Most antibiotics are either compounds produced as defensive mechanisms by various organisms, or are synthetic derivatives of the same. That's given bacteria a h

Most antibiotics are either compounds produced as defensive mechanisms by various organisms, or are synthetic derivatives of the same. That's given bacteria a huge quantity of time to evolve resistance mechanisms, and the scale of horizontal transfer of genetic material across the entire bacterial kingdom means once we started using antibiotics ourselves it was easy to select for the genes that already existed. COE2-2hexyl is predicted to be an attack vector that hasn't occurred in nature before - based on the paper, the original observation was that in an attempt to engineer compounds to permit electron transfer across bacterial membranes, but in some cases these compounds had the side effect of inhibiting bacterial growth. They iterated through various versions and found one that simultaneously killed bacteria and had no obvious toxic impact on mammalian cells. An entirely novel attack mechanism means there's been no pressure for bacteria to evolve immunity at any point, so there's unlikely to be any handy plasmids sitting out there carrying an immunity gene and just waiting to be selected for. This doesn't mean it's impossible for immunity to be developed, but the suggestion seems to be that COE2-2hexyl interferes with well-conserved complex pathways. If there were simple mutations that granted useful immunity then the experiment in which they tested multiple generations of bacteria with increased mutation rates would probably have shown that up. That implies that it's more complicated - for instance, a mutation might reduce the effectiveness of COE2-2hexyl but in doing so might also reduce the fitness of the carrier so much that it doesn't matter. A mutation in another component of the relevant pathway might compensate for that, but you're now requiring both those mutations to end up in one bacterium before there's a fitness advantage. For a sufficiently complicated and well-tuned pathway, there may be no simple set of mutations that gets you to that point - any realistic path might involve passing through a phase of significantly reduced fitness, and that might be an insurmountable hurdle. Or the only realistic way to deal with this might be to evolve an entirely new pathway, and that's definitely not going to happen overnight. You're right that it's only a matter of time until resistance develops. The question is whether that time is on the order of years, decades, centuries, or millennia. A sufficiently long window means that we don't really have to care - either we'll be able to crank out alternatives by then, or we'll be in no position to produce synthetic antibiotics anyway. (Wrote my undergrad thesis on this, but that was over 20 years ago so please assume that I'm mostly wrong)