Could I use the internet without a proper web browser?

After questioning my usage of Safari and wanting to stop using it, it remained my main browser, despite its flaws, including the lack of options when it comes to default search engines.

Before accepting my fate as a Safari user, I had tried other browsers. Because I still refuse to consider a Chromium-based browser, even if their legacy as battery hogs on the Mac may not be a thing anymore, my only real options were Firefox (or Firefox-based browsers like Zen Browser) and Kagi’s Orion, which is the only browser based on WebKit that is not Safari (I think).

These attempts at switching from Safari, while encouraging at first (uBlock Origin in particular is a great content blocker), didn’t work out for various reasons that I may get into in a future post. On the iPhone, Quiche Browser was very close to succeeding, but as long as it’s not also available on the Mac, I stick with Safari.

At first, I believed that my main frustration with these browsers was due to a lack of habits, that I needed to be a little bit more patient and, more importantly, indulgent, before being able to adopt them and replace Safari. Why expect perfection from other browsers when fewer flaws would still be an improvement?

But I was wrong. The issue with these alternative browsers was not that I’m too used to Safari; the issue was that they are just web browsers, all frustrating by essence. To quote Laura June from back in the golden days of Twitter, “Tip @Techmeme every browser is a piece of shit.” (this tweet is now long gone)

This is when I realised that the best web browser I could use, the one that would annoy me the least, would actually be the one I barely touch.

The web browser is my most-used app, whether it’s on my laptop or my phone, so how could I “barely touch” the web browser? How would this all work in this alternate dimension? Well, I’m glad you asked.

Let’s fall into the rabbit hole of how I would use a computer without a main browser, and consider the current use cases:

• Favourite sites: Visiting homepages daily

• Default link opener: Opening links from other apps (default browser)

• Bookmark management / reader: Reading, watching, and organising URLs

• Search engines: Helping me find answers

Favourite sites

RSS is great for many things, and I use it for something like 150 websites, but for daily news, it can be a bit noisy. For instance, the Guardian’s RSS feed would completely submerge all my other feeds, and I think that chronological news only makes sense in some contexts. A curated and well-organised homepage is a great way to solve this problem: the most important news of the day is near the top of the page, other news is organised by category, etc. Therefore, some bookmarks live on my new tab page for quick and frequent access to some of these homepages.

That use case alone would make the web browser a central app, right?

Well, it turns out that for these websites that publish a lot of daily articles, there is a pretty great alternative to their website homepage, one that even has the word “news” in its name: the newsletter. If I replace my news homepages with curated newsletters from the same publishers, I could, in theory, not need to use a browser to access curated, hierarchical news.

On top of these daily or weekly newsletters, I guess I could also use dedicated mobile apps or Safari’s web app, where I can add the site to the dock and forget that there is an actual browser behind the site. I could also subscribe to some of these newsletters via RSS, which would make my RSS reader my main and most-used app to use the web.

Both solutions would be easy to adopt, but I would miss having easy and fast access to the URL bar by pressing or Command + L or T.

Default link opener

Emails, web apps, and RSS feeds are all very good, but what happens when I click a link from a message, an email, or an RSS item? What happens when I need to type a URL? I thought about this, and for this use case, I guess MacOS’s useful LookUp feature — the little “web view” that can be activated by tapping a link with three fingers — could work; too bad it can’t be more easily customised and the window size seems doomed to be too small.

I can always use a minimal browser specially set up exclusively for these use cases and define it as the default browser of the system.

This browser would maybe have some sort of reader mode activated by default, or JavaScript disabled. In fact, it would be more like a “web reader” or “web player” than a web browser.

To keep this minimal browser out of view (no icon kept on the dock), typing and opening URLs would be done via apps like Alfred or Raycast, or the upcoming version of Spotlight.

Bookmark management / reader

For managing bookmarks and even reading some articles directly, an app like GoodLinks would be my obvious solution: I can add a link to it from basically anywhere there is a share menu. For articles I know I will want to read later, I wouldn’t even need to open the default browser before saving it to GoodLinks: I could do it right from an email or truncated RSS item.

For saved links that wouldn’t display well in GoodLinks’ reader view, the minimal default browser would come in handy, or even better, I could switch to a more powerful app like EagleFiler and save links in the Web Archive format.

For things like video streaming, I still wish some of the platforms would at least offer an app like Apple TV, but I guess the Safari web app solution could be good for sites like Netflix, Crunchyroll, or YouTube (which would make it interesting privacy-wise too, as Safari web apps have a separate set of cookies).

Search engines

Searching outside the web browser would, for example, mean using the ChatGPT Mac app, which is pretty decent as a Mac app, or an app Sindre Sorhus’ QuickGPT. And just like that, an “answering machine” would replace the traditional web browser for most, if not all, search requests. I know an increasingly large number of people around me add an A.I. app to their homescreen, which wasn’t really the case just a few months ago.

And it works great, at least for the users.

Still, I don’t want to do that just yet and let go of going to traditional search engine result pages, mostly because this is how I have experimented and loved the web for… check notes… decades.

Also, there is an ethical aspect that is increasingly hard to ignore with A.I., especially for web search. Mathew Ingram recently published a great and well-documented recap of the current situation.

On one side, A.I. companies are raising unprecedented amounts of funding to build their products. On the other, they charge users monthly subscriptions to use these products. Yet no one is paying for the source materials used to build these products, let alone asking where they come from or whether the extraction process might harm the source of said materials.

To me, it feels more or less like a colonisation of the web.

1. An A.I. company arrives in a place it doesn’t own and where it wasn’t invited (scraping the web without asking if it’s OK to do so).

2. They exercise total disdain for the local population and culture (stealing copyrighted works of art, putting pressure on third-party servers, with no financial compensation whatsoever, let alone offering a convenient way to opt out).

3. They exploit available resources thanks to their wealth and power (building LLMs with huge data centres and resources). In the case of the web specifically, they do so without caring about the impact. They may very well leave ruins behind, and they know it, whether they are ruins of business models or culture, but they just care about growth and revenue. Never mind that, in doing so, they make producing the original building material worthless, or at least far less valuable.

4. They justify their actions by saying they are improving the quality of life and basically stating that this is somehow the march of civilisation (so maybe instead of complaining and sounding the alarm we should help building an AGI, worship, and applaud).

Obviously, in this exaggerated view, things are looking quite bad: we can argue that this is how capitalism is supposed to work, which certainly doesn’t mean that everything is right and fair.

While I agree that copyright can’t really work in the case of A.I. training, there must be some kind of compensation for authors or a better way to credit the source. But despite the limitless amounts of money at their disposal, none of these companies have found a solution to this serious problem, nor even see this as a problem.

This is why, for now, as long as this view lives rent-free in my mind, I’m not very keen on embracing A.I. search for the web.

Proofreading, outlining, brainstorming: these activities are not really threatening the already sick and tired link economy and business models of most publishers. But search? It doesn’t feel right, not only in the deprivation of traffic this generates, but even as a web enthusiast, I don’t find this use of A.I. particularly appealing. It feels like using a service like ChatGPT or Mistral’s Le Chat for web searches would be like travelling the world and only eating food from the hotel because it’s more convenient. The same goes for A.I. results inside Google.

I’m obviously even less interested in using or even considering A.I.-focused web browsers like Comet or Dia, arguably the most up-yours move an A.I. company could make to the web it took so much away from.

It is also true that as long as Safari forces me to use either Google or Bing-based shitty search engines (sorry DuckDuckGo), I am more and more reluctant to engage in traditional search, and using a dedicated A.I. shortcut or app to start these searches instead of my web browser has never been so tempting.

Why am I getting so worked up about A.I. and search? After all, the web won’t disappear, just like print didn’t disappear when radio or TV arrived. It did take a big hit though, and changed forever. I guess the web too will evolve and morph into something new. Maybe for the better. Maybe for the worse.

Because A.I. search is increasingly replacing traditional search, most of the web we know today will disappear faster than previously expected. Good riddance I might say for a lot of it. If these A.I. “portals” and new “search answers” end up killing the mafia-like SEO industry, I’ll raise my glass to that. A lot of websites are awful and already doomed anyway, so why care so much about how the web could change?

Because a lot of the good websites risk disappearing too, or never getting started in the first place. Even if many publishers succeed at monetising without any search traffic, as they should, I think this will impact everyone indirectly. Behaviours change and internet users will end up spending less time browsing the web than they previously did, replacing some of that time with consuming its essence without actually visiting websites or subscribing to a new publication.

Manu published this beautiful quote this week, which reminded me why the web is so fun:

The internet is not kinda shit right now. Not even in the slightest. The internet was and still is a fucking awesome and magical place. It’s a place where you can browse thousands of blogs. It’s a place where you can press the button and get a useless website. It’s the place where you can admire rotating sandwiches or stare at a random park for a minute.

A.I. answers may very well feel magical on their own, but they won’t be able to provide this experience. This is why this “colonisation of the web” is worrying, because there is so much to enjoy and so much to lose. The upcoming lack of visibility and discoverability plague that will eventually kill a lot of publications and ideas, sadly, won’t differentiate between good or bad websites.

---

Right, where was I? Ah yes, could I use the internet without a proper web browser?

Imagining a hypothetical computer behaviour without the traditional use of the web browser raises a lot of questions in my head, as you can see. In theory, my plan of circumventing its central position could work, even if it would require some effort, some of which I will probably never consider. So yes, I could, but no, I won’t.

I already enjoy a good chunk of the web without needing Safari, with my RSS reader, my email app, and my read-later app, all providing a fantastic experience for what they do. I can’t say the same about Safari, or any other web browser. But I still need it, and, in many ways, I somehow still like it.

Will we soon access the web like we currently access email, meaning mainly not via a dedicated app but via an app that does other things? Like using an A.I. first app to search the web, just like we use the web browser for accessing email. Or starting the default system app that we don’t use that much, but works fine for that one time we have to do something specific, like using Apple Mail to send a non-professional email.

Or was it the plan all along? This would certainly explain the neglect Safari has had to endure in recent years. Using a dedicated A.I. app to search the web, or the Google app for instance, if search results don’t point to actual web pages (or barely so), could mean that browsers as we know them are doomed to change, and living their final years.