Before antibiotics one could get an antiserum against each of many nasty infections. The rise of antibiotics displaced these drugs - even for some things (such as some forms of meningitis) where an antiserum against the particular organism, did a better job.
This actually made some sense. Antibiotics were broader spectrum, so (even after drug resistant bugs became common) you were likely to find one that worked in time to save the patient. Antisera, on the other hand, were very bug-specific.
If multiple drug resistance makes antibiotics nearly useless, perhaps it's time to revive antiserum use.
We now have the technology to rapidly identify the target organism(s) in a disease process, so we can rapidly select the correct magic bullets. And we also have the technology to make specific antisera by the bucketful.
And without the side-effects of making it by exposing an animal (like a "serum horse") to a pathogen and then (once it's developed an immunity) extracting the (horse-type) antibodies to this - and to everything else its immune system doesn't like - to make the drug. Instead we can make human monoclonal antibodies to just one target.
We can also engineer an immunization by chopping out the DNA for some conserved region snippet of some pathogen's accessible surface markers, splicing it with neighboring coding that will make the immune system take note and building it into an otherwise (and still) harmless bug - either to make an active ingredient for an immunization cocktail or a variola/polio style live-virus challenge. The bug has a very hard time evolving resistance because a conserved region of some component of its molecular machinery is usually conserved because has to be the way it is for it to work.
This is already being done to some extent. Seems to me it's time to stop crying about the end of antibiotics and focus on this set of approaches - which should be very lasting.