I upvoted Peter's and Mayo's answers and would like to add this:
There are different groups of users. One interface can't possibly satisfy everyone or be optimally easy for everyone to use. Do user research and determine what the common shared qualities are within each user group. That will result in data-driven personas.
When you have personas, you can compare their main points, goals, frustrations, and scenarios to determine what the primary persona is. The primary persona represents the user group who won't have their needs met by an interface built for any of the other personas.
To use an example: what end user would want to try to read a WordPress blog through the site's admin dashboard every time they visited the site? Most users who are not technical or unfamiliar with WordPress would feel lost seeing this, and they probably wouldn't get any value from the site. That doesn't make them stupid. The site they are viewing just is not built to meet their needs. They just need to see the site from the perspective of an end user, not from the perspective of someone who works for the site they are trying to view.
Similarly, a WordPress developer who wants to be able to control the PHP code behind a page wouldn't be able to do that as easily from looking at the site from an end user's perspective as they would if there were (or were also) a text editor in which they could easily make their changes.
So in that example, I identified two personas: an end user and an admin with development skills. These are both examples of primary personas. They need different interfaces (a publicly viewable site and an admin dashboard, in this case) to be able to get value from the site and to do their work effectively (which are different goals).
However, you can't put all end users in the same category. I've never seen user research tend toward that.
There are other categories of personas too, adapted from this summary.
A secondary persona would have most, but not all, of their needs fulfilled by an interface built for the primary persona. The primary persona's interface could be modified to also meet their needs. A separate interface would not be required.
Supplemental personas are other personas whose needs are met by an interface built for a primary persona.
Customer personas aren't the end user but are responsible for buying the product. The interface should be built for the end user instead of them, but it is still important to consider their needs. For example, if our WordPress site is a game for kids, their parents would be the customers who buy products related to that game which would be offered on that site.
Served personas are affected by your interface even though they don't use the product you are designing. For example, for designers of software used in commercial settings, the managers of people using that software might not actually be using that software. But since that software will still affect the quality of work that their employees do and the amount of time it takes them to do it, their managers are served personas.
Negative personas are who you are not designing an interface for. For example, if you are designing an IDE for software developers, you are not designing it for non-tech-savvy people who are not interested in programming.
So no, it's not user stupidity. It's poor design, probably resulting from a poor design approach.
- Are your users S.T.U.P.I.D.? from Boxes & Arrows.
- Stop calling your users stupid from Everyday IX.
From there, it's a good segue into Peter's answer.