> Is a robot taxi really cheaper to make than a manual car? If we only look at the future costs to supporting a consumer vehicle over the long term there could

> Is a robot taxi really cheaper to make than a manual car? If we only look at the future costs to supporting a consumer vehicle over the long term there could be significant savings by avoiding those costs. No need to overengineer - design for X km not Y years. No need to engineer for fixing some accidents - maybe cheaper to write off vehicle: removes infrastructure costs and reduces parts inventory. No need to support vehicle after Y years - reduce long term maintenance. Design for fleet refreshes every Y years. Designing vehicle to be written off on accounts (depreciate - no resale): paper gains matter hugely to corporations. Optionality... If robotic hardware is improving rapidly then may be cheaper overall to build new vehicles than upgrade older fleet. Design for recycling. Handover old fleet from US to Mexico - keep using old vehicles in jurisdictions where liability and modern branding is less important.