I can't speak to the particular browser application. I haven't installed it and probably never will, but the language around text interfaces makes the OP sound... uninformed.
Graphical applications can be more beautiful and discoverable, but they limit the user to only actions the authors have implemented and deployed.
Text applications are far more composable and expressive, but they can be extremely difficult to discover and learn.
We didn't abandon the shell or text interfaces. Many of us happily live in text all day every day.
There are many tasks that suffer little by being limited and benefit enormously by being discoverable. These are mostly graphical now.
There are many tasks that do not benefit much by spatial orientation and are a nightmare when needlessly constrained. These tasks benefit enormously by being more expressive and composable. These are still often implemented in text.
The dream is to find a new balance between these two modes and our recent advances open up new territory for exploring where they converge and diverge.
Am I the only one that interpreted OP in a way that they weren't opposed to neither CLIs, TUIs, nor GUIs at all? The topic wasn't "textual interface VS graphical interface", but "undocumented/natural language VS documented/query language" for navigating the internet.
In addition to the analogy of the textual interface used in Zork, we could say that it'd be like interacting with any REST API without knowledge about its specification - guessing endpoints, methods, and parameters while assuming best practices (of "natural-ness" kind). Do we really want to explore an API like that, through naive hacking? Does a natural language wrapper make this hacking any better? It can make it more fun as it breaks patterns, sure, but is that really what we want?
I'm not focused on this particular browser or the idea of using LLMs as a locus of control.
I haven't used it and have no intention of using it.
I'm reacting to the OP articulating clearly dismissive and incorrect claims about text-based applications in general.
As one example, a section is titled with:
> We left command-line interfaces behind 40 years ago for a reason
This is immediately followed by an anecdote that is probably true for OP, but doesn't match my recollection at all. I recall being immersed and mesmerized by Zork. I played it endlessly on my TRS-80 and remember the system supporting reasonable variation in the input commands.
At any rate, it's strange to hold up text based entertainment applications while ignoring the thousands of text based tools that continue to be used daily.
They go on with hyperbolic language like:
> ...but people were thrilled to leave command-line interfaces behind back in the 1990s
It's 2025. I create and use GUI applications, but I live in the terminal all day long, every day. Many of us have not left the command line behind and would be horrified if we had to.
It's not either/or, but the OP makes incorrect claims that text based interfaces are archaic and have long been universally abandoned.
They have not, and at least some of us believe we're headed toward a new golden age of mixed mode (Text & GUI) applications in the not-so-distant future.
CLIs are still powerful and enjoyable because their language patterns settled over the years. I wouldn't enjoy using one of these undiscoverable CLIs that use --wtf instead of --help, or be in a terminal session without autocomplete and zero history. I build scripts around various CLIs and like it, but I also love to install TUI tools on my servers for quick insights.
All of that doesn't change the fact that computer usage moved on to GUIs for the general audience. I'd also use a GUI for cutting videos, editing images, or navigating websites. The author used a bit of tongue-in-cheek, but in general I'd agree with them, and I'd also agree with you.
Tbh, I also think the author would agree with you, as all they did was making an anecdote that
s/Take/Pick up/
is not that far off from annoying its users than
s/atlas design/search web history for a doc about atlas core design/
is. And that's for a product that wants to rethink the web browser interface, mind you.
We are more rapidly heading towards (or already in) a future where the average household doesn't regularly use or even have a "computer" in the traditional sense, and a CLI is not just unused but entirely non-existent.
Anil is an old-school technologist. He helped Movable Type which was arguably the first blogging CMS. He also worked at Fastly and Glitch. I'm sure he knows how CLIs work and what they're good for. And surely no one here is suggesting that for 99% of normie users that they'd be comfortable with CLI just because it's good at piping one command output to another. Even those of us who are proficient use GUIs more often.
Graphical applications can be more beautiful and discoverable, but they limit the user to only actions the authors have implemented and deployed.
Text applications are far more composable and expressive, but they can be extremely difficult to discover and learn.
We didn't abandon the shell or text interfaces. Many of us happily live in text all day every day.
There are many tasks that suffer little by being limited and benefit enormously by being discoverable. These are mostly graphical now.
There are many tasks that do not benefit much by spatial orientation and are a nightmare when needlessly constrained. These tasks benefit enormously by being more expressive and composable. These are still often implemented in text.
The dream is to find a new balance between these two modes and our recent advances open up new territory for exploring where they converge and diverge.