During the course of my career, the internet has evolved in some interesting ways. I found myself coming online during the late 90s. Netscape3 had just dropped. The Usenet was still a thing, websites functioned as basic sources of information.
The first half of the 2000s focused in different ways of turning the internet into a slick advertising friendly sales brochure. Waves of gold rushes came through. CSS changed a lot here. Social media added a propagation layer. Security became a concern as bad actors exploited everything from webforms to hosting providers.
Around 2006, web apps gradually became thicker, heavier, and less efficient as mobile became the main UI for the network. Some good things happened. Connections got faster, UI's became simultaneously heavier and less effective, while also becoming more effective and lighter in spots. It's a weird dichotomy.
By 2015 abstraction spiralled out of control, to the point where an entire new generation of developers forgot even the basics of responsible design. Client side payloads ballooned as all pretense of resource consumption disappeared entirely.
The upside is that this same period created a much cleaner separation of concerns, drawing a hard line between API's and user interfaces which was much needed.
After the advertising model
I think the next few years are going to usher in the end of the advertising model, except in video, where it's going to survive. The larger internet is going to slim down again, and contract to something more like its original form.
The death spiral of the client side continues today.
In the next few years, nearly all programming languages will disappear, replaced by vector systems, reasoning back end, and world models that connect to specialized back processes, markdown, and simple HTML.
Oddly enough, what was artificial about the internet will begin to fade away, and the original repository will re-emerge as AI becomes the defacto client side and intermediary layer for just about everything.
In the next year, there will be a Netscape moment, where agents become useful enough to be widely adopted.
This is the world we need to start planning for.
The internet isn't dying.
It's getting back to basics, and the centralization of it all is beginning to crack at the seams.
Letting some of the older more problematic monetization models die isn't necessarily a bad thing.
What to build for
To survive this shift, start thinking in terms of lightweight e-commerce, real engagement building, donation-based infocom, crowdsourced value creation, and B2B systems that deliver rich, actionable insight.
This is an opportunity for those willing to look ahead.
Everything is back on the table.