Where was the proof presented for this? That idea doesn't sit well with me at all. Maybe they meant to say that it's hard for most people to want to sit still and stare at text all day. It's hard to ignore urges to go out rather than stay inside and read tutorials on the web. I don't think programming itself is hard. Most people program. It's just that they do it without thinking about the steps they're programming,…
The evidence presented is fairly poor, and consists of:
(1) failure rates of university first courses in programming for majors (which could indicate how hard it is to learn programming -- or how poorly designed university first courses for majors in the field are for the audience they attract.)
> I don't think programming itself is hard.
Actual coding isn't all that hard, once you have the foundations. But structured problem solving is a key foundation, and isn't all too often deliberately and effectively taught outside the context of programming (and, unfortunately, often isn't effectively and deliberately taught with programming, either, but programming seems to be the most common context where it is taught.)
> Most people program.
I don't think that's even approximately true. Most people do not plan out detailed step-by-step process from inputs to desired outputs that are expected to be successfully implemented by an execution agent that is perfectly literal and exact and exercises no independent judgment and has no ability to fill in the blanks and do what you probably meant given their understanding of the desired outcome even though you failed to specify it.
Even preparing step-by-step procedures for human agents is hard, not done a lot by most people, and done poorly by most people who do it, because most of them do so without good training.
System/process analysis skills that underlie both programming and procedure development may not be hard, but they are not skills that most people are taught effectively.