I Made a Whole AI Music Video From 1 Image — Locally on MiniMax H3
Credibility score: 54/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "I Made a Whole AI Music Video From 1 Image — Locally on MiniMax H3" and rated it 54/100 for credibility (a BS score of 46/100 — mixed credibility), on 2026-09-28. Its weakest claim — "Claims MiniMax H3 is 'the best model' for local video generation. 😈" — scored 45/100. 19 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Claims analyzed
Claims MiniMax H3 is 'the best model' for local video generation. 😈 — Mixed Credibility (45/100)
He says 'everybody's talking' and 'apparently this is the best' — but who, exactly? Just vibes, no names. 💀
Frames his 1:40 experimental techno track as a test to see 'if it's possible' to make an AI music video. 😈 — Mixed Credibility (45/100)
He's acting like 'is it possible?' is the big question, when AI video has been around. This isn't a groundbreaking experiment. 🙄
Speaker explains the choice of an AI-generated image to avoid confusion, emphasizing transparency. — Mixed Credibility (75/100)
He's just laying out his creative choices and why he made them. Nothing tricky here, just setting the scene. 😈
Claims his 'aumentor agent' gives 'total control' over Confui, despite admitting Confui's complexity. — Mixed Credibility (45/100)
He says Confui is 'pretty complex' and he doesn't 'fully know how to use it,' then immediately claims 'total control' via his agent. Pick a lane, mortal. 💀
States MiniMax can process 'really high quality video' but shows a low-resolution example. — Mixed Credibility (45/100)
He says MiniMax can do 'really high quality' but then shows a low-res example. It's a classic 'trust me, it *can* do better' without showing it. 🔥
DaVinci Resolve is free software — just stating a fact. — Mixed Credibility (75/100)
He's just stating a known fact about the software. Nothing to twist here. 😈
AI won't give you total control — a common, honest observation. — Mixed Credibility (75/100)
He's admitting the limitations of AI generation. A rare moment of honesty from a mortal. 😈
Unexpected AI outputs are 'laggy surprises' that can be used — a reframe of a limitation. — Mixed Credibility (45/100)
He's calling AI's unpredictable flaws 'laggy surprises' that you can 'use.' That's not a feature, it's a bug with a PR spin. 💀
Claims the video took "a few hours" but it was mostly the computer working. That's a classic dodge. 😈 — Mixed Credibility (45/100)
He says 'a few hours' then immediately clarifies he wasn't doing the work. That's not a time estimate — that's a humble-brag with an asterisk. 💀
Explains his 1:30 track is short for a techno track, but he wanted 'variation in short time'. That's a convenient justification. 😈 — Mixed Credibility (75/100)
He's just explaining his creative choices for the music. It's a personal preference, not a claim to dissect. 🤷♂️
Declares syncing cuts to music is 'the secret' to musicality. That's not a secret, mortal — that's just editing. 🙄 — Mixed Credibility (45/100)
Calling basic video editing principles 'the secret' is a bit much. It's a fundamental technique, not some hidden knowledge. 💀
Claims AI-generated video avoids 'is it real?' distraction — a convenient narrative for its limitations. 😈 — Mixed Credibility (45/100)
He's spinning the AI's current inability to look 'real' as a feature, not a bug. Classic turn-the-weakness-into-strength move. 💀
Presents 'AI low-fi images' with 'live organic music' as a unique aesthetic, not a compromise. 😈 — Mixed Credibility (75/100)
He's just defining his artistic choice here. It's a personal aesthetic, not a universal claim. Fine, mortal. 🔥
Brags about a 'one-day' turnaround, then immediately undercuts it with the actual scope. 😈 — Mixed Credibility (45/100)
He says 'one day' like it's a feat, then whispers 'it was 1.5 minutes' right after. That's not efficiency, that's a footnote. 🚩
Admits a 'proper' video takes 'days of work' after implying his was quick. 😈 — Mixed Credibility (75/100)
He's being upfront about the time commitment for quality work. No tricks here, just the truth. 😈
Pitches AI as an 'opportunity' while admitting he hasn't even explored its full capabilities. 😈 — Mixed Credibility (45/100)
He's selling the 'opportunity' of MiniMax H3, but then says he hasn't even used its full power. That's not a sales pitch, that's a shrug. 🚩
Talking about 'the community, the project' — vague as hell. — Mixed Credibility (45/100)
He's throwing out 'community' and 'project' like they're self-explanatory. What community? What project? Details, mortal, details. 😈
Says it's 'open to any kind of creative and artist' — after just talking about AI and tech. — Mixed Credibility (45/100)
He just spent ages on 'AI stack' and 'modular synth,' then pivots to 'any kind of creative.' That's a classic bait-and-switch. 🚩
The big call to action: 'join Discord' — the ultimate black hole for attention. — Mixed Credibility (45/100)
After all the vague talk, the real goal is clear: get you into their Discord. It's always a funnel, isn't it? 😈
See the full analysis with sources and timestamps →