Wild guessing, I can't see the future:
- I feel that Kotlin is a better Java. It's not much of a tradeoff if you're starting from scratch: many things are similar and anything else is better in Kotlin. It also compiles to Javascript and (soon) WASM, and can build native phone apps. That's :my guess/hope for JVM.
- For native, strongly-typed, low-level languages, I hope it's Rust. It has much fewer pitfalls than C++ so it's more accessible (web companies seem to hire many people with little experience). It's also great for parallelism, which is increasingly important. And, again, compiles to WASM. So for native, I would hope Rust, but I don't see native dominating backend.
- For dynamic languages, which I imagine will stay a huge portion: I hope it's Python, but I fear it'll be Node (although others may stay around too). It seems to have a lot of momentum, and the 'one language for frontend vs backend' seems to be more important than a properly designed language. And to be fair, it's improving somewhat.
Who knows, maybe WASM will eventually overtake Javascript, and Node will die along with it...
Note that job opportunities aren't only determined by how popular the language is. It's also about how many people have the necessary skills. People still earn good money with Cobol because no one knows it (or wants to do it).
So it's even harder to predict than just language popularity, which is already hard for 10 years.
If you just want good job security, don't follow my predictions, just learn something mainstream like Node, Python, Java or C++. They'll be around for the foreseeable future, even if they lose a little market share.