Prediction: By the time it's economically feasible to re-write the infrastructure of the internet, coders will be automated.

Current Weight: 80%

Reasoning

The languages of the web (html/js/css/+ various frameworks) are -- in some ways -- an awe-inspiring feat of sophisticated engineering. And in other ways, they can be incredibly frustrating^[1][2]. It'd be cool if there were other options. But my understanding is: whatever the limitations of the modern web may be, the cost of redesigning outweighs the benefits.

However, the benefits (more dev-friendly, more resource efficient, etc)^[3] might start becoming more appealing in the future -- and the cost of switching might drop (because of the army of robot coders, which we'll refer to as, the army of robot coders).

But, by the time the army of robot coders can reinvent the internet, it won't make sense to write the internet for human developers -- because they won't be ones doing the development [~]. There will probably need to be some human-readable "description" of what the codebase is doing, but there would likely be pressure towards making the underlying language preferably friendly towards automated coders [~]. We may get 500 iterations of the "languages" of the internet, but human coders will only have ever directly interacted with the first (html/css/js).

I don't have a prediction on what the time period might look like; this is just a prediction about the ordering of events^[4].

[~] : debatable

---

1. see this hilarious take-down ↩︎

2. A setup using html, css, js, liquid, and mustache leaves you coding in 5 different syntaxes. (syntaces?) ↩︎

3. assuming these are legit ↩︎

4. though my current gut says 50 years ↩︎