People crave authentic, human-made music and instinctively reject AI-generated performances even when technically impressive, because the value lies in the realness of the artist, not just the sound.
Lex Fridman argues that even highly polished AI music remixes ultimately feel soulless, and that listeners fundamentally want authentic human artists rather than AI-generated substitutes for iconic performers. ✦ AI generated
Lex Fridman · Lex Fridman · 2026-03-01 · original ↗
starts at this moment · 108:27
For a soul remix, it actually lacks soul... when I listen to soul or blues, I think I really want, in that case, to know... I don't want an AI B.B. King, I want the real B.B. King. And, and I, if I know, if any AI is involved in the B.B. King process, I'm tuning out.
verbatim transcript · starts at 108:27
108:27- Yes. - For a soul remix, it actually lacks soul. But it made me think of, like, when I listen to soul or blues, I think I really want, in that case, to know... I don't want an AI B.B. King, I want the real B.B. King. And, and I, if I know, if any AI is involved in the B.B. King process, I'm tuning out. - Yes. - And the, I don't think I'm being curmudgeonly old dude
108:58in that. I think we humans want authenticity. - So when AI, when I first started making these AI videos, it started back in 2023, I made my first one, and I would take my phone, come up in the kitchen, I'd play a song, and then my youngest and Dylan... My youngest, Layla, and I have three kids, and my oldest, Dylan, as soon as I played it, "Why are you listening to AI?" And it's like, oh my God,
109:29instantly. It's like, how do you know? Oh, it has this ringing sound in the thing. So it took me probably about four or five days to figure out, "Okay, what are they hearing that I'm not hearing?" So I did it, I separated all the parts, and what they're hearing was the artifacts that are in the vocal reverb. That sound- that were... That made incomplete- It just couldn't do the ambiances correctly, right? Because it's trained
109:54on... A lot of these AI programs are trained on very low bitrate MP3s, right? So, they feed all this stuff in there. So, they're getting really inferior information on the in the training process, whereas now when they make these deals with the major labels, they'll get the multitracks, and they'll get high-quality WAV files to train from, right? And whoever opts in, they get the solo vocal tracks. You know, if Ed Sheeran wants to do it,
110:22or Drake, or whoever wants to give their voice to it, let it do its thing, and then get the royalties from it. I'm not saying that any of them are doing it. I'm just giving an example. But every time that I would do it, I could be down the hall, and I would play something in my phone just to see if they'll be like, "Why are you listening to AI?" They can instantly tell.
- ·AI soul remix sounds impressive but lacks soul
- ·Polish isn't the point; authenticity is
- ·Even subtle AI involvement changes the experience
- ·Wants the real B.B. King, not an AI version
- ·Any AI involvement makes him tune out
- ·Value lies in artist's realness, not just sound