Alternative perspective: we were thinking to hire an M.Sc. thesis worker to do some work in Prolog and we definitely contemplated how to make the interview harder. The point was not to be monsters. Instead, Prolog is harder than JS and thesis is shorter than a few-year contract and there is less help on SO. If we hire someone who can barely write Prolog or understand what RDFS inteference is, the output will be nearly zero. Also, if the person struggles to complete something in 6 months, they will get anxious too and nobody wins. But obviously not a regular "hard" algorithms interview. Also, a senior engineer was essentially suggesting ways to completely rewrite my Prolog code in PRs for quite a while. So I hope you can appreciate why Clojure and other niche shops are a bit more selective.
Prolog is not harder than JavaScript, it is just different. I would hire somebody that is interested in Prolog and good with data structures and algorithms in general, and by good I mean knowing the nitty-gritty behind them. In my experience, exotic languages are not enough, they are not good indicators that you have a prospect in front of you.
