Lea Verou
Last year, Kilian predicted that the way we write CSS was about to change forever, and this year's results validate his prediction.
Interop 2022 has brought browsers together working towards the same goals, and several features went from "this will never happen" to "this is now happening". Parent selectors (:has()) and container queries (@container) are two big ones in that category. For years it was hammered into developers that these could not happen due to performance implications, yet it turned out that both were possible.
Subgrid on the other hand was always on the roadmap, it was just a very big chunk of work. It is now supported by both Firefox and Safari, and being actively implemented in Chrome, so we will probably see much higher usage next year.
I predict that the stars of 2023 will be native Nesting, and color manipulation. Nesting is currently the primary reason people still use preprocessors, so being able to rely on it natively will be incredibly freeing.
Working with color in CSS is also about to get a lot more exciting. There is a swath of features being implemented across the board as we speak. By the end of 2023, we will probably be able to rely on wide gamut colors and device independent colors (color(), lab(), lch(), oklab(), oklch()), and we'd be able to manipulate them in a basic way through color-mix().
All of this together means that 2023 is likely to be a very exciting year for CSS. I look forward to all the amazing implementations to come, whether my predictions turn out to be correct or not. What a fantastic time to be writing CSS!