If AI ends up being as transformatively disruptive to labor markets as suggested, it's more important than ever to consider how to make companies pay into workforce development systems they benefit from, so the costs of innovation are not merely - and inadequately - socialized like they usually are.

The bandaid is seemingly a win for all involved, but each workforce project comes at a high public cost per participant, and most people affected by these labor market changes do not end up the beneficiaries of these programs - as the # of slots are generally crafted to meet a given industry's needs

I think the why is pretty obvious: Well-meaning actors throughout get to boast about "X dollars spent" on "Y number of (marginally better) jobs" for "Z underserved populations" while simultaneously subsidizing corporations' training and recruitment, freeing scarce political capital for other issues.

Unfortunately, this tactic works time and time again because it's more politically popular for politicians and bureaucrats alike to default to solving issues of "skills shortages" rather than root causes like systemic disinvestment by employers chasing short term profits over sustainability.