I agree here - this puts pressure on the company to make sure their drivers are (1) more efficient (2) make more $$ and (3) stick around for a while which then

I agree here - this puts pressure on the company to make sure their drivers are (1) more efficient (2) make more $$ and (3) stick around for a while which then goes back to (1) more efficient, and generates a cycle of improvement that benefits everyone. Of course this requires the company to make an upfront investment, but exactly as you said human capital is an asset, not a cost, so IMO it's a solid investment to make. The alternative is the current uber-model which is: (1) Focus on bringing as many people in as quickly as possible (2) try to pay them as little as possible as soon as you can which results in (3) high churn (probably disproportionately churns out the higher-quality folks who won't work for so little) and then leads you back to (1). This results in low quality and high recruitment costs and depletion of the labor pool. Uber did reach a solid state in most markets, but I would argue that they could have spent less on recruiting and retained a higher quality fleet for similar long-term costs, with the added benefit of not looking like a-holes for the past couple years.