As an outsider I assumed it took GM a substantial investment just to realize how far out of their depth they were. It made sense to cut their losses once they figured this out.

Having experience and capability to manufacturer cars has approximately zero benefit to create a self-driving software/sensor stack. It would make more sense for Adobe to create a self-driving car than GM.

Cruise was being operated as a separate company though. As a default, GM could have just not done anything and let Cruise operate as if it were independent. Any synergies (personnel, manufacturing expertise, etc) would have just been a bonus. And if they didn't want the financial exposure, they could have spun it out again.

Instead they chopped it up for spare parts, specifically, sending some Cruise personnel to work on deadend GM driver assistance tech and firing the rest. Baffling.

I remember GM cars in Herzliya, Israel with cables and cameras held by duct tape circa 2019 after Andrej Karpathy already presented end to end neural network training for Autopilot in Tesla. Looked like very late to the party.