Ask HN: Does this old horse have a few years of useful work left?

So, my most recent stint at Microsoft just didn't work out. The tech was cool, but the day-to-day workflow was tedious and torturous. I only held out for 4 months - my shortest gig ever, but it really was the most I could muster. I really wanted it to work, but ultimately couldn't strike that bargain with myself.

So... now I'm back to living on couches and wondering / researching where I might still be able to be useful and productive in the world. It's not an easy question. There are more technology stacks than ever, and I'm suffering very severe paradox of choice. I sure miss the "good old days" when you just focused on writing a single app in C that ran on a well-understood machine, or even a single well-understood OS API. But those days are mostly gone. Now it's all distributed, multi-level conglomerates of varied frameworks and languages sort-of working together. I can't decide whether I have any interest in that. I sort of envy the people who can simply state, "I am a Rails dev" or similar.

I feel a lot of pressure to pick a niche. Something. Anything. Because living on couches gets old, fast, and even more dire financial straits await after that. So I'm being driven by the stick, and not the carrot - which is an unpleasant position.

I found a great game (on Steam) called TIS-100, where you progress by writing small, well-defined programs in assembly language, for a strange, highly constrained imaginary processor. It's great for its pure distillation of machine-level programming. There was a time when people used to get payed for doing what you do in this game. In other words, I fear I'm getting old, and my career options have just begun to seem uncomfortable.