But most importantly it’s fun. I don’t program games much these days, I work on comparatively “serious” apps like compilers, but games were a deeply motivating experience that kept me going through the peaks and troughs of learning to program.

In 10 years, I have no doubt that a vast majority of business&game will be created without worrying, let alone touching the items you listed.

(I’d love to believe in such a future, but I’ve personally seen very little evidence to suggest what you’re claiming, so it’s hard to just take it on face value.)

Regarding the impact of Roblox, I think in the next 10-15 years (maybe sooner) we will see a game like application (maybe it will be called a virtual world or something else, but I'm guessing it will be called a video game in the mainstream) that has users and reach approaching a major social network

I don't think Roblox will be that thing, but it could

My thoughts are also that the primary business model for this will be virtual goods (content, cosmetics, etc) not advertising

Also regarding learning programming in games not teaching "serious business logic", programming is programming at some level. There are always domain specific challenges but many of the core skills are transferable

Most business applications are relatively easy to implement compared to games, although games are easy compared to system software and mission critical real-time systems

Going more out on a limb: I think that there is a chance that we will soon see the emergence of a mainstream (say, top 50) programming system that can be used for general purpose programming (apps, servers, business, art, game dev, etc) that is itself as aesthetically dynamic and engaging as a video game, and will look almost entirely unlike programming as we know it today (so — not text, and not node-and-wire). I think the existence of such a language would do more to advance our field than any effort to reduce the burden of manual state management or client-server interaction. Yes, it's possible to do better at those things, but that's just an optimization of our existing practice, not the opening up of an entirely new kind of practice.

For context, I'm the lead dev on a SaaS used globally by (eg) an S&P100 company for 6 years running. I write a lot of business logic, a lot of essential complexity, trying to quickly turn around features to keep clients happy. I want to do it using something so drastically richer than what we have today that you wouldn't even recognize it as coding. I do not want to switch to yet another text language in a static editor which forces me to play compiler in my head across a 100k LOC codebase, no matter what incidental complexity that language kills. Give me a kind of programming that lets me be alive inside my essential complexity, the way games let me be alive inside a system of play.

@Ivan Reese we should chat. I strongly believe your are describing what we are building.

JavaScript was considered a toy, Unity was a toy when it first shipped

I'm biased though as I work close to this space and was working directly on a VR programming environment built on top of Unity :)