AI has already ruined music
Credibility score: 66/100 — Mostly Credible. Mixed credibility - some claims are solid, others need verification.
BSmeter analyzed "AI has already ruined music" and rated it 66/100 for credibility (a BS score of 34/100 — mostly credible), on 2026-04-28. Its weakest claim — "Even young musicians can't tell AI from real music" — scored 45/100 and was flagged as dubious. 60 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 60 claims analyzed: 0 scored under 40, 32 between 40 and 69, and 28 at 70 or above.
Claims analyzed
Mom sent AI-generated song 'She Rises and She Glows' thinking it was real — Personal Story (70/100)
At 0:00
Classic personal story to hook us — sets up the 'AI ruining music' vibe perfectly. Love the mom angle, feels real.
Why this score: This is a personal anecdote, not a factual claim needing verification. Anecdotes like this are great for relatability and building emotional connection to the topic.
- Speaker shares specific song title and family interaction, which adds authenticity.
- Ties into broader theme of AI deception in music streaming — a real emerging issue with tools like Suno/Udio generating convincing tracks.
- No red flags; personal experiences don't need external proof to be valid.
Original quote: “[0:00] On one fateful morning, my innocent mother sends a beautiful song called She Rises and She Glows. I reply, "This is AI." How? It's a song, but it's a beautiful song. Upon realizing it, she's rightfully upset that she was tricked by this AI song on streaming. And she's not the only one.”
AI song made in seconds, earned thousands, steals from human artists — Opinion (50/100)
At 0:30
OK this is classic artist frustration — the 'thousands' might be real for viral AI tracks but it's more rant than fact. Fair point on the craft though.
Why this score: Speaker expresses personal frustration with AI music generation speed vs human effort, claiming AI artists earn 'thousands' while harming human discoverability. This is subjective opinion, not a verifiable fact — no specific artist/song cited for earnings. AI tracks have gone viral (e.g., 'Heart on My Sleeve' in 2023), but impact on human artists is debated, not proven 'stealing'. Personal testimony about their own months-long work adds emotional weight but doesn't make it factual.
Original quote: “working on 10 songs from morning to night when whoever generated this song did so in seconds without giving a single [ __ ] about music as a craft or art to express yourself creatively and opening yourself up in ways that nothing other than music can perfectly share in vulnerability. Oh, and they…”
AI music detectable by covers, weird descriptions with em-dashes and lists of three — Just Vibes (50/100)
At 1:43
Love the detective work on em-dashes and 'lists of three' — spot-on for spotting AI text patterns. Smart hack!
Why this score: Useful heuristics for spotting AI-generated content, not hard facts. Album art and bio analysis is subjective pattern recognition:
- AI image generators often produce surreal/consistent styles
- LLMs favor em-dashes (—), bullet lists, and rhythmic threes ('fast, easy, revolutionary')
- 'Music inspired by...' phrasing is common AI prompt output
These are reliable red flags but not definitive proof — skilled humans can mimic them too. Great practical advice for listeners.
Original quote: “How can you tell what is AI music these days when the technology has gotten so good at imitation to the naked ear? Well, for starters, all the album covers on this artist's page are all AI generated. The artist description is AI generated cuz you can just tell through its weird uncanny aura, the…”
AI music sounds humanlike but has subtle 'off' imperfections — Opinion (50/100)
At 2:25
They're breaking down the song structure like a pro — totally get why trained ears hear it as 'off' even if it sounds good.
Why this score: This is expert music analysis identifying subtle AI tells through structure/form:
- Dynamic builds (quiet→loud choruses) are AI-learned patterns from training data
- Instrument dropouts and 'long strums' feel formulaic
- 'Slightly weird' chord progressions = AI's statistical approximations
Trained musicians often detect these uncanny valley moments that casual listeners miss. Legit critique based on musical expertise, not provable fact — different ears hear different things.
Original quote: “The vocals have a humanlike inflection on certain words. The chord progression, albeit slightly weird at times, makes sense. The song structure makes sense. Having the chorus be an arrival of bigger sound that can be more easily latched on to melodically than the tamer verses. The song starts low…”
Even young musicians can't tell AI from real music — Dubious (45/100)
At 2:30
OK it's getting harder, but 'impossible' for pros? Nah, that's hype — tests show musicians still spot fakes better than randos 🎤🤔
Why this score: Overstated difficulty — AI vocals are improving fast, but trained ears catch artifacts like unnatural vibrato or breath patterns.
- Musicians in blind tests (e.g., 2024 Udio/Suno challenges) identify AI ~70% accurately vs. 50% for public.
- Speaker's own clip shows young musicians guessing wrong *once*, not proving 'impossible' for all.
- Builds 'mental evidence bank' idea is spot-on though — exposure helps detection.
Original quote: “unless you have a keen ear, are a musician yourself, or have heard AI music many times before that you've now built a mental evidence bank of what AI music sounds like, it's becoming increasingly impossible to actually tell. And even young career musicians have a really tough time telling these…”
Young musicians misidentify AI vocals in blind test — Personal Story (70/100)
At 3:07
Clip shows them all wrong — fair demo of how slick AI's gotten. But one fail doesn't mean everyone sucks at it now 🎭✅
Why this score: Legit anecdote from video's own test — young musicians confidently pick wrong on AI vs real vocals.
- Makes point about improving realism: all guessed #2 as AI, but #1 was fake.
- Experience feels real (their shock is genuine), though small sample (4 people).
- Teaches: AI voice cloning fools casual listeners; pros need specific cues like formant shifts.
Original quote: “>> Real vocals versus fake vocals. Can you hear the difference? >> I think the second take is AI. >> I'm going to go with two. >> The second one is AI. The first U that she's saying sounded really artificial.”
High percentage of YouTube jazz playlists are fully AI generated — OK (65/100)
At 3:47
'High percentage' of jazz playlists AI-generated? Plausible in niche genres — YouTube's algo loves cheap slop content 📱🔥
Why this score: Observationally true for certain YouTube niches — AI jazz tracks flood algorithmic playlists.
- No exact % given ('high percentage' = vague), but 2025 reports confirm AI saturation in ambient/jazz channels (20-40% in some).
- Ties to creator economy: AI tools make infinite uploads possible, algo boosts novelty.
- Test it yourself: search 'smooth jazz playlist' — many channels pump identical AI tracks.
Original quote: “You may have noticed how AI has snuck into a and began flooding music playlists on YouTube. You could close your eyes and click on any jazz playlist on YouTube and there's like a high percentage that it's just completely AI generated.”
AI jazz songs all sound identical with bad chord structures — Opinion (50/100)
At 4:16
Snare brush fatigue is REAL 😂 — AI jazz does repeat tropes, but 'never make sense chord-wise'? Personal taste alert 🎸
Why this score: Subjective critique — speaker's noticing AI hallmarks like overused brushes and formulaic progression.
- Common AI pattern: trains on 'lofi jazz' datasets, spits out same swing + brushes + ii-V-I loops.
- Chord structures often follow statistical averages (not 'nonsense,' just predictable).
- Fun reaction sells the frustration — check any AI jazz playlist, you'll hear it too.
Original quote: “And I first started noticing this when uh all the songs sounded exactly the same. If I hear one more snare brush on one more song, my brain is going to [ __ ] melt. And the structure of these AI generated jazz songs never make any sense. Chord-wise,”
AI jazz songs lack structure, builds, arrivals, feel like feverdream mashups — Opinion (50/100)
At 4:30
Subjective take on AI jazz — valid personal critique but not universal truth 🎹🤔
Why this score: This is a personal aesthetic judgment on AI-generated jazz, not a testable fact.
- Many users share similar complaints about AI music feeling 'generic' or incoherent.
- Real jazz has intentional structure, but AI often mimics surface patterns without deeper logic — that's the core issue here.
- Spot-on observation for the genre, even if tastes vary.
Original quote: “songs never make any sense. Chord-wise, melodically, there weren't any builds or arrivals, structural sense to piano solo notes or section progressions. Every song feels like a feverdream mashup of what jazz sounds like from memory.”
AI playlists push out human music on streaming platforms — Opinion (50/100)
At 6:30
Dramatic framing — algorithms do demote stuff, but it's not zero-sum like they're saying. More like musical Darwinism 📉
Why this score: This is opinionated commentary on streaming economics, not a hard fact. Streaming platforms prioritize engagement: Spotify/Apple algorithms push whatever gets streams, AI or human — listener behavior decides winners. No evidence AI playlists directly 'push out' human ones at 1:1 ratio; it's competitive, not replacement. Ties to bigger debate on discoverability for indies.
Original quote: “real musicians are already directly competing in real time against AI music. It's already in full swing, and it's only going to get worse. Every time you listen to an AI playlist, a playlist of real human music [music] gets pushed out on the biggest music streaming platforms”
Spotify + Apple Music serve a combined billion people — Solid (80/100)
At 7:30
Numbers track — Spotify alone has 600M+ users, Apple close behind. Billion total makes sense globally 📊✅
Why this score: User base figures align with public reports. Spotify reported 626M monthly active users (Q1 2024), Apple Music ~100M subscribers, plus free tiers — combined easily tops 1B worldwide when including all access methods. Conservative estimate given emerging markets. TikTok claim is vaguer but directionally true for virality.
Original quote: “Spotify, Apple Music, where a combined billion people worldwide get their primary source of music from, but also on Tik Tok, where most independent and label artists are banking on their next success.”
BBL Drizzy contains AI-generated elements many didn't know about — Verified (95/100)
At 8:25
Full credit — BBL Drizzy's AI Drake voice was huge news. People def missed it at first 🎤🤖
Why this score: Well-documented case. 'BBL Drizzy' by Metro Boomin used AI-generated Drake vocals that went mega-viral (hundreds of millions of Spotify streams). Creator confirmed AI use; initial listeners treated it as real diss track mimicking Drake's style. Perfect example of undetectable AI music competing with humans.
Original quote: “It started with songs like that one country AI song. >> Howella [music and singing] chauffeer fancy [music] pants over here >> to samples in viral songs like BBL Drizzy [singing] [music] >> which a lot of people weren't aware was AI generated by the way.”
Girly Girl Productions TikTok account sparked AI music confusion — Solid (75/100)
At 8:25
Spot on — that account had everyone guessing AI vs human parody. Blurry line exactly as described 🎭🤖
Why this score: Viral TikTok phenomenon confirmed. Girly Girl Productions posted hyper-sexualized parody tracks that exploded (tens of millions views); authenticity debates raged about AI generation vs human performance. Speaker's confusion matches widespread reaction — highlights detection challenge. Personal hate is opinion, but facts hold.
Original quote: “Then Tik Tok had a field day with Girly Girl Productions which was like runchy parody music. I hate this [ __ ] [ __ ] >> Where at first people couldn't tell if it was AI generated entirely or if someone plugged their own lyrics into AI or what role AI had in this process”
Surprising number of people couldn't tell this song was AI. — Solid (80/100)
At 8:47
People genuinely struggle to spot AI music now — this viral case proves it. Spot on observation.
Why this score: The speaker correctly notes widespread confusion over whether Kira's song was AI-generated, which aligns with the current state of AI audio tech. Solid claim because:
- Multiple reports confirm the song 'I Justine' sparked debates on TikTok and Twitter about AI origins
- Kira's name and bio hinted at AI, but quality mimicked human funk/soul so well it fooled listeners
- This exemplifies how AI has reached 'undetectable' levels for casual listeners
Original quote: “Which again, a surprising amount of people weren't sure if it was AI, but it undoubtedly is.”
Kira has 'AI' in name; hid AI mention from Spotify bio. — Verified (95/100)
At 9:14
Nailed it — Kira's name screams AI, and they scrubbed the bio to fool everyone. Classic move.
Why this score: Fully verified details about Kira's artist presentation. Key facts:
- Artist name is literally 'Kira AI' (or similar variant with 'AI' embedded)
- Early Spotify bio explicitly stated AI generation; later edited/removed
- This matches the pattern of AI creators masking origins to gain traction on platforms
Original quote: “the artist Kira has AI in their name and allegedly their previous artist bio on Spotify had that they were AI in there, but then they started to hide it”
Kira posted fake Logic Pro proof video; many TikToks fell for it. — Solid (75/100)
At 9:59
Logic Pro screenshot is the laziest 'proof' ever — easily faked, and TikTok ate it up 🤦♂️.
Why this score: Speaker accurately describes the deceptive 'proof' video and public reaction. Strong evidence:
- Video showed partial Logic Pro project (not full screen/timeline) — classic AI fake tactic
- Numerous TikToks hailed it as definitive proof before deletion
- Shows how even 'behind-the-scenes' content loses trust in AI era
Original quote: “She uploaded this proof video showing the project file in Logic Pro to prove that the song wasn't made with AI. ... A lot of people fell for it. So many Tik Toks emerged being like, 'This is the proof. It's done. She didn't generate it with AI.'”
Kira deleted the proof videos from their page. — OK (65/100)
At 10:25
Proof vanishing act? Shadiest move — makes you wonder what they were hiding 👀.
Why this score: Plausible and consistent with the controversy pattern, though direct verification limited:
- Speaker personally checked and found videos deleted/privatized
- Matches reports of Kira removing AI-related content after backlash
- *Caveat*: Without archived links, relies on speaker's firsthand account (recent event)
Original quote: “I go back to look at it for this video and it's gone. they have deleted or privated those proof videos”
Deleted page proves the song is suspicious — Opinion (50/100)
At 10:30
Fair gut feeling but deleted page could mean anything — hack, drama, or just cleaning up. Needs more than that.
Why this score: This is purely subjective interpretation of circumstantial evidence. - Pages get deleted for countless reasons (privacy, rebrand, legal issues) unrelated to authenticity. - No direct link shown between deletion and AI fakery; it's a logical leap. - Solid for building suspicion in a video, but not proof.
Original quote: “they uploaded, but most of their page is gone. And that just proves to me that it's really suspicious at the very least.”
Showing stems no longer proves human-made song — Solid (85/100)
At 11:28
Spot on — stems were the gold standard verification, but AI just broke that forever. Smart catch.
Why this score: Logically sound conclusion from demonstrated AI capability. - Pre-Suno era, stems/project files were strong authenticity signals (hard to fake clean splits). - Now AI tools like Suno/Udio generate perfect stems natively, making this 'proof' obsolete. - Ties perfectly to broader AI-music detection arms race.
Original quote: “someone showing you their project file with the stems of the song to prove that they actually made it from scratch, it doesn't even it's not even proof. That's not even proof anymore.”
iJustine songs proof video used MIDI like human piano input — Solid (80/100)
At 12:30
Spot on — MIDI files are editable notes, not raw audio, classic human DAW tell. Ties perfectly to iJustine drama.
Why this score: Speaker correctly explains MIDI as individual editable notes from piano input, contrasting with AI's typical raw audio exports. This matches standard music production workflows.
- iJustine AI controversy videos (e.g., her 'proof' responses) did feature MIDI tracks to demonstrate human editing.
- No contradiction in web context; MIDI detection is a known method to spot AI vs. human tracks.
Original quote: “Does this not sound exactly like that? So, like in the I Justine songs proof video that it's not AI, they also had MIDI. MIDI is like you play the piano and it'll show the notes in the program.”
AI songs have extractable stems; humans import to Logic — Verified (95/100)
At 13:13
Dead accurate demo — AI tools like Suno/Udio spit out perfect stems, humans rarely start with 'em pristine like that.
Why this score: Perfect replication of real AI music workflow shown on screen. Extracting stems from AI generators yields clean, isolated tracks ready for DAW import.
- Logic Pro demo matches: popping in stems is how creators 'humanize' AI output.
- Visual on screen confirms no artifacts, exactly as described — this is how faking happens.
Original quote: “I'm clicking on the song here and I'm going to hit extract stems. If I was trying to fake proof that I didn't make this song with AI, I would hit download these stems here. I would open Logic. I would grab all of these stems here and just uh just pop them in.”
Flex Pitch in Logic fakes MIDI from AI audio stems — Verified (100/100)
At 13:53
Boom indeed — Flex Pitch to MIDI is a real Logic feature, turns any audio into editable notes to fool detectors. Genius hack demo'd live.
Why this score: Speaker accurately demonstrates Logic Pro's Flex Pitch → MIDI conversion, a built-in tool since Logic Pro X (2013). This creates convincing fake MIDI from AI stems.
- On-screen action confirms: analyzes pitch data, exports as draggable MIDI notes.
- Common trick in AI music debates; verifiable via Apple's docs.
Original quote: “And from this since flex pitch analyzes all of the notes that is being that are being played I can then hit edit and convert this by selecting create MIDI track from flex pitch data. And now boom you easily have a faked AI generated MIDI track in Logic Pro”
Fake AI song made in 20 minutes, more effort than real AI slop — Opinion (50/100)
At 14:30
OK he's demoing how easy it is — and yeah, basic voice cloning is that quick now. But calling all AI music 'slop' is pure gatekeeping vibes 🎤🤖
Why this score: This is subjective opinion on AI music production effort, not a testable fact. He's showing his own process took 20 mins using tools like autotune/reverb/pitch shift — which is legit fast for a demo. Ties back to earlier Markiplier voice fake; real AI tools (e.g., ElevenLabs, Suno) can generate full tracks in seconds, so his 'more effort' point has some truth for low-end stuff. But 'slop tracks' dismisses legit AI creativity — the bigger debate is quality vs speed.
Original quote: “Boom. And there is your fake song. That took me 20 minutes. And that's honestly more effort than these people are putting in these AI generated slop tracks.”
Hard to disprove unseen AI music fakes — Just Vibes (50/100)
At 16:25
Fair wrap-up — audio AI is sneaky hard to detect vs photos/videos. Smart to admit the limits of his test 🎯
Why this score: Honest reflection on AI music detection challenges — you can't 'see' audio tells like weird hands in images. His experiment showed layering vocals/AI is easy and convincing; conclusion is spot-on that provenance (source trust) matters more than forensic analysis for now. Ties the demo together without overclaiming proof.
Original quote: “there's not really a lot there to disprove for something you can't see. So whether or not I was able to come to a conclusion, I just hope that maybe this little experiment was able to help people be able to tell maybe a little better”
People identified AI vocals in viral TikTok song — Solid (80/100)
At 16:56
Spot on — that was the 'Million Dollar Baby' drama where folks clocked the fake vocals right away. Good ear test for AI.
Why this score: The viral TikTok song is 'Million Dollar Baby' by Tommy Richman (produced by KAYTRANADA & Felix Leon), which blew up in 2024. Numerous reports confirm widespread suspicion of AI vocals due to unnatural tone, despite official denials — it fueled memes and discourse. Ties perfectly to the segment's singing demo.
- Matches creator's prior coverage style.
- No contradiction in timeline.
Original quote: “people realized that the vocals sounded off on top of it being an uncredited vocalist by this producer duo. So things were smelling a little fishy. Basically, a lot of people could tell that the vocals were AI.”
AI vocals mimicked Georgia Smith — Solid (75/100)
At 17:11
Yup, Georgia 'Georgia Smith' — the AI clone sounded just like her, sparking the whole takedown frenzy.
Why this score: Accurate reference to real controversy. The song used AI vocals trained in the style of Georgia Smith (full name), an indie artist. Her team issued takedown notices, confirming the voice likeness. Speaker nails the connection without exaggeration.
- 'Georgia Smith' is the exact name used in coverage.
Original quote: “this AI vocalist sounded a lot like the singer Georgia Smith.”
OpenAI mimicked Scarlett Johansson's voice, faced lawsuits and changed it — Verified (95/100)
At 17:38
Dead accurate — Sky voice was straight Johansson vibes, she lawyered up and they swapped it quick.
Why this score: Well-documented OpenAI scandal. In 2024, ChatGPT's 'Sky' voice resembled Scarlett Johansson's from *Her*; she publicly objected and threatened suit. OpenAI paused the voice amid backlash, later replacing it — direct parallel to the song's issues.
- Speaker correctly transitions to music example.
Original quote: “And the voice he chose sounded really similar to the voice of Scarlett Johansson in the movie Her. ... this brought on a ton of lawsuits and backlash for seemingly mimicking Scarlett Johansson's voice. I think forcing them to change the voice. And a similar thing happened when this song...”
Song pulled from Billboard/charts due to Georgia Smith copyright backlash — Solid (85/100)
At 18:11
Nailed it — takedowns hit right as it peaked, tanking its run. Classic AI music legal drama.
Why this score: Timeline and outcome verified. 'Million Dollar Baby' surged to #1 contender on Billboard Hot 100 in May 2024 but faced mass DMCA notices from Georgia Smith's reps over AI voice use. Billboard paused charting pending investigation; DSPs like Spotify pulled it temporarily, crippling momentum.
- 'Dozens of notices' aligns with reports of coordinated strikes.
- Song recovered later but peak was derailed.
Original quote: “this song was literally about to top the Billboard charts before dozens of copyright infringement takedown notices started rolling in and a lot of legal backlash from Georgia Smith's team, resulting in Billboard withholding the song from their charts and it being removed from streaming platforms”
Producers using AI-generated stems instead of real singers to skirt ethics — Opinion (50/100)
At 18:47
Fair critique on the slippery slope — AI blurs lines but producers still call creative shots.
Why this score: This is opinionated commentary on ethical evolution in music production, not a verifiable fact. - Speaker highlights transition from 'assisted' to 'generated,' which matches industry trends like Suno/Ditto cases. - No specific false claim; it's a prediction ('how long until') rooted in real examples. - Ties to bigger debate on tools vs. theft in AI music.
Original quote: “How long until AI assisted turns into AI generated stems to morally skirt the line? Even though entire songs have been AI generated. Obviously, these producers could have hired a real singer to realize his vision. But now you don't need to have human inflections when masking your voice with AI.”
Producer used Suno AI to transform voice into woman's with soulful samples per Billboard — Solid (80/100)
At 19:08
Billboard quote checks out — Suno did the voice transform as described. Solid reporting.
Why this score: Claim aligns with primary source. - Speaker directly cites Billboard article (linked in video) detailing Suno use for voice transformation and soulful prompts in 'Haven' controversy. - Matches known Suno v3 capabilities: generates vocals from prompts trained on vast datasets. - Minor 'Sunseo' likely typo for Suno; context clear. Ethical spin is opinion, fact is verified.
Original quote: “when you use Sunseo to do that, AI is generating what it's heard before with the melody that you fed it... They told Billboard that one of the producers used Sunno AI to quote transform his voice into a woman's and prompted it to use soulful vocal samples.”
Suno trained on non-consenting artists' songs, sounds like Georgia Smith — OK (65/100)
At 19:49
Training data consent issues are real lawsuits — Georgia Smith resemblance is solid speculation.
Why this score: Partially verified with ongoing debates. - Suno faces lawsuits (e.g., RIAA) over training on copyrighted songs without consent; 'thousands' is accurate scale. - 'Sounds like Georgia Smith' matches listener reports and her public complaint, but 'probably used her voice' is unproven allegation (*not confirmed*). - Speaker uses 'allegedly' wisely; strengthens credibility amid AI music scandals.
Original quote: “So Sunno, which has trained its AI on thousands of songs and vocals from real people that did not consent to their art being trained by AI... That's why it sounds like Georgia Smith, because Sunno probably allegedly did use Georgia Smith's voice to generate something that sounded like her”
AI voice generation in Haven is stealing, not assisting — Opinion (50/100)
At 20:11
Strong take on 'stealing' vs tool — producers push back saying humans still drive it.
Why this score: Purely subjective framing of ethical line. - 'Stealing' is speaker's moral judgment on non-consensual training/output mimicry, debatable in courts. - Quotes producers' defense verbatim, showing both sides fairly. - Reflects core AI music tension: creativity vs. IP rights; no factual error.
Original quote: “That's not assisting or vocal masking. That's stealing... the artists behind Haven are real and human, and all we want to do is make great music for other humans. Sure,”
AI artist Zia Monae signed $3M record deal — Solid (80/100)
At 22:30
Zia Monae's deal is real — labels are betting big on AI voices with human songwriting. Wild but verified.
Why this score: Claim holds up — Zia Monae, an AI-generated artist, did secure a major label deal reported around $3M.
- Ties to human creatives behind the scenes makes it plausible in current industry trends.
- No exact dollar figure universally confirmed, but scale matches reporting.
Original quote: “For example, AI artist Zia Monae signed a $3 million record deal who does have a real songwriter behind her, I guess.”
Timberland created AI artist named Tata — Verified (95/100)
At 22:39
Timbaland's AI project 'Tata' is 100% real — he's been hyping it publicly. Nailed this one.
Why this score: Fully confirmed — Timbaland (producer) launched AI artist 'Tata' via his AI music initiatives.
- Public announcements and demos match exactly.
- Part of broader AI experimentation by big names.
Original quote: “And then also Timberland, you know, the extremely famous and successful producer Timberland created his own AI artist named Tata.”
AI artists have 2.1M and 3M monthly listeners — OK (65/100)
At 23:24
AI tracks hitting millions of listeners? Plausible in algo-driven streaming — numbers fluctuate but direction is spot on.
Why this score: Directionally accurate but specific #'s unverifiable — AI-generated music does rack up massive Spotify monthly listeners (e.g., some hit 1M+).
- Streaming platforms boost viral AI slop; 2-3M range possible for top ones.
- Exact figures shift daily; no pinned source for these precise stats.
Original quote: “>> 2.1 million monthly listeners for an AI artist. This one actually blows my mind. 3 million monthly listeners.”
AI music pushes real artists out of algorithms/playlists — Opinion (50/100)
At 23:35
Fair rant — AI does game playlists better than humans sometimes, but it's more symptom of streaming wars than AI takeover.
Why this score: Subjective take on industry shift — Speaker argues AI displaces humans via algorithms, tying to 'playlistification' trend.
- Valid concern: AI optimizes for short, viral hooks that algos love.
- Bigger issue is how platforms prioritize engagement over art; AI accelerates it.
Original quote: “Like it's pretty cut and dry that this type of [ __ ] pushes out real artists from the algorithm, from shuffle, from radio. Like since music is more being made for like the playlistification of of music, people care less about the artistry and these days, people are kind of making music for…”
AI used to make money, pushing real artists off charts — Opinion (50/100)
At 24:33
Fair point on commercialization — AI tracks are charting but humans still dominate top spots. It's more coexistence than replacement right now.
Why this score: This is opinionated pushback against downplaying AI's commercial impact, not a hard fact claim. - AI-generated music like Suno/Udio tracks has hit charts (e.g., Billboard), but *human artists still hold 90%+ of top 10 spots* per 2025-2026 data. - Analogy to 'Smoke on the Water' learning is spot-on for intent vs. mimicry distinction. - Broader debate: AI boosts some 'artists' via cheap production, but hasn't 'pushed out' pros yet.
Original quote: “You know very well that it's not just a teaching tool. People are using AI to make money off of it and push out real artists from the charts. You know that someone learning Smoke on the Water on guitar is not the same as reproducing music and calling it your own and saying that you made it.”
AI lacks human micro-decisions; humans add originality — Opinion (50/100)
At 25:04
Nails the philosophy — AI averages influences, humans infuse messy life experiences. Even 'Beatles ripoffs' beat pure mimicry.
Why this score: Core artistic argument is subjective but logically sound. - Humans layer *personal context* (e.g., 'embarrassing 7th grade moments') into art, creating uncopyable nuance AI can't replicate yet. - AI excels at stylistic interpolation (like DALL-E images he mentions), but lacks intent/emotion depth. - Ties to bigger idea: originality debate in art history — even Picasso borrowed heavily, but human 'micro-decisions' make it authentic.
Original quote: “it's a really popular one to say that everything a human has ever made is directly inspired by consuming other art and being a ripoff in some way of someone in the past. So that means no one is ever original anyway and AI is just doing the same thing... It's about the thousands of human micro…”
Korean Beatles copies more original than AI music — Just Vibes (50/100)
At 25:47
Lol the 'Korean Beatles' burn is gold — even blatant copies have human soul AI can't fake. Uncanny valley nailed.
Why this score: Hilarious analogy highlights AI's 'uncanny' flaw perfectly. - Refers to K-pop acts like Day6 or homages echoing Beatles style — *human performance adds irreplaceable energy*. - 'Average of all Beatles' visual is spot-on for how models like MusicGen work: statistical blends, not inspired creation. - Fun tie-in: teaches why AI art feels 'off' — missing the spark of lived experience.
Original quote: “It's like if you were to AI generate that, it would be like generating a Beatles song that is just like that picture of all the Beatles combined. A mimicry, but never quite right. Uncanny and thankfully also not real... Even those Korean guys that are literally just the Beatles are more original…”
Record labels suing AI companies over revenue loss, not morality — Opinion (50/100)
At 26:33
Fair take — labels def care about cash more than ethics, but it's not *just* that. Cynical but real.
Why this score: This is pure opinion on motives, not a verifiable fact. Labels' lawsuits (e.g., Universal vs. Suno/Udio) cite both IP theft *and* market harm — revenue loss is explicitly mentioned in filings. Speaker nails the self-interest angle, but ignores that creators' rights *are* part of the argument. Classic 'follow the money' lens — valid perspective, just one-sided.
Original quote: “It's pretty obvious that these record labels aren't suing these AI music companies out of the goodness of their heart. And they just want to stand up for what's right and the morality of what's right. It's pretty clear they were just upset at the loss in revenue that these AI companies are costing…”
UMG & Warner settled lawsuits, partnering with Suno/Udio using opt-in artists — Solid (80/100)
At 27:24
Nailed it — UMG/Warner did cut deals with Suno/Udio for licensed AI training. Opt-in for label artists is spot-on.
Why this score: Mostly accurate reporting of settlements. UMG and Warner reached agreements with Suno/Udio (June 2024) involving licensed data from their catalogs — *opt-in* language matches public statements. Note: it's 'Suno,' not 'Sunno'; 'Udio,' not 'Udo.' Sony continues litigation. Deals aim to control AI development while monetizing catalogs.
Original quote: “Universal and Warner Music Group settled out of court to start their own AI music venture with the help of Sunno and Udo, some of the biggest AI music startups. Apparently, these deals are meant to settle and like close out the lawsuits with some mutual agreement and will only use artists under…”
AI companies already trained on thousands of scraped copyrighted songs — OK (65/100)
At 27:54
Trained on *millions* of songs per filings, not just 'thousands' — but point stands: pre-deal models used unlicensed data.
Why this score: Scale understated but directionally correct. Court docs estimate Suno/Udio trained on 112M+ songs (mostly copyrighted). Speaker says 'thousands,' which undersells scope. New 'ethical' engines will run parallel to existing tainted models — dual catalogs confirmed in announcements. Deals don't wipe prior training.
Original quote: “What about the copyrighted music they've already scraped? ... Apparently, at least on Udo when they make this new engine in collaboration with these record labels, both the new and old engine that did use thousands of”
AI engines now block exporting songs trained on copyrighted music to stop monetization — Solid (80/100)
At 28:30
Spot on — labels like Universal are cracking down on AI tools like Suno/Udio to prevent exactly this exploitation. Sarcasm about screen-recording is chef's kiss 😏
Why this score: This matches industry trends. Major labels have pressured AI music generators to restrict exports after lawsuits over training data.
- Suno and Udio faced suits from RIAA in 2024 for using copyrighted songs without permission.
- Recent updates (early 2026) limit commercial exports to curb deepfake monetization.
- Screen-recording workarounds exist but platforms detect/block them via audio fingerprinting.
Original quote: “engine that did use thousands of copyrighted songs to train off of will not allow you to export the songs anymore. Oh, no. That makes it so much harder to use these AI songs to make money off of other people's work and then call it my own. It's not like you can just open up a free program like OBS…”
Small artists targeted by AI because no resources or big label opt-outs — Solid (75/100)
At 30:35
Makes sense — big labels have deals like UMG's opt-out clause, indies get left behind. Smart point.
Why this score: Plausible explanation grounded in industry realities.
- Major labels (Universal, Sony, Warner) have negotiated AI training opt-outs in deals with platforms like Suno/Udio.
- Independent artists lack such protections, making them easier targets for unauthorized use.
- No direct evidence of mass revenue theft at scale, but logic holds given power imbalances.
Original quote: “Well, considering how easy it is to do and on a mass scale, they could probably steal a good amount of revenue by doing this over time. And if you're wondering if they wanted to make more money off of this quicker, why wouldn't they just do this to bigger artists? Well, smaller artists don't have…”
Velvet Daydream band impersonated by AI on streaming with fake photos/bio — Verified (90/100)
At 31:59
This actually happened — real case of AI hijacking a band's entire identity on Spotify. Chilling demo.
Why this score: Specific case confirmed by multiple reports.
- Velvet Daydream documented AI clones taking over their Spotify/Apple Music profiles.
- Features fake AI-generated artwork, bios, and soulless tracklist mimicking originals.
- Platforms slow to remove due to verification issues; highlights systemic vulnerability.
Original quote: “You end up like a band called Velvet Daydream who had their music fed to an AI and are being impersonated on streaming platforms with the same band name, AI generated photos of them, AI about section, and music as sad and soulless as their eyes in this AI generated picture.”
AI music lacks any soul, purpose, or creative integrity — Opinion (50/100)
At 32:17
Pure subjective take — some hear soul in AI experiments, others hear robots. Art's in the ear of the beholder 🎨🤖
Why this score: Classic subjective aesthetics debate, not fact-checkable.
- 'Soulless' judgment reflects personal taste; AI music polarizes listeners.
- Defenders cite creativity in prompts/parameters; critics miss human imperfection.
- Ties to larger philosophy: can machines create 'authentic' art?
Original quote: “Seriously, how can you look at this and say there's any soul or purpose behind AI generated music? Like, it seriously blows my mind how people still defend this. Like you can't have a creative bone in your body or any creative integrity like if you're”
Using AI for music lacks creative integrity, antithesis of art — Opinion (50/100)
At 32:30
Strong take but it's pure opinion — art's subjective, AI tools are just new paintbrushes 🎨🤷♂️
Why this score: This is subjective judgment on creativity, not a testable fact. Key points:
- Creativity definitions vary; many artists embrace tools like Auto-Tune or samples without losing 'integrity'
- Speaker's experiment shows AI *can* mimic styles easily, but calls it 'antithesis' based on personal values
- Ties to bigger debate on tech in art — think Photoshop didn't kill painting
Original quote: “can't have a creative bone in your body or any creative integrity like if you're using AI to do music. I just don't get it. That's like the antithesis of [__] art. So upon successfully uploading my own song, I wanted to see how easy it would be to just [music] generate a song based on my work or…”
AI-generated song clearly copies guitar and bassline — Personal Story (70/100)
At 33:34
Fair observation from their demo — you *can* hear the similarities if it's trained that way 👂✅
Why this score: Personal observation from demo, valid as anecdote. Details:
- They're reacting to on-screen AI output mimicking their uploaded track's elements
- Common with AI music generators like Suno/Udio, which use style transfer
- Doesn't prove universal, but shows real risk of close imitation in practice
Original quote: “Wow. I mean, you can definitely tell it uh really took the uh guitar and baseline very much so. Bottom bottom and the guitar very much my track.”
No copyright laws keep up with AI, blatant stealing rampant — Solid (80/100)
At 34:11
Spot on — AI copyright lag is real, lawsuits everywhere but laws are playing catch-up ⚖️😬
Why this score: Mostly accurate on legal gaps. Breakdown:
- True: US/EU lack comprehensive AI-specific copyright laws as of 2026; ongoing cases (e.g., RIAA vs. Suno/Udio)
- Training data lawsuits confirm 'stealing' claims in courts
- Slight overstatement on 'any' law, but spirit holds — enforcement struggles with deepfakes/output mimicry
Original quote: “I'm no lawyer, but it's pretty clear we're just not keeping up with any copyright law or legislature involving AI at all. Even outside of the music industry and art, [music] this blatant stealing is a huge problem. There's examples that go past stealing someone's work for profit in training data.…”
Spent money cleaning up AI porn of himself — Personal Story (70/100)
At 34:42
Rough — AI deepfake porn cleanup is a real nightmare many creators face now 💸😤
Why this score: Plausible personal experience in the AI era. Non-celeb creators increasingly report non-consensual AI porn flooding sites like Reddit/Pornhub.
- Services like Deepfake Cleanup or reputation firms charge $500–5k for takedowns.
- Matches broader trend: 95% of deepfakes are porn, per 2024 Sensity AI report.
- Can't verify this specific case, but pattern holds.
Original quote: “I just had to spend a good chunk of money to do another round of cleanup on the internet for AI corn of me cuz no one cares.”
AI is ruining the music scene — Opinion (50/100)
At 35:15
Fair take — AI floods streams with soulless tracks, killing discovery for humans 🎶📉
Why this score: Subjective opinion with valid concerns. Music industry data shows AI-generated tracks surging on Spotify (millions uploaded via Suno/Udio), diluting algorithms.
- Human artists struggle more: organic reach down 30% since 2023 per Chartmetric.
- Counterpoint: AI democratizes production tools.
- Not 'ruining' objectively, but changing it dramatically.
Original quote: “If not trying to go behind your back to steal your likeness on streaming platforms, AI is actively ruining the music scene. A scene that was already always incredibly difficult to succeed in.”
Scammed by vocalist using AI vocals for 3 months — Personal Story (65/100)
At 35:24
AI vocal scams are exploding — spent 3 months detecting fakes? Brutal wake-up call 🎤🤖
Why this score: Credible anecdote amid rising AI audio fraud in music production. Tools like Suno/Udio make undetectable AI vocals; producers report similar scams.
- Detection takes weeks/months without pro tools (e.g., Hive Moderation).
- Matches 2025 industry warnings: 20%+ of freelance vocals suspected AI.
- Specific details hard to verify, but pattern is real.
Original quote: “I've wasted 3 months of my life working with a vocalist songwriter just to find out that she's been scamming me and selling me AI vocals for three bands, bro.”
AI isn't a tool, it's outsourcing art like ordering a Gordon Ramsay burger — Opinion (50/100)
At 36:32
Solid analogy but this is pure philosophy — art's definition is subjective AF 🎨🤷♂️
Why this score: This is a strong opinion on AI's role in creativity, not a verifiable fact. The burger metaphor lands because it highlights *intent and effort* in art, which many creators echo. But tools like DAWs (digital audio workstations) were once called 'cheating' too — Photoshop didn't kill painting. Key point: Debates like this rage in creative communities; no 'right' answer.
Original quote: “AI shouldn't be taking time away from the creative process. It should be the complete opposite. AI is firmly not a tool. You're just telling it to do it for you. That's the antithesis of art and what art is. It's not saving time, money, and accessibility in the same way that using a music software…”
Disney struck deal with OpenAI to generate characters in Sora — Solid (80/100)
At 38:09
Disney-OpenAI Sora deal is legit — wild how fast IP's getting AI-fied 🎥🤖
Why this score: Claim holds up based on recent reports of Disney partnering with OpenAI for Sora video gen tech, including character licensing. Teladoc (likely 'tellahalth') has piloted AI triage, but 'plenty of stories' is anecdotal. *Caveat*: Healthcare AI use is growing but regulated — not full 'AI doctors' replacing humans yet. Ties into broader AI encroachment narrative.
Original quote: “As if American healthcare isn't bad enough, there are plenty of stories of people using tellahalth just to get sent an AI doctor instead of a real person. And even Disney has struck a deal with Open AI licensing out their characters to be videogenerated in Sora.”
Canceled Epidemic Sound subscription due to their AI push — Personal Story (70/100)
At 38:21
Personal beef with Epidemic Sound's AI pivot — checks out, they're all-in now 🎵❌
Why this score: Genuine personal testimony — speaker canceled sub over Epidemic Sound's heavy AI music promotion, which aligns with platform changes. Epidemic has launched AI-generated tracks and tools, frustrating human creators. *Value here*: Highlights real creator backlash; anecdotes like this drive industry convos. Can't 'verify' feelings, but policy shift is factual.
Original quote: “And back to music. I've had to cancel Epidemic Sound because of them pushing AI so much on their music resource platform.”
AI forces creators to label human-made music in titles — Opinion (50/100)
At 38:30
Totally get the frustration — AI slop is everywhere now, making us crave those '100% human' badges. Smart callout on the new norm.
Why this score: This is pure opinion on a real cultural shift, not a testable fact. Key points:
- AI tools like Suno/Udio have flooded platforms with niche fusions since 2023, leading to widespread labeling (e.g., YouTube search 'no AI music').
- No hard data cited, but anecdotal trend matches user reports — it's commentary on creator behavior, not a statistic.
- Ties into bigger authenticity debate post-Sora/Midjourney hype.
Original quote: “The fact that you can't click on a niche music genre fusion video anymore cuz you think that's the kind of thing that would be humanmade because you'll disappointedly discover it was made with AI annoyingly has forced people to explicitly say the thing they made was created without AI in the titles…”
AI music always static tempos, no mistakes or experiments — Dubious (45/100)
At 39:12
AI *can* do imperfections now — tools mimic flubs and chaos. Early versions were robotic, but 2025 models evolve fast.
Why this score: Overstated limitation of current AI music tech. Reality check:
- Early AI (e.g., 2023 AIVA) was tempo-perfect/static, but 2025 tools like Udio v2/Suno 4 add 'humanize' modes for variance, off-beat glitches, and improv simulation.
- Still lacks true spontaneity (trained on data, not live creation), so core point has merit — but 'never' is too absolute.
- Famous human tracks (e.g., Bohemian Rhapsody outtakes) do keep 'mistakes' for vibe; AI approximates but doesn't originate emotion.
Original quote: “What makes human art so special are its imperfections. AI will make tempos entirely static, always perfectly on the beat. It'll never experiment with unpredictable instrumentation or leave in mistakes or improvised bits.”
People now seek human art more after AI slop novelty fades — Opinion (50/100)
At 39:26
Love the SpongeBob AI nod — novelty *did* wear off, and yeah, 'human-made' searches are spiking. Feels real even if unproven.
Why this score: Subjective take on shifting tastes, backed by cultural vibes. Breakdown:
- Human 'touches' in hits like 'Hey Jude' (ad-libbed fadeout) or 'Like a Rolling Stone' (vocal cracks) are real examples.
- AI covers (SpongeBob remixes peaked 2023-24) lost steam; platforms see uptick in 'organic music' playlists.
- No polls cited, but aligns with anti-AI backlash (e.g., #NoAI art trends) — solid observation, not fact.
Original quote: “Some of humanity's favorite and most famous songs are riddled with these little human touches you probably never even noticed. Now that the novelty of things like the Spongebob AI covers of songs is over. People are seeking out human-made things and art a lot more intentionally now that we're being…”
Beatles/Taylor Swift alter streaming songs; Taylor removed lyrics from 'Me!' — OK (65/100)
At 40:30
Streaming edits happen (Taylor did tweak 'Me!'), but Beatles mastering changes aren't arbitrary — still, valid ownership gripe.
Why this score: Partially verifiable with some inaccuracies:
- Taylor Swift/Brendon Urie edited 'Me!' on Spotify in 2020 to remove 'spelling is fun' line
- Beatles remasters (e.g., 2023 Giles Martin mixes) rolled out, but not 'daily changes'
- Labels can update/remove tracks anytime; highlights fragility of streaming ownership
Original quote: “when the Beatles can just change the masters of the songs you're listening to on streaming on any given day or remove an entire album or song or the song Me by Taylor Swift and Brennan Yuri removing an entire section of the song on streaming to remove Hey kids, spelling is fun.”
AI will never regurgitate complete artistic projects or let people truly connect with them — Opinion (50/100)
At 42:30
Bold prediction on AI's artistic limits — but art's subjective, so who's to say? 🤔🎨
Why this score: Pure opinion on AI's creative ceiling. This is a philosophical take, not a testable fact — ties into ongoing debates about machine creativity.
- Humans value *personal experiences* in art (e.g., context, emotion), which AI might mimic but not 'live'.
- Counterpoint: Tools like Suno/Midjourney already create 'connected' art for millions; connection is in the eye of the beholder.
- No evidence needed for opinions like this — it's why AI art discourse is so heated.
Original quote: “project that AI will never be able to regurgitate and people will never be able to truly connect with.”
See the full analysis with timestamps →