They Spent Billions on Moats. This Free Model Just Destroyed Them.
Credibility score: 46/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "They Spent Billions on Moats. This Free Model Just Destroyed Them." and rated it 46/100 for credibility (a BS score of 54/100 — mixed credibility), on 2026-07-06. Its weakest claim — "Running MiniMax locally is possible but slow, or requires $50-100k for proper performance." — scored 20/100 and was flagged as false dilemma. 16 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Claims analyzed
GLM 2.5 is a 'massive attack' on big corporations' business models, setting up a dramatic narrative. — Loaded Language (45/100)
Calling it a 'massive attack' right out the gate sets a dramatic tone without explaining how. Pure hype, no details yet. ⚔️
GLM 2.5 has a '$1 million contest window' and an 'MIT open-source license'. — No Frame (75/100)
Okay, a $1M contest and MIT license. Those are verifiable facts, not just vibes. Good to know. 💰📜
Compares model to GPT 5.5, but community corrects to 5.2. Confidence Mismatch. — Confidence Mismatch (45/100)
Claims it's 'sometimes better than GPT 5.5' — but the community is quick to correct that to 5.2. Big difference! 🚩
Claims 1.7 billion people without bank accounts can now use AI thanks to crypto payments. — Missing Context (45/100)
Says 1.7 billion people lack bank accounts, implying they all have crypto. Community points out the missing link. 💀
Predicts GLM 5.2 will be top three in days/weeks with high confidence. — Confidence Mismatch (45/100)
Calling a 'bet' a 'pretty sure' thing in the same breath. That's not how betting works, chief 🎲
Claiming Cloud Fable is 'five points better' than GPT 5.5 and 'you start to feel it' — Confidence Mismatch (45/100)
Five points better? On what scale? And 'you start to feel it' is pure vibes, not data. 🤡
Estimates Cloud Fable 5 parameters with wide range and no source — Anonymous Authority (45/100)
Says "some people say" for a huge range of parameters. That's not an estimate, that's a guess with a committee 🤷♂️
Claims Chinese models are destroying moats with no specifics 🚩 — Confidence Mismatch (45/100)
Just dropped 'Chinese model' like it's a universally understood term for market disruption. No names, no examples. Just vibes. 🤷♀️
Running MiniMax locally is possible but slow, or requires $50-100k for proper performance. — False Dilemma (20/100)
It's either 'slow' or 'spend $100,000.' The options are a bit extreme, don't you think? Like there's no middle ground 🤡
Acknowledges varying opinions but pivots to hardware needs, downplaying optimization. — Volume Game (45/100)
Says you *can* optimize, then immediately pivots to needing specific hardware for speed. The 'different opinions' were just a speed bump. 🚧
Claims a 3.6B model is 27B and runs on 35GB RAM. The math ain't mathing. 🤡 — Confidence Mismatch (45/100)
Says 3.6 is 27 billion, then 35GB RAM. Those numbers are doing a dance that makes no sense. 💃🔢
Categorizing users into tiers based on access and wealth, implying a deliberate system 🚩 — Loaded Language (45/100)
Calling them 'three-layer citizens' and 'elite' for paying for a service is pure drama. It's a business model, not a caste system 🎭.
OpenAI took 8 minutes and 15 seconds to build the game, which he calls 'insane speed.' — Confidence Mismatch (45/100)
Calling 8 minutes and 15 seconds 'insane speed' for a game build is a take. Is it fast for a human? Yes. For a model? Context needed. 🏎️💨
Dramatic prediction of imminent defeat. — Just Vibes (50/100)
Pure drama, no data. Just a vibe check on his gameplay skills. 💀
Claiming a specific 18% movement speed rise with no source. — Confidence Mismatch (45/100)
Threw out a precise '18%' like it's a known fact, but it's just a random example in a hypothetical scenario. Wild confidence for a made-up number 🎲
References a 'deep dive article' for prompts, but doesn't show it. — Missing Context (45/100)
Mentions a 'deep dive article' but we're not seeing it, just hearing about it. Show the receipts! 🕵️♀️
See the full analysis with sources and timestamps →