> On one side there is a massive amount of smart rust programmers who are super young in all the rust communities

I think actual Rust programers maybe in reasonable numbers but RIIR evangelists would be in massive numbers.

I can imagine fifth epoch engineers only doing code review and prompt engineering or handling major incidence once the LLMs handle the code writing.

My son is the only one in his class that has a PC.

There is going to be an astounding gap between those in the know and those who are not.

I've noticed the same. I'm convinced that the generation that is under 16 today will on average know a lot less about computers compared to 20 years ago.

Not my son, he's counting binary and hex, built his PC with me, ran network cables with me, is coding his own games for himself and his friends.

Assuming some form of ubiquity for AI tooling people like him are going to build amazing things.

There have always been and will always be outliers that are super into anything. My point was that the kids who didn't really care about computers when I was in school still knew a lot more about computers than the kids today who don't really care about computers

20 years from now your son is founder/CEO of a large gaming company and my son a union worker raising slogan outside offices "Free gaming credits are Human rights!"

Unfortunately your son is uncommon and for every kid who "counts in binary" there are dozens of kids in Juvenile Hall, on food stamps, no job, government handouts, sub-100 IQs. Your son may generate untold wealth and a portion of that will go to feeding the masses.

I can imagine first epoch engineers doing this, they are the ones who are in most need of giving their wrists a rest and less keystrokes.

Aren't Nth-epoch engineers supposed to be interviewing stakeholders and formulating requirements ?

I've also recently convinced myself that the 5th Epoch is coordinated attempt by marketing professionals inside of Silicon Valley companies to encourage a generation of programmers to not learn anything about computers or how to program them.

In the future the only thing you the engineer will do with a computer is type into a text area, hit the submit button, and be returned variations on strings of text that seem to say something important but upon careful considering do not mean anything at all.

We can only hope the next epoch will bring about some kind of enlightenment that will herald a new era of blog posts that convey information across the Internet.

Is it really necessary to learn things in depth when AI can do it for you? e.g. Recent Gemini 1.5 result of learning a language just from grammar book? If AI can do this, can't it be expected to excel at lower-level stuffs trivially which are more deterministic than a human language. After all, we don't learn / teach assembly as our first programming language.

> can't it be expected to excel at lower-level stuffs trivially which are more deterministic than a human language

Our current LLM's excel at nondeterministic problems (style transfer, summarization, etc), not at deterministic ones.

Not as the first language, but as a mandatory part of any respectable CS curriculum. Not unlike a software engineer that doesn't know (and doesn't learn) assembly, writing code with an AI producing output the human doesn't understand means that whole classes of bugs are un-debug-able. I have a little faith that AIs will be as reliable as modern compilers, but who knows what the future holds.