The first time I heard Freddie Mercury sing a duet with Michael Jackson, I had to pause. Not because it was technically impressive, the harmony was pitch-perfect or their voices blended like a vintage wine. I paused because that duet was like a dream coming true. Almost too good to be true. I remember sitting there, phone in hand, grinning like a teenager hearing something forbidden. This wasn’t a remix or a tribute act. It wasn’t someone doing impressions. It was them. See, as silly as it may sound, back then, these things actually had to happen in order to, well, happen. A “what if” remained exactly that. Now you say: “Well, thanks captain obvious.” But I’m serious. Because today, we’re officially in an era where AI can sing like anyone. Not just technically “in their voice,” but with enough nuance, emotion, and soul to make you lean in and say, “Wait… is this real?” And now that the tech is out in the wild, used by fans, labels, artists, and rogue bedroom producers alike, it opens a very specific door in my brain: dream duets.
I’ve always been fascinated by musical “what ifs.” What if Kurt Cobain had done a track with Billie Eilish? What if Amy Winehouse had harmonized with Adele? What if Bowie and Prince had made a slow-burning ballad together? Well, I made a list. Call it delusional. But here are the dream AI duets I’m itching to hear, and why I think they’d work, even if the real artists never set foot in the same studio.
Whitney Houston x Beyoncé – “I Will Always Run the World”
Tell me this wouldn’t shatter streaming records. Whitney, with her effortless power and operatic control, paired with Beyoncé, whose voice has grown richer and more emotive with each album. It’s not hard to imagine Beyoncé taking the first verse in that slow, deliberate style she uses on ballads, building up to Whitney sweeping in with the kind of chorus that knocks you back. You’d get velvet and fire. Control and explosion. Two women who mastered vocal precision and turned it into emotional weaponry. This one almost makes me angry we never got it in real life.
David Bowie x Lady Gaga – “Starman Born”
This isn’t even a stretch. Gaga has practically made a second career out of honoring Bowie, from her Grammys tribute to her glam-rock-inspired performances. But this wouldn’t be some cheap copy-paste homage. Gaga’s voice was made for this kind of pairing. Imagine an AI mashup where Bowie whispers the verses like he’s telling a cosmic secret, and Gaga belts the chorus like she’s announcing the birth of a new galaxy. It would be perfect.
Johnny Cash x Lana Del Rey – “American Sorrow”
This one gives me chills just thinking about it. Johnny’s baritone, aged and gravelly in those final Rick Rubin recordings, paired with Lana’s dreamy, haunting Americana? That’s a duet that writes its own movie. You could layer their voices over a sparse acoustic track with the occasional swelling of strings and slide guitar. Maybe something about lost highways, motel lights, and national disappointment. It would hurt in the best possible way. Who knows, maybe someone gets inspiration from reading this and mashes them up in some magical AI tool. You have my permission!
Kurt Cobain x Billie Eilish – “Smells Like Teen Depression”
Yes, the title’s tongue-in-cheek. But the concept isn’t. There’s something poetic about pairing the angst and detachment of Cobain’s lyrics with Billie’s whisper-sung vulnerability. Both artists share that rare ability to sound completely uninterested and deeply emotional at the same time. It’s that Gen X and Gen Z overlap where numbness isn’t apathy. I imagine Billie murmuring a verse, maybe even reinterpreting a Nirvana lyric in her style, before Kurt breaks through with one of those ragged, desperate choruses. Raw.
Elvis Presley x Frank Sinatra – “Blue Moon Rising”
You want smooth? You want class? AI can deliver a duet from the gods of mid-century cool. Elvis with his Southern charm and sultry drawl. Sinatra with that velvet phrasing and swagger. It would be a crooner showdown. Or maybe a crooner collaboration, because in my mind, they’re sipping whiskey in tuxedos, taking turns on verses, then harmonizing over a big band arrangement that swells like an MGM musical climax. You could release this one on vinyl just to keep the illusion alive.
Amy Winehouse x Adele – “London Rain”
This one hurts. We were so close to having it. Two voices from the same city, born a generation apart, yet carrying the same old-soul melancholy. The difference is in the edge. Amy sounded like she might cut you. Adele sounds like she’s already been cut. Their AI duet would be bluesy, maybe even jazzy. A slow piano ballad with swells of horns and a splash of gospel toward the end. Too bad it didn’t happen in real life.
Nina Simone x Kendrick Lamar – “Mississippi Fire”
I want this to exist more than I want most things. Nina Simone’s defiant, trembling piano. That voice that sounded like it had been forged in smoke and steel. And Kendrick, sharp as a scalpel, weaving rhythm into poetry. Together, they’d sound like a protest, a prayer, and a prophecy all at once. I imagine Nina pounding the keys while Kendrick delivers a verse that floats somewhere between “Sing About Me” and “Alright.” Then they both join in a call-and-response chorus, cutting straight into the soul. If AI can give us this moment, I’m listening.
Britney Spears x Madonna x Charli XCX – “Queen Energy”
Yes, I’m shameless. This would be an AI pop fever dream. Picture it. Charli brings the glitch-pop chaos. Britney delivers that unmistakable breathy charm. Madonna struts in with the icy confidence of someone who’s been ruling the charts since the Reagan era. Together, they could rewrite the rules of the ultimate girl power anthem. AI could stack harmonies that never quite existed, bend styles across decades, and make it sound like a multi-generational club banger.
Thom Yorke x Bon Iver – “Autumn Error”
Sometimes, I crave sadness. One that reminds you you’re human. A duet between Radiohead’s Thom Yorke and Justin Vernon of Bon Iver, generated through AI, could break hearts in subtle, quiet ways. Breath-heavy vocals layered over glitchy beats and minimal synths. A song that feels like waking up from a strange dream you’re not sure you want to forget. Even the mistakes would be beautiful.
Layne Staley x James Hetfield – “Ride the Dirt”
The moment I’ve heard first Metallica’s cover of Alice in Chains back in the day, I knew Layne and James should do a collab. James Hetfield’s signature growl kicking off the track with that metallic edge, part preacher, part executioner. Then Layne Staley slips in, spectral and wounded, threading harmonies like cigarette smoke through Hetfield’s iron riffs. This wouldn’t be a clean, radio-friendly track but a sludge and sorrow. A dirge with teeth. A song that sits somewhere between Dirt and Ride the Lightning, held together by mutual trauma and furious introspection. I’d want Jerry Cantrell’s ghost riffs in the background and a slow-burning chorus that hits like a freight train made of regret. Layne sings from beneath the surface. James answers from the edge of the pit. AI can’t feel the pain those two carried, but it can imitate the howl. And maybe that’s enough to give us one more glimpse of what could’ve been. One more reason to crank the volume and remember that even when they were falling apart, they still sounded like gods.
So, Is This Okay?
As much as I would love these duets to happen, and as much as these are technically achievable, there’s still one issue. Should we do it? It’s one thing to chat with AI companions and chatbots trained on internet data, spitting out stitched-together responses from the collective noise of online posts. But generating something modeled on a real person, their voice, their face, their presence, is something else entirely. That crosses from imitation into reanimation. Who knows, maybe one day AI-generated duets will be so common we won’t even blink. Maybe artists will license their voices like fonts or stock audio. Maybe a whole generation will grow up assuming “Beyoncé x Whitney” was a real song, not a generated experiment. Or maybe we won’t do any of that because of the ethical reasons. Because if you do it, who owns the song? Who made the song? But as a music lover, I can’t help it. I want to hear more. More fantasy. More fusion. More voices from different timelines singing in harmony.