Antibiotics

Resistance isn't the bacteria getting cleverer, and it isn't the individual germ toughening up because it met a drug. It's evolution, running at a speed we can actually watch. A single bacterium can divide every twenty minutes or so; a population of billions can turn over dozens of generations in a day. In any population that large, random mutation guarantees that a few cells will happen to carry a change — a slightly altered protein, a pump that spits the drug back out — that lets them survive a dose that kills the rest.

Then the antibiotic does exactly what natural selection needs. It clears out the susceptible cells and leaves the field to the survivors, which multiply into a population that the drug no longer touches. The drug doesn't create the resistant trait; it selects for it. This is why finishing a course matters, and why using antibiotics for viral infections they can't touch is so costly: every exposure that doesn't finish the job is a training session, handing an advantage to whatever happened to survive.

For most of the antibiotic era, the surplus felt infinite, and we treated it that way. Doctors, facing an anxious patient and an uncertain diagnosis, wrote prescriptions for infections that would have cleared on their own or that no antibiotic could reach — a sore throat, a cold, a cough. The drug felt free and harmless, so the default was to reach for it. Multiplied across billions of prescriptions, that habit became one of the main engines of resistance, spending down a shared resource one reasonable-seeming decision at a time.

The larger drain is harder to see because it happens on farms. Globally, a large share of all antibiotics produced — by many estimates the majority — goes not to sick people but to healthy animals, mixed into feed to prevent disease in crowded conditions and to make animals grow faster. The same drugs, or close cousins of them, are used at enormous volume in settings where the whole point is low-level constant exposure. From an evolutionary standpoint, this is close to a worst-case design: exactly the conditions that select for resistant bacteria, which then move through the environment and into people.

What kept the antibiotic era going wasn't restraint — it was resupply. As long as new drugs kept arriving, resistance to the old ones was survivable. The shelf refilled itself. The reason the situation feels newly urgent is that the shelf has largely stopped refilling. Since the late 1980s, the number of genuinely new classes of antibiotics reaching patients has dwindled to almost nothing. We are still relying, in large part, on chemistry discovered decades ago, against bacteria that have had decades to adapt.

The cause is less scientific than economic, and it's a peculiar trap. A new antibiotic that works well is one a responsible health system will use as little as possible, holding it in reserve precisely because overuse would burn it out. So the better the drug, the smaller its market. Compare that with a medication for a chronic condition, taken daily for years, and the math is brutal: a drug company can spend a billion-plus dollars developing an antibiotic that, by design, it is discouraged from selling. Several firms that specialized in new antibiotics have gone bankrupt even after getting drugs approved. The reward structure actively points away from the thing we most need.

Fleming's warning in 1945 turned out to be less a prediction than a description of the whole system in miniature. Use too little, use it carelessly, and the bacteria come back stronger — he saw the evolutionary logic before it had a name. What he could not have seen was the second failure: that decades later the pipeline of replacements would run dry not for lack of science but for lack of a business case. The drug that stopped working, in the end, stopped being replaced.