The justification text on the website positions Mercury as an experiment targeted at "humane" factors so in that sense I think it is a concept of what could be. But in application I don't know who would be able to use this and for what purpose it would be useful for.

The idea of having a predictive fluent interface is a hot topic but this interpretation seems capped at creating and replying to messages and emails and even then it seems much easier to use their native counterparts as they offer complete utility that isn't hidden or requiring multiple windows (not to mention they exist within eco systems which do everything else we are accustomed to). For example in the Mercury email flow the user needs to search for and initiate separate modules to view a calendar, and another to find the location for an event from an email. Compare this to baseline existing apps like Outlook and Gmail which create calendar events that populate title, time, people and location from the content of an email in one action. On top of that they offer everything else we would expect to do with the information and actions associated with an email in one complete space.

If there was more detail around everyday behaviour that required more than inputting text and hitting send I might be able to assess it differently but the current use cases demonstrate that it is restrictive and slower than existing services for no benefit. I'm not trying to poo-poo this I believe wholeheartedly in experimenting but in the global application intent space I think projects like Dialog Flow are much more "humane" in their implementation. They do this by being flexible to exist within services and products at a native level which is much more democratic, and they do it invisibly for the sake of optimisation. I know comparing intent services with something claiming to be an intent driven"OS" isn't an exact comparison but there is a reason why there isn't anything closer.

I offer this critique after reading the Mercury medium article which makes some big claims in terms of what it achieves. The author sets out to solve the problems of traditional desktop OS interfaces and cites messy mac desktops cluttered with folders and application windows. He claims to have designed this solution for people with low attention spans and identifies himself within that category.

But the problem I have is that the end result does not offer a replacement for anything close to an OS level interface, and on top of that I doubt anyone would be capable of understanding how to use it. Maybe other people in the design space would understand its operation from a theoretical stand point but I just can't imagine anyone I know wanting to approach this, including those I know with lower attention spans. The author hints at this himself where he disregards user research to support the need for his design "...(where’s the data?, they ask), I stand by my words and my truth".

It bugs me because Mercury doesn't actually fix anything, its nowhere near an OS or even a complete app, and it's not that innovative to modify Apples notification centre design into a larger space, add Alfred command line shortcuts, and call it an intent driven OS validated by nine months of what is largely siloed work. Not once do we hear about testing it with a real audience, or see any user research or data to support any of the decision making. In nine months. Hell, you can't even undo an action! I just don't understand how the outcome can be celebrated given its goals let alone its execution.

okay, how do i launch a web browser? how do i play my games?

this says:

The clutter we take for granted in today’s operating systems can be overwhelming, especially for folks sensitive to stimulation. Mercury is respectful of limited bandwidths and attention spans.

but it doesn't say why clutter is bad. designer's have an unhealthy fetish with "dealing with" other peoples' "clutter". this idea has turned my whole device into a list of reminders basically -- useless.

and not to mention, the concept of "fog" is probably the dumbest metaphor you can use for a user interface. nothing is good about trying to navigate in "fog" -- it's disorienting and spaceless and frustrating.

No Apps or Folders. Mercury fluidly assembles content and actions based on your intentions. So you can focus on the destination, not the many ways to get there.

oh my god this is so stupid

Declaring the “Review Inbox” intention generates a space with all your unread mail and incoming communications. You can navigate between different flows in a Space by swiping up and down, or by using the up and down arrow keys.

I get a hundred emails a day. why are you obsoleting my mouse? what are you solving here? you're plowing me into my emails 1 by 1. i'm getting a migraine.

this whole OS concept is a tyrannical takeover of my digital life. i have zero freedom in this as a user -- I just have to hope your shitty AI can determine what "smart actions" to take on my to-do list, and that's supposed to be an operating system... okay.

Pros: This is absolutely beautiful. Seems like a lot of thought went into this UI.

Cons: My interpretation of this concept is like hiding an entire OS behind a hamburger menu or chat bot. Discoverability would be far worse. Users don't know what they can do without seeing it.

Relying on "intent" is dangerous and puts more mental strain on your users.

Surprised that this is so desktop/tablet focused rather than mobile, which is the future.

This kind of projects never work with real data or real life software. How an Excel file would look like? How can you do file transfers, etc. You know what I mean.

It a highly speculative concept, and for that I give this a huge thumbs up. However I would love to see a similar process done lean, including problem/solution fit and user testing. It would help him build a stronger case. But all in all, great fkn project for what it is– a speculative experiment.

looks like an Apple TV chatbot. don't see the appeal.

We totally need more poetic copywriters in tech.

• Dimmed UI > Kiri, Japanese for fog
• UI animations > Daoist Choreography
• Light grey drop shadow against dark background > ethereal glow of moonlight

Such beauty and elegance......lol

The menu text mixing with the content on scroll is triggering my OCD.

I always applaud these things for bringing new ideas and experimentation. But damn it cuts out a lot of the current desktop experience. Maybe it's not even an OS but like a monitor that hangs on your wall where you can manage todo's and maybe that's it.

Keeping the full-featured open desktop OS for other tasks.

Really wish that anyone who believes a messy desktop covered in windows is bad would actually spend an hour or so watching someone who prefers it use one.

Multitasking is not a bad thing or difficult.

I'm interpreting this as a command line with a GUI layer. I can see the appeal of that in some cases but for the most part, it seems like it will have the same hurdles as most smart assistance. They are a bit of a black box and it's very hard to know what they can or can't do and what the correct syntax is. Having to constantly consult the "more actions" for each new application/function seems like a bit of a pain. Especially on an OS level, there are so many commands and action that you can do at any time, the list would be absolutely massive.

Another thing that is very unclear is how you would actually access files. At some point, I will have to browse my files since most applications or functions tend to generate new files (word processors produce text files, image editors create images, etc). For example, I want to send some photos to my colleague, I have to find the photos somehow and maybe they are massive RAW files so I want to zip them. How do I find and select the photos I want to send? I can sort of imagine how the zip/compress would work but where does the resulting file end up? And after it is sent, will the zip be deleted or does it just sit somewhere on the hard drive where I can't see it?

Sounds like a cool concept and I like the focus on flow and simplification. I'm looking forward to seeing what else comes out of this but as it stands right now, as with most concepts, the more you look at it the more questions it generates :)

Just for clarification, this is done by https://jasonyuan.design. I just wanted to spread it to other people and see what others think about it.

This looks so clean, great job, not sure how well it would work tho. However the font used "Söhne by Klim Type Foundry", i can't seem to find this font anywhere?

Article: https://uxdesign.cc/introducing-mercury-os-f4de45a04289

This thing is clean AF Andrei... Good job!