No, that far overestimates the speed and impact of natural selection. Having <1% of population die before breeding because of some factor (no matter if drug ov

No, that far overestimates the speed and impact of natural selection. Having <1% of population die before breeding because of some factor (no matter if drug overdoses or something else) is not a sufficiently big pressure to nudge towards an opioid resistance in just a few generations; if there today were some meaningful sub-population with such a resistance, an effect so tiny as the current overdose deaths might result in that resistance spreading over hundreds or thousands of generations; it would take a few more orders of magnitude more overdose deaths to have an effect in just a few generations. Similarly for hormonal birth control, especially if it fails, it has an above-average likelihood of being aborted. I mean, all these things can happen, but not after we're dead, but at the time scale of year 3000+ by which time I'd expect that medical technology would allow us to control these genes directly, making the direction of natural selection irrelevant.