3 AI Tracks From Recent Years That Made the Internet Stop and Listen

AI music used to sound like something trapped inside a broken elevator. You could hear the idea, but you also wanted to apologize to your ears. Then, almost overnight, AI tracks started showing up with real hooks, real controversy, and real numbers. Suddenly, people were asking uncomfortable questions like: “Wait, is this song fake?” and “Why is the fake one kind of catchy?”

This selection was created as a culture-and-tech roundup by https://joi.com/chats, looking at how synthetic voices, virtual artists, and prompt-made music are changing online entertainment. And honestly, the weirdest part is not that AI can make music. The weirdest part is that some of it is getting stuck in people’s heads.

Below are three AI-related tracks from the last few years that show where the trend is going: viral parody, fake celebrity vocals, and full-on AI artist branding.

1. “BBL Drizzy” — King Willonius / Metro Boomin

Rating: 9/10  Best for: meme culture, hip-hop history, and people who accidentally dance to internet drama.

“BBL Drizzy” should not have worked as well as it did. On paper, it sounds like chaos: a comedian uses AI to make a fake retro-soul diss song, Metro Boomin samples it during the Kendrick Lamar and Drake beef, then the beat becomes a remix challenge. That is not a normal release strategy. That is the internet throwing spaghetti at a ceiling fan and somehow inventing a genre.

The original “BBL Drizzy” came from comedian King Willonius, also known as Will Hatcher, who used AI music tools to create a 1970s-style soul parody. Reporting around the track identifies Udio as the AI music generator behind the original sample, while Metro Boomin later turned it into the instrumental “BBL DRIZZY BPM 150.mp3.”

The numbers are not small either. Metro’s SoundCloud upload shows around 9.14 million plays, and the original King Willonius YouTube upload shows around 1.8 million views.

What makes this track important is that it did not feel like a sterile AI demo. It felt like a joke, a sample, a beat, a meme, and a cultural moment all at once. That is why it traveled so fast. It had the messiness of a human joke and the speed of AI production. The machine helped make the sound, but the timing, concept, and punchline were very human.

Programs/tools used: Udio for the AI-generated original track; Metro Boomin’s production turned it into a hip-hop instrumental.

Why it matters: “BBL Drizzy” showed that AI music does not have to arrive as a fake pop star in silver sunglasses. It can sneak into mainstream music through sampling, memes, producer culture, and remix challenges.

2. “How Was I Supposed to Know?” — Xania Monet

Rating: 8.5/10  Best for: people interested in AI artists that are no longer just internet jokes.

Xania Monet is one of the clearest examples of the “AI artist” era. She is not a human singer in the traditional sense. She is an AI-powered R&B persona created by Telisha “Nikki” Jones, a Mississippi poet and creator who writes the lyrics and uses Suno to turn them into songs.

“How Was I Supposed to Know?” is the track that made people pay attention. It is emotional, polished, very much in the adult R&B heartbreak lane, and exactly the kind of song that makes listeners argue in comment sections. Some hear a moving ballad. Others hear the beginning of a music-industry headache while wearing lip gloss.

The track has real numbers behind it. Kworb listed 14,098,420 Spotify streams for “How Was I Supposed to Know?” as of its April 30, 2026, update, while the official YouTube video shows more than 3.17 million views.

The controversy is part of the story. Jones has said she writes the lyrics from personal poems and then works through prompts in Suno, generating many versions until she finds the one that fits. That makes Xania Monet different from a lazy “press button, get song” stereotype. There is still human direction, taste, editing, and emotional input involved. Whether that makes the final result “authentic” is exactly the argument everyone is having.

Programs/tools used: Suno AI for music generation, with lyrics and creative direction by Telisha “Nikki” Jones.

Why it matters: Xania Monet shows the next stage: not just AI songs, but AI-branded artists with catalogs, fan bases, deals, and chart activity. This is no longer just a novelty clip. It is a business model.

3. “Heart on My Sleeve” — ghostwriter977

Rating: 8/10 Best for: anyone who wants to understand why the music industry started sweating.

“Heart on My Sleeve” is probably the most infamous AI track of the current era. It appeared in 2023 and used AI-generated vocals that sounded like Drake and The Weeknd, even though neither artist was involved. The track was uploaded by the anonymous creator ghostwriter977, went viral, and was later removed from major platforms after Universal Music Group objections.

Before it was taken down, the track reportedly reached around 600,000 Spotify streams, 15 million TikTok views, and 275,000 YouTube views. Guinness World Records also notes that the #heartonmysleeve hashtag had accumulated 52 million TikTok views during the viral peak.

Musically, the song worked because it sounded plausible. Not perfect, not fully human, but close enough to make people do a double-take. That was the scary part. It was not some robotic novelty with obvious seams. It felt like a leaked collaboration from an alternate timeline where copyright lawyers wake up screaming.

The exact software used was not publicly confirmed in the same clean way as Udio for “BBL Drizzy” or Suno for Xania Monet. Reporting described it as AI voice-model work trained or built around the artists’ vocal styles, which is precisely why it became such a legal and ethical flashpoint.

Programs/tools used: exact tool not reliably disclosed; the track used AI voice-cloning / vocal-model techniques imitating Drake and The Weeknd.

Why it matters: This was the warning shot. “Heart on My Sleeve” made the industry realize that AI music was not only about fake bands or funny experiments. It could imitate a superstar identity, generate huge attention, and create legal problems before breakfast.

Final Take

These three tracks show three different futures for AI music. “BBL Drizzy” is the fun one: AI as meme fuel, sample culture, remix bait, and internet comedy with a beat. “How Was I Supposed to Know?” is the business one: AI as a branded artist project with real streams and a serious push. “Heart on My Sleeve” is the dangerous one: AI as voice imitation, controversy, and a copyright nightmare in a hoodie.

The ratings are not just about whether the songs are “good.” They are about impact. Did the track move culture? Did it start arguments? Did it show a new use case? Did people actually listen? By that standard, all three matter.

AI music is still messy. Some of it is forgettable. Some of it feels like a demo made by a microwave with confidence issues. But the best-known tracks prove something important: listeners do not only care whether a song is made by a human or a machine. They care whether it hits, whether it surprises them, and whether it gives them something to talk about.

That may be uncomfortable for traditional music culture, but it is already happening. The next big AI track might be a parody, a virtual R&B ballad, a fake duet, or something nobody has a name for yet.

And yes, someone will probably say, “This is the future of music,” while someone else replies, “No, this is the end of music.” The truth is probably less dramatic and more annoying: it is both a tool and a problem.