In our consumer society, we already expect to get monetized for anything that we happen to leave on the table. We can't afford space to have our music anymore, so it's been digitized and rented to us. We don't understand computers, so someone else owns one for us. We have trouble making friends, so platforms use us to track each other. But how is that relegation of control happening in the web as the browsers we use every day try to become all things to all people?

Originally, the web was simple and anyone could pretty much read through its code. HTML has always represented documents with things like sections, lists, or tables, and it also unlocked the potential of having links to other documents. Today, a website can still be something that you use, take a look at how it was made and you can modify it, even make something of your own. This hacking spirit that is now seen as fringe or even criminal is actually a core part of the web's success. 

Browsers started as rendering engines for text and media loaded from the network, but they have become small operating systems where you can interact with content on one side and investigate how it works on the other. All browsers come with dev tools and can be extended to allow users to change the behavior of any website. Even the address bar is something users can edit to avoid getting tracked or to share specific parts of a page. Even just having multiple tabs is a way to leverage different options for some important task we're working on. 

But we're again leaving all of that on the table. When we lock ourselves inside mobile operating systems, multiple tabs are hard to manage, addresses are hidden away and dev tools are not even an option. We even forget what browsers are and they become just a window opened and controlled by the facebooks or the tiktoks. Now we may feel like there's some browsing we should be doing, but we have no idea how. Maybe somebody can even browse the web for us. And that's where the AI hype slides in to apparently save the day. 

AI is another loop in this use-it-or-lose-it merry-go-round. Large Language Models (LLMs) can be trained in your data and run in your home computer (yes, the one most people don't have). The local version of an AI future may not be as powerful as the one we're moving towards, but having the benefit of trust is something that could unlock a unique potential for machines to truly assist us. Experts worry about AI alignment when it's primarily humans that are miss-aligned. We're not even giving machines a chance to truly help us as we're having them align with mega-corporations and then correctly expect them to not really work for us. 

Like many other AI advancements, it would be great to have browsers that browsed the web for us, if we could trust them 100% with all of our data. Trust them even 200% as they learn from our personal information and find better ways to help us. We would have no trouble having them monitor us as we browse the web to learn how we like to do particular things. As actual personal assistants, we would then trust their help a lot more because there wouldn't be anyone else in the room taking notes and controlling them. 

But things are coming together in a different direction. Google and Microsoft are moving their browsers towards "chatting with your tabs" and pushing search to become more autonomous. With browsers that browse themselves, the address bar that doubles as a search bar will probably now triple as a task bar. Smaller players like Opera, the Browser Company, Kagi, or Perplexity all have different ideas vying for the same push. The best case for users — the one with the least amount of ads and tracking — will involve paying for these new services. Most users can't afford to, but those that can at least incentivize businesses that work for their direct customers instead of other stakeholders. 

And coming last into this race we still have Apple. Their efforts with CoreML show that they're at least partially interested in making AI work locally on their hardware. But although Safari is a browser deeply integrated with their operating systems, strategic investments into software don't seem to come easy for Apple and their relationship with developers has degraded. It's doubtful if they can leverage being the last ones to come to the table and even that may still require new leadership at Cupertino. 

Finally, for those who want an independent browser engine, Firefox has no vision for the future, but a new fork of it called Zen is showing that there's still a lot of work to do if you love the open web. Meanwhile, Ladybird is maybe releasing and alpha version for early adopters in 2026 and Servo is focused on providing a solid WebView API that can be embedded in applications. The story of web browsers has entered a new chapter and although we are many pages in, the web is far from over.