I’ve been a happy Arc user for the past few years. I found the user experience to be really effective, space management was quite workable, the vertical tab management with autoclose worked very well for me, and I found myself actually using the space bookmarks… and the browser worked. I don’t think I’ve ever used a browser that worked as well for me like Arc did.
I used past tense there intentionally. And Arc still sort of does work for me, but…
Recently, Atlassian purchased The Browser Company (“TBC,” the company that put Arc together) and emphasized future support for Dia, an “AI browser” and the focus of The Browser Company over Arc for quite some time now.
Arc’s been on life support since Dia was announced; the idea was originally that Arc would revolutionize browsing (and for me, it did) but the impact of AI changed things; The Browser Company created Dia as a heavily-AI-integrated browser and started focusing on that, with Arc getting updates primarily from Chromium. Arc worked - and works - but as a living, breathing, growing product, I think it’s pretty much done.
So with the purchase of TBC by Atlassian, I decided it was time to see what Dia was about. Downloaded, signed in, cranked it up, checked it out… and found it wanting, somewhat through unavoidable circumstances.
Part of the problem is that Dia is new. It’s still in beta, after all, so wrinkles are expected; there are extensions and things that don’t work yet. That’s annoying, but honestly, it’s fine - I’m a developer, that’s what happens with new and beta products. It meant I’d still need to rely on another browser for some operations (like Arc, or Safari, with Arc being the go-to), but surely that requirement would diminish over time.
The problem, though, wasn’t with extension support, but with the browser’s gestalt, or the way it was intended to be used. The whole design is around AI enhancement of human interaction, with tools to help you write and work.
On some levels, I get it. I use AI tooling. I run my drafts through AI for revision and tuning and analysis, for example. But it’s a deliberate act on my part: I choose where and when to engage the AI.
Dia might fully support that level of optional engagement… but the AI is pervasive, and AI use is expected and encouraged by the browser.
I don’t like that. Maybe I’ll change my mind over time, but for now, I find my interactions with AI to be best effective when they’re tightly controlled, and when I have my hands on the wheel at all times. I choose how to engage; I don’t choose which engagement suits me best. It’s an act on my part, not a filter.
So Dia, with its emphasis on AI engagement and its incomplete state where some features just don’t work yet… well… if I’m moving on from Arc, and I think I rather need to, it’s not to Dia yet. And may not ever be.
I tried the Zen browser, which promises a “calmer Internet” - I’m all for that! But Zen has some of the same problems of Dia - it’s not fully cooked yet, and while setup worked well and it had some features that were indeed reminiscent of Arc (the vertical tabs, for example), the lack of extension support made it, too, feel incomplete. Overall, I’d give it a thumbs up, but as far as serving as a live working browser, I don’t think it’s all there yet.
So now I’m back to Brave. I’ve had a number of people ask me why I didn’t just use Brave when I mentioned my browser explorations (or even before: “What, you use Arc? Why? Brave’s right there.”) but the thing is: I’ve used Brave. It’s okay. It’s better than some, worse than others, and it never felt like it truly fit how I browse; I don’t like the horizontal tabs (one of the things that made Arc so appealing), and it’s just… there.
But I’m willing to accept that my 18-ish impressions so far (this is not my first try of Brave) might have lacked commitment and immersion, so I’m back for try #19. It’s configured, it’s working, I’ve put in the various bookmarks and extensions I rely on, and I’m hunting for lifestyle extensions that make using the browser more tolerable for how I use a browser (autoclosing tabs that stay open and unused for a length of time, and so forth) and I’m trying to learn how to enjoy the tab groups as a replacement for Arc’s spaces. (So far: nothing doing. Tab groups suck by comparison.)
I will say that it’s snappy enough, and the extension support has been workable; there’re things that don’t work in Arc that are working fine in Brave, so that’s a plus, and the reported memory usage for Brave is much nicer than any of the other new browsers I’ve tried recently.
We’ll see how it goes.
I would love to see Atlassian look at Arc and say “Hey, this actually kinda rocks” and revive it as an actively developed product and fix some of the extension support issues where it struggles - but they seem all in on Dia, so my hopes in that area are quite limited. Zen might get there, too, but it has much farther to go than even Arc; at least it’s still being developed, I guess.
As a result, here’s hoping that I figure out how to fix Brave’s quality-of-life features for how I work and browse, or that I adapt.