Having recently started on my journey through JavaScript, HTML, and CSS, I haven’t explored many other UI Frameworks besides Semantic UI. However, despite my limited and somewhat glitchy knowledge about Frameworks, I can say with absolute certainty that Semantic UI makes web development much less painful than coding in strictly HTML and CSS. While HTML and CSS are the essential backbones of creating websites, their cumbersome and time consuming downsides in the realm of aesthetics far outweigh any primitive want to code strictly in HTML and CSS.
Semantic UI allows developers to program their websites in the way that they would describe their websites to others, and the advantage of this is that it not only aids developers in remembering keywords, but also it aids in cutting down confusion between developers. For example, in Semantic, if you wanted a large, red button on your page, you could write ‘large red button’ and Semantic and more importantly other developers would understand what you had intended for the item to be. Coding strictly in HTML and CSS, there could be confusion unless there were specific IDs and classes given, but even then, not all programmers label their items in logical ways. Also, the format of the object could be different according to what one programmer considers ‘large’; however, in Semantic, there is a frame of reference given since ‘large’ is a predertmined size, so everyone who is familiar with Semantic will know what size object they should expect without needing further information.
Although I haven’t been exposed to other UI Frameworks yet, for the sake of code readability, team projects, and your own state-of-mind, I would highly recommend using one rather than developing strictly with HTML and CSS.