happen, and what were the key factors that triggered them? I need a clear breakdown of the timeline showing when funding dried up and research progress stalled.