Trained on 1,000,000 more examples of natural language than code?

When the case occurs tens of millions of times within the corpus that is used to train the model, itâ(TM)s no longer a corner case.

I (with another engineer) recently designed a pump that needs to move a fluid in both directions. When starting, stopping, or reversing directions, the pump accelerates and decelerates along a programmable curve. The amount of fluid moved is the area under the velocity curve. It needs to be measured. To compute it, I take a definite integral. There, I've used calculus as a SW engineer. This was not the first time.

Code libraries that deal effectively with dates, times, and calendars has sophisticated support for daylight savings time: that they know when different regions change their clocks, how to convert between DST and non-DST time zones, etc. Will all that code break if it thinks the clocks are switching twice a year, but they aren't doing so any longer?