The people who build all the solid, crucial parts of the infrastructure and tools that the vibe coders depend on. People who are pushing forward with new ideas. People who are building solid, snappy, delightful software. People who take joy in understanding the logic and creating solid, novel abstractions.

It is such a big category error to say "we're building simple apps without looking at code, therefore no one reads code anymore."

I'm not sure if this comment is meant as a little joke or observation, but I keep on seeing serious comments like this. It is serious anti-intellectualism to scoff at people reading programming books or understanding things.

Over the last few weeks, I've been eliciting a compiler for an experimental programming language called 5emel from Claude Opus 5.5. To give you the flavor, here's a randomly selected function from its compiler:

    # The table of the program's own functions, by their descriptors, which
    # the runtime searches for --memo-every and --memo-all.
    def function_data(:defs) =
      :rows = loop :next(:r = defs, :acc = []) (
        | [] = r; acc
        | [Def { :fname, _, _, _ }, ...:more] = r;
          next(more, [acc, "  .quad d_", fname, "\n"])
      );
      ["  .balign 8\n  .globl rt_functions, rt_nfunctions\n",
       "rt_nfunctions: .quad ", show(length(defs)), "\n",
       "rt_functions:\n", rows]

There's a lot of code in 5emel I haven't read; there's something like 15000 lines of TypeScript and JavaScript and 3000 lines of C. I've read almost none of any of that. I haven't even read all of the compiler, which is only about 1600 lines of code.

Despite that, I think I've been putting my extensive past reading of programming books to pretty good use in this process. 5emel is directed at finding a workable, though not optimal, general solution to one of the famously most difficult problems in computer science, and if it is successful, I think it'll be a real advance — and one that LLM coders can use as well as human coders. Better, maybe.

Maybe, six months from now, AIs will be good enough to devise better programming languages on their own so they can program things more easily. They don't seem to be able to do that right now, though, so my skillset doesn't seem completely obsolete yet.

However, I play go recreationally, even though AlphaGo already plays go better than any human ever could, and possibly better than any human ever will. Often go players consult KataGo or a similar AI to improve their game. So I anticipate that, even if my programming skillset is completely surpassed by AIs, I'll probably still enjoy doing things like 5emel (if I have the chance), just like I enjoy playing Minetest and walking around my neighborhood at sunset.

By the way, there's probably enough information in the above snippet to recreate all of the important parts of 5emel. I fed it to an Incognito Claude session and asked it to explain it, and it explained basically all of the significant aspects of the language, although it wasn't sure about some of them. You could probably elicit the rest of the compiler and runtime system from it, although you'll probably have to fight your way through a lot of the same problems I have.

I still think the return-on-time from a good book is greater than a lecture or video or discussion. There is just so much refinement and information density, plus the ability to skip around to find interesting content is hard to match. (there is also something to be said for reinforcement of memory via spatial awareness while reading a physical book that is lost on a screen).

An interactive portal to a digital book could rival or surpass the experience, but I haven't encountered that yet.

Also, this guy:

https://www.reddit.com/r/vintagecomputing/comments/1amon8p/d...