Spotlight: Unrush — music that claims to sense your state before you search
Unrush is going after a fairly specific problem: the mental tax of picking the right sound for the moment you're actually in. Not "a lo-fi playlist" or "rain sounds," but something that adjusts as your state changes, without you having to search, skip, or curate anything yourself.
The positioning is the interesting part. Unrush's own framing draws a line between what it calls "Passive Search," the Spotify/Apple Music model where you type a mood and get a static playlist, and "Proactive Sensing," where the app is supposed to perceive your real-time state and generate an adaptive, effectively infinite soundscape around it. They're calling this "agentic AI," which is a bold word to put on a music app, and it's clearly meant to separate them from the crowded lo-fi-and-ambient-noise category rather than compete inside it.
That's a smart wedge if they can pull it off. Focus and sleep audio is a saturated space full of interchangeable playlist apps. Framing the product as sensing-driven rather than search-driven is a genuinely different mental model, and it's the kind of positioning that's easy to say and hard to prove.
And here's where I have to be straight with you: when we went to check the product firsthand, the live page we pulled didn't render, it just returned a "this page couldn't load" error with a reload button. So everything above comes from Unrush's own meta description and tagline, not from hands-on use of the app. We can't tell you what the actual listening experience is like, how the "real-time state" is detected (phone sensors? manual check-ins? something else?), what the sound generation actually sounds like, or what it costs. Those are the questions I'd want answered before writing a longer verdict.
Who should try it: if you're the kind of person who already reaches for focus or sleep audio daily and you're tired of manually switching playlists as your mood or energy shifts, this is worth a look, if only to see whether the "sensing" claim holds up in practice. Anyone curious about applied AI in wellness products, rather than AI as a marketing sticker, has a reason to poke at this one too.
Who should skip it: if you're happy with a good playlist and a skip button, there's no obvious reason to switch yet, the pitch is about reducing friction and personalization overhead, not about sound quality or catalog size. And if "agentic AI" language makes you skeptical rather than curious, that's a fair reaction until there's more to point to than a tagline.
Where I think this goes next, and this is a read, not a fact: the products that win in this space usually have to prove the sensing/personalization loop is real and not just a fancier randomizer, and they have to do it fast, because "AI-powered" is now the default marketing move for basically every wellness app. The wedge here (proactive sensing over passive search) is genuinely a better story than most competitors have. Whether Unrush can back it with a product that feels different in the first five minutes of use is the thing to watch. Worth checking back on once the site's up and there's something to actually click through.
Try Unrush: unrush.music
See the launch: Unrush on welaunch.sh