> An example of unnecessary generality is provided by the requirement that all procedures should be recursive. In ordinary computing -as distinct from symbol manipulation- it will be found that the need for procedures to be recursive is the exception rather than the rule, and the requirement that all procedures should be recursive leads to inefficiency, since a recursive procedure is both longer and slower than a nonrecursive one.
I wonder which things we currently optimise in the name of smaller and faster will seem as quaint six decades hence?
(I get three very different reverse translations of the above from three different engines, so in case it helps I had been attempting to riff on emergency exits vs calm hearts. originally inspired by https://news.ycombinator.com/item?id=42851495)
DeepL returned "Very much in the government, very much in the field."
Dulles seems to have been reflecting on a result of game theory that says it can pay off for an asshole in a positive sum game to attempt to take more than their numeric share of the positive margin off the table.
If kaizen is the way to approach local optima, do we have a single word to describe seeking global optima? (someone once said compiler optimisations would more honestly be called ameliorations; I guess that could make a "local optimum" also an "ameliorum" point?)
What's the usual connotation of 非? Thus far I only know it from emergency exit, so reading "not-" to form "unusual/abnormal exit" seemed perfectly logical.
(placeholder for a more analytic response. I wonder if the best sota non-assaholic LLM [loss-loss-models?] with accompanying pinyin would be able to translate your duilian to the basin of the global optimum
I wonder which things we currently optimise in the name of smaller and faster will seem as quaint six decades hence?