Stuff that kills bacteria has to also not kill people.
I saw a TV programme where scientists had a big tray of nutrient gel. They divided the tray into sections. The fist section was normal nutrient gel. The next contained a normal dose of antibiotic. The next section was 10x, the next section was 100x and the final section was 1000x a regular dose. This was far more than humans could tolerate. They stopped at 1000x because that was the limits of solubility - they couldn't disolve more antibiotic into the gel.
Scientists then dropped a bit of e.coli on the normal gel and set their time lapse cameras.
In just a few weeks the bacteria had evolved to live on all sections of the gel.
(This was a BBC 'Horizon' programme called "defeatng the superbugs")
Worth noting: the bacteria at the end are not the same as the bacteria at the start.
Antibiotic acts on a particular mechanism to kill the bacteria. In order to overcome an antibiotic threat, bacteria have to either modify or drop any functions associated with that mechanism.
If any of those functions help them to infect humans, then the bacteria at the end of the process will be generally less able to infect humans.
I saw a TV programme where scientists had a big tray of nutrient gel. They divided the tray into sections. The fist section was normal nutrient gel. The next contained a normal dose of antibiotic. The next section was 10x, the next section was 100x and the final section was 1000x a regular dose. This was far more than humans could tolerate. They stopped at 1000x because that was the limits of solubility - they couldn't disolve more antibiotic into the gel.
Scientists then dropped a bit of e.coli on the normal gel and set their time lapse cameras.
In just a few weeks the bacteria had evolved to live on all sections of the gel.
(This was a BBC 'Horizon' programme called "defeatng the superbugs")