I think you've made a fair point, which I did have in mind but didn't write down - that there is a dichotomy between embedded and cloud systems, and the definition of software development will be to a lesser extent defined by the embedded side. Apple, for example, will have clout.
But long-term I think the cloud providers will assimilate so much power that the embedded side will follow its diktats.
Big iron mainframes have longevity, but will absolutely die out close to complete extinction - I've worked on those systems, I understand their strengths and the legacy issues, and there's no way that cloud isn't going to gobble up that market, it's just going to take a long time (as you say, beyond 2030).
But long-term I think the cloud providers will assimilate so much power that the embedded side will follow its diktats.
Big iron mainframes have longevity, but will absolutely die out close to complete extinction - I've worked on those systems, I understand their strengths and the legacy issues, and there's no way that cloud isn't going to gobble up that market, it's just going to take a long time (as you say, beyond 2030).