Your mistake is believing that people invest in self-driving cars as a sort of charity or social service in the same way as drug clinics. Even if that were the case, I think there's a reasonable case to be made that solving self-driving cars would be a better lives-saved/$ investment than opioids. Even if it was a comparable investment, it wouldn't make addicts as socially-productive a group as crash victims--this would take tens/hundreds of millions more in social services.

Even neglecting that, crash victims are more worthy (on average) than overdose victims, as they come into their fate largely on accident, while overdose victims largely make an informed choice to roll the dice. I don't like saying this; it sounds judgmental and it certainly won't be popular, but this is the logical conclusion. Fortunately, the choice between self-driving cars and combatting the opioid epidemic doesn't remotely hinge on this point.