If you have been on TikTok or X recently, you have probably seen two people who are very obviously not rappers performing a rap song with unsettling accuracy. That is the Hotel Lobby AI trend, and underneath the joke is a format worth understanding — because the part that travels is not the video, it is the back-and-forth between two voices.
This page explains where the trend came from, why it works, and how to build your own version with audio you actually own.
Where the trend came from
The source clip is older than the trend. In June 2022, Quavo and Takeoff performed their song "Hotel Lobby" as the duo Unc & Phew on A COLORS SHOW. They stand side by side against the show's plain orange wall and pass verses back and forth over a single microphone hanging from the ceiling. The video passed 100 million views.
The AI version of it took off four years later, in September 2026. Early clips featured two cats recreating the performance; then a TikTok account posted a version recast with characters from the TV series Suits, which drew millions of views in under two weeks. After that the format spread quickly: film characters, political figures, historical pairings, sports stars, couples — the joke is always the same, two subjects who should never be trading bars together.
Quavo noticed. In late September he reposted the untouched 2022 clip and called it "The Original."
Why two voices beat one
The format works for a reason that has nothing to do with AI:
- The handoff is the engine. Every time the second voice comes in, the picture has a reason to change, so the cuts feel earned.
- Contrast does the rest. The best versions pair two characters who should never rap together. The bigger the gap in size, age or mood, the harder each line lands.
- Your ear starts predicting. Once you hear the pattern, you keep watching to see where the next swap lands.
That is why a solo rap video is flat by comparison: a duo has a relationship, and the relationship is the content.
One thing we would not do
The viral versions recast two real performers — and Takeoff was killed in 2022, which is why some fans said the meme felt wrong. Beyond the taste question, using a real performer's face or voice without permission carries personality-rights and copyright exposure.
Our suggestion: keep the format, drop the borrowed likeness. Invent two characters, write your own track, and you own the result outright — including the right to use it commercially.