It's a long way around but I think/believe we'll have some general purpose computing/systems research wins, that start to expose some computing as fun, interesting, casual, and from there we'll see a budding "heh, neat" attitude.
The atmosphere around computing had gotten hypersaturated by the intense capital flows & software heavy-industries that have built up. Serving the industrial need has been an overehelming priority. And it's really outshadowed much of the fart around & make fun stuff element that made it a scene of interest for so long.
I continue to tend to believe there's a ton of looming change in what computing will be when it grows up more. While most of that work gets concentrated in refinement & iteration & support of heavy-industrial software, we also keep finding/stumbling on stupendously more straightforward ways of doing things (react components/vdom, then hoc, then hooks). The ability of very small ideas to more capably express computing has enormous implications.
The web is interesting here because the page is a bit of a canvas, resculptable by anyone knowing a little introductory javascript+html+css. Whatever innovation computing has, ultimately most of it strongly contrasts, is esoteric, no matter how useful it could be. Almost all the work is sent through an extremely small funnel where a handful of people are at all familiar with the system. For bigger libraries, it might be hundreds. It's bad that most works are obscure & inaccessible, but worse, most software being opaque means rven though it's everywhere around us, we never can learn it. The world is obscure, it's mechanism happening at nanoscale levels we cant get to & which are still down-compiled assembly codes with most of the meaning wiped away.
Computing needs a kick towards extropic ends. Creativity will re-emerge, but only if there's an exosystem or ecosystem of computing, and right now we are small isolated islands, producing appliances, fixed systems, not computing software. Somewhere computing has to start letting users back in.