Lessons learned

What has this exercise in treating the snarxiv as oracle revealed?

First, it confirms that the snarxiv could serve as an idea generator for physicists in search of inspiration. Particle physics contains a large number of possibilities, and one can easily discover whole new classes of model just by combining these possibilities. Perhaps no-one has ever thought to examine whether T-duality is possible in new inflation, or whether the AdS/CFT/unparticle correspondence can be adapted so as to employ the topological A model in the bulk, or whether cosmic rays might be generated by instantons in the interstellar medium, or whether neutralinos in a generalized RS1 scenario might show up as a TeV-scale fluid - but now, those ideas are out there, ready for the scrutiny of a skeptical world! This is the amazing fact: the space of possibilities in physics is extremely large, and it seems there is no end of work to be done in exploring them.

Second, it confirms that a few hours studying papers from the snarxiv might be a productive way to kill time. This may sound like the previous point repeated, but here I'm talking about a psychological perspective. Of course it would be possible to overdose - I am not about to conduct the experiment of reloading snarxiv and interpreting its output until I can't bear it any more - but employed judiciously, snarxiv may provide the weary physicist with a form of recreation which nonetheless stimulates their professional imagination in a helpful way.

I still find it somewhat disturbing that a mindless computer script can perform this stimulating function. But perhaps if I was a computer scientist, more familiar with the difficulties of large search spaces, it would seem less surprising.