Almost Alive's Evan Kanter is asking a question the music industry hasn't figured out how to answer yet.
- May 15
- 3 min read

We have spent several years now debating what AI can and cannot do to music. Whether it flattens or enriches. Whether it produces or merely mimics. Whether the human hand needs to be visible for something to carry genuine feeling. The conversation has generated considerable heat and considerably less light.
Then something like Deep Down arrives, and the argument shifts.
Almost Alive is the project of Evan Kanter, a New Jersey artist and producer who has, across a growing catalog of rock records, been quietly stress-testing one of the more interesting propositions in contemporary music: that instinct and algorithm, placed side by side and pointed in the same direction, can produce something that hits as hard as anything made the old way.
Deep Down — the first single from forthcoming album Undercurrent, due June 5th — is his most direct test of that proposition yet. Created using Suno AI and ChatGPT, it is a grunge rock track of genuine force. Thunderous. Driving. Emotionally unguarded in the way the best of the genre has always been. Not a simulation of feeling, but feeling itself, shaped by tools that Kanter has spent years learning to use with intent.
The reference points are canonical: Pearl Jam, Nirvana, Foo Fighters, Soundgarden. The sounds that first formed him as a musician — and the sounds, he has found, that AI production can render not as pastiche but as genuine inheritance.
The album title is worth sitting with. Undercurrent. The thing that moves beneath the surface — that has always been moving, even when nothing above the waterline suggested it. Kanter describes the emotional territory of Deep Down in similar terms: the things you believe you've left behind, only to find they've never fully gone. The unresolved. The persistent. The feeling that outlasts your belief that you're finished with it.
It's a subject that suits grunge well. The genre has always had a particular relationship with what won't stay buried — with the emotional textures that more palatable rock tends to smooth over. Undercurrent doesn't smooth. It excavates.
Where previous Almost Alive records — Hypnotica, Pulse — explored cinematic scope and hypnotic electronic depth, Deep Down strips that apparatus away. What remains is momentum. Guitars that don't ask permission. A rhythm section with somewhere to be. Vocals that carry the weight of the lyric without announcing it.
The AI question, here, becomes genuinely interesting. Because what Kanter is doing with these tools is not generating music — it is generating his music. The instinct, the emotional intelligence, the grunge DNA absorbed over years of listening and living: these are his. The algorithm is in service of something pre-existing, not producing something from nothing.
This distinction matters more than the debate usually allows. The instrument has never been the point. The point has always been what the player does with it — what interior truth they're trying to externalise, and whether the tool in their hands is capable of carrying it across.
On Deep Down, it is.
Almost Alive is, by Kanter's own framing, a project at the intersection of human intensity and machine precision.Born in Jersey. Built with AI. Powered by rock. It's a tidy formula that somewhat undersells the genuine experimentation beneath it — six albums deep, a growing catalog that has earned real traction on streaming platforms, a sound that has evolved without losing its core identity.
Undercurrent arrives as the project's most focused statement. Two further singles precede the full June release — Hit Refresh on May 8th, and a third alongside the album — each presumably tightening the frame around what Kanter is building toward: a grunge record for 2026, made with the tools of 2026, that sounds like it was always going to exist.
The future of rock, Almost Alive suggests, is not a rejection of its past. It is a return to its emotional roots — just louder, sharper, and built by whatever means necessary.



Comments