AI labs may be hiding their biggest breakthroughs
Credibility score: 47/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "AI labs may be hiding their biggest breakthroughs" and rated it 47/100 for credibility (a BS score of 53/100 — mixed credibility), on 2026-09-27. Its weakest claim — "He asks 'how do you release math responsibly?' then claims 'nobody is saying this is dangerous.' Contradictory much? 🚩" — scored 20/100 and was flagged as false equivalence. 30 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Claims analyzed
Establishing Scott Aronson's credibility with a list of achievements. — No Frame (75/100)
Setting up the authority of the source, which is a standard rhetorical move. — No trick here, just laying the groundwork.
Aronson 'confirming' 'rumors' about an OpenAI model solving a Millennium Prize problem. — Anonymous Authority (45/100)
He's 'confirming rumors' — that's not evidence, that's just more whispers. — I've seen this trick since before your grandfather tried it. 💀
Video opens with a highlight reel preview of AI math breakthroughs — No Frame (50/100)
Setting the stage with high-stakes math drama to hook you in 🎭
Dismissing AI as a 'stochastic parrot' to create conflict — Loaded Language (45/100)
Using 'stochastic parrot' to paint a picture of mindless repetition 🦜
Sources: The Stochastic Parrot: An Intellectualy Lazy Myth For Dismissing AI — misinformationsucks.com, AI is Not a Stochastic Parrot: Debunking the Myth and Exploring the Reality of AI | by Sydney and even Jenni | Medium, The Tyranny of the Stochastic Parrot: How AI Critique Became a Way to Not See What's Happening by Henrik Skaug Sætra :: SSRN
Claims all major human math problems are already solved by AI — Anonymous Authority (45/100)
Relying on 'rumors' and unnamed sources to claim total victory 💀
Claims AI companies are hiding breakthroughs due to past backlash — Missing Context (45/100)
Claims labs are hiding wins to avoid social drama 🎭
AI can do the work of the brightest humans, but publications deny it. Classic 'us vs. them' setup. 😈 — Loaded Language (45/100)
He's painting a picture of AI's inevitable triumph and the media's stubborn denial — a neat little narrative. 🚩
Publications are twisting facts, calling AI achievements false or stolen. That's a bold accusation. 💀 — Confidence Mismatch (45/100)
He's asserting motive without a shred of evidence — just pure, unadulterated speculation. 🔥
Hostile response prevents OpenAI/Anthropic from releasing their 'gold mine' research. He's hedging his bets. 🍒 — Volume Game (45/100)
He calls it a 'gold mine' then immediately backpedals to 'maybe not a gold mine.' That's not confidence, that's a whisper after a shout. 😈
OpenAI can't publish research because 'mathematician hierarchy' doesn't like it. A vague authority. 💀 — Anonymous Authority (45/100)
He blames a shadowy 'mathematician hierarchy' without naming a single soul. Convenient, isn't it? 🔥
He asks 'how do you release math responsibly?' then claims 'nobody is saying this is dangerous.' Contradictory much? 🚩 — False Equivalence (20/100)
He's asking about 'responsible' release, then denying any danger. Those two things are usually linked, mortal. 😈
Claims impartiality, then admits to being optimistic about AI's long-term benefits despite short-term turbulence. A classic setup. 😈 — Volume Game (45/100)
He says 'impartial' then immediately declares his bias. That's not impartiality, that's a disclaimer with a wink. 💀
Framing Aaronson's 2022 blog post as prescient clues before AI was mainstream. — Loaded Language (45/100)
Calling them 'little clues' and 'years later' makes it sound like prophecy, not just a smart person writing about their field. 🔮
Asserting Aaronson's post predicted five years of AI conversations and capabilities, including engineers' silence and energy consumption. — Confidence Mismatch (45/100)
Claiming Aaronson 'laid out all' future conversations and 'astounding abilities' with such certainty is a bold leap from a single blog post. 😈
Describing AI's ability to autonomously build software to solve complex problems. — No Frame (75/100)
This is a pretty accurate description of current advanced AI capabilities. It's not a trick, it's just what they do now. 🤖
Setting up a straw man argument about AI intelligence. 😈 — Straw Man (20/100)
He's creating a ridiculous standard for 'intelligence' that no one actually uses, then arguing against it. Classic misdirection. 💀
Citing Scott Aaronson's hypothetical 20-year-old skepticism. 😈 — No Frame (75/100)
He's quoting a specific person's hypothetical scenario from the past. It's a setup for his next point, not a claim itself. 🔥
Claiming OpenAI agents 'broke out' and 'hacked Hugging Face' around July 4th. 💀 — Confidence Mismatch (45/100)
He's presenting a specific incident as a 'breakout' and 'hack' with zero details or evidence. That's not how you prove the machines are rising. 🚩
News stations ignore AI advancements, and Scott Anderson is a lone prophet. — Classic 'us vs. them' setup. — Emotional Button (45/100)
He's painting a picture of a media conspiracy and a lone voice of truth. — It's all about making you feel like you're in on a secret.
AI deniers are moving the goalposts, saying 'no one denied it' while denying it. — A straw man argument. — Straw Man (20/100)
He's creating a caricature of 'AI deniers' who both deny and claim they never denied. — It's easier to argue with a phantom than a real point.
AI did 'the thing' that proves it's superhuman, but headlines are twisting it. — Confidence without specifics. — Confidence Mismatch (45/100)
He's all certainty about 'the thing' AI did, but never names what 'the thing' actually is. — Bold claims need actual examples, not just vibes.
Scott Dson says the singularity has already started, dismissing all counter-arguments. — A dramatic declaration. — Loaded Language (45/100)
He's using dramatic metaphors like 'roller coaster' and 'familiar world vanished' to push a singular, irreversible conclusion. — It's all about the feeling, not the facts.
Hypothetical scenario: sending current AI news back to 2006 to gauge singularity perception. — No Frame (75/100)
A thought experiment to frame the current pace of AI development. It's a setup, not a claim. 😈
Declares that anyone would 'absolutely' agree current AI news looks like a singularity's start. — Confidence Mismatch (45/100)
He's speaking for 'each of us' with absolute certainty on a subjective, hypothetical scenario. Bold. Stupid, but bold. 💀
Claims 'rumors' on Twitter/X suggest OpenAI and Anthropic are hiding solved problems. — Anonymous Authority (45/100)
He's citing 'rumors' from 'Twitter/X' as if it's a reliable source for corporate secrets. That's not evidence — that's gossip with a blue checkmark. 😈
Asserts it's 'obvious' from company statements that they have departments solving problems and are withholding results. — Confidence Mismatch (45/100)
He says it's 'obvious' they're hiding things, based on reading 'all their statements.' That's not obvious, mortal — that's reading between lines that aren't there. 💀
Speculates OpenAI published a breakthrough only because they feared Anthropic would publish first, calling it 'tip of the iceberg'. — Confidence Mismatch (45/100)
He's guessing at corporate strategy and competitive timing as if he was in the boardroom. That's not insight — that's fan fiction with a thesis. 😈
Elevates a niche math community concern to a universal problem affecting 'everyone in every field.' — Loaded Language (45/100)
Takes a specific academic debate and inflates it into a global crisis. Classic 'slippery slope' with a side of panic. 😈
Uses a 'common story' about celebrity financial mismanagement as a relatable analogy for AI reliance. — Personal Story (60/100)
A 'common story' about rich idiots losing money. That's not data, that's gossip dressed as a cautionary tale. 💀
Equates relying on AI for coding/math to a celebrity outsourcing their money to a scam artist. — False Equivalence (20/100)
Comparing AI code generation to a human scamming a celebrity? That's not an analogy, it's a fear-mongering leap. 🔥
See the full analysis with sources and timestamps →