I think everyone (or nearly everyone) should be taught a minimal amount of coding, not so that they can code, but so that they can appreciate what coding can do (and so they can decide whether they are interested in learning more.)
Here is a parable.
Lecturer was approached by Researcher. Researcher was working with DNA sequences and had received a large computer file with many thousands of DNA sequences. These sequences all had a few characters at the beginning and end which were artefacts of the amplification and sequencing process, and needed to be removed before the sequences could be used by the next stage in the process. This was the second such file Researcher had worked with - the previous time, Researcher had spent about a month editing the file in a text editor to remove the surplus characters. Now they dreaded having to do it again, and hoped Lecturer could provide a better way. Lecturer promptly solved the problem in under a minute with a one line Unix command.
Had Researcher had an idea of what programming can do, they'd have sought this help when they received the first file, and saved a month of extreme drudgery. (Incidentally, this really happened, my current boss was Lecturer.)
I present here (not for the first time) the Woodhams Hierarchy of Epistemological Categories:
1) Stuff that you know
2) Stuff that you know where to find out
3) Stuff that you know that somebody knows
3a) Stuff that you know that nobody knows (a category irrelevant to this discussion but important to scientists.)
4) Stuff you know nothing about
(Compare to the Rumsfeld Epistemological Categories.)
In the parable, 'how to best modify these DNA sequences' was initially in category 4 for Researcher, but would have been category 3 if they'd ever done some simple programming. The difference between category 4 and category 3 cost them a month. The difference between category 3 and category 1 cost them perhaps 20 minutes - instead of writing the one-liner themselves, they had to find somebody who could write it for them. This pattern is typical - when considering shifts in categories (from 4 to 3, from 3 to 2, and from 2 to 1) the benefit of shift 4 to 3 is greatest, and the cost (i.e. acquiring the knowledge) is lowest.
To be a functioning person, you need stuff in category 1, but people usually undervalue categories 2 and 3, which can cover very much more knowledge than you can fit in category 1.