General reply to express thanks for and respond to the several replies my first comment here received. First of all, I wrote in the original reply, "his warnings should be taken seriously," so I am by no means disagreeing with the quoted CDC expert about his general policy recommendations. Indeed, I am on record here on HN[1] saying that I'd like to see the United States follow the lead of the EU immediately in banning antibiotics used in human medicine (especially last-line antibiotics used for antibiotic-resistant human disease) from use on farms. That seems like a sensible way to segregate the ecosystems in which microorganisms develop antibiotic resistance by haphazard evolution through natural selection.
I respectfully note the disagreement of several participants here have with me. They disagree with my disagreement with the interviewed expert's casual language about thoughtful agency by bacteria. Noting your statements that this is just everyday conversational usage, I stand by my objection to that fallacious manner of speaking. Most people in the English-speaking world think erroneously about biological evolution and especially about the implications of evolution for human medicine. I am not alone in thinking that popular thinking about biology needs to be improved by rejecting the idea of organismal agency in evolution,[2] although it is remarkably hard to find this kind of careful thinking by a Google search amid the flood of webpages that specifically assert a purpose or intention to evolution by natural selection. Ernst Mayr, mentioned in one reply to my first comment here, is indeed a rare example of a reliable author on that issue. Human beings have a cognitive bias of attributing agency to inanimate objects,[3] and the reactions to my first comment suggest how deep-seated that cognitive bias is.
The entire revised introduction to the thirtieth anniversary edition of Richard Dawkins's book The Selfish Gene[4] appears to be available for free online reading, and Dawkins writes explicitly about what level of analysis can be useful when pretending that genes have intentions and goals while thinking about problems in evolutionary theory. The University of California Berkeley website about evolution has a good page about antibiotics[5] that makes helpful medical practice and public policy suggestions with more cautious language.
There is a testable prediction here of course: either bacteria outrun human antibiotic development or they do not. I'm confident that more bacterial lineages will be wiped out before human medicine is seriously compromised by natural selection of bacteria that are resistant to most antibiotics used in human medicine. The precautionary principle suggests that we follow the quoted expert's policy recommendations, but meanwhile conduct research for a deeper understanding of the biochemistry of harmful microorganisms and their ecology.
[1] https://news.ycombinator.com/item?id=5674781
[2] http://whyevolutionistrue.wordpress.com/2011/02/21/natural-s...
[3] http://cogsci.stackexchange.com/questions/3951/is-there-a-co...
[4] http://books.google.com/books?id=koaD_Aod_V0C&printsec=front...
[5] http://evolution.berkeley.edu/evolibrary/article/medicine_03
Two other comments at this level express the nub of the argument:
I'm slightly bugged by "desire" type language because it does actually cause problems in how laymen think about evolution. It's not uncommon to hear them make the mistake of thinking evolution will "think ahead". But the bigger problem I have with it is the "anything we do ... bacteria will eventually discover".
Some problems are just very difficult to evolve around, and it's hard to predict what they'll be even if you have a complete working knowledge of an organism's genome and biological workings. If the problem space doesn't contain a solution that can be climbed toward without sacrificing fitness in the short term, it is unlikely that it will ever be found by evolution.
Evolution is actually a really crappy optimization algorithm. The reason it has worked so well is not that it is so effective, but because it's the only game in town.
That is very well put, and deserves your upvote.
At what point do we start designing RNA with a delivery mechanism that attaches itself to a specific bacterial type and eliminates it?
A good question. It is an empirical question about the future, so it has no definite answer yet, but that is the general research direction to follow to get around the legitimate problem mentioned in the article kindly submitted here.