Did AI Solve One of Math’s Hardest Problems?
Credibility score: 70/100 — Mostly Credible. Mixed credibility - some claims are solid, others need verification.
BSmeter analyzed "Did AI Solve One of Math’s Hardest Problems?" and rated it 70/100 for credibility (a BS score of 30/100 — mostly credible), on 2026-10-05. Its weakest claim — "OpenAI announced they 'solved' a Millennium Problem with AI — a bold claim, indeed." — scored 45/100 and was flagged as confidence mismatch. 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
AI has solved the Navier-Stokes equation — a bold, attention-grabbing opener. — No Frame (75/100)
Starting with a bang, asking a question that sets up the whole video. It's a hook, not a lie. 🔥
Defining the Navier-Stokes equation and its connection to Newton's laws — straightforward explanation. — No Frame (75/100)
A basic definition of the Navier-Stokes equation and its relation to Newton's laws. Nothing tricky here, just setting the groundwork. 😈
Introducing an expert on Navier-Stokes equations and fluid dynamics. — No Frame (75/100)
They're just bringing on their guest, setting the stage for the topic. Nothing to dissect here. 😈
Defining 'fluid' as anything continuous that flows and takes its container's shape. — No Frame (75/100)
A basic definition of a fluid. Simple, direct, and accurate. No tricks, just physics. 😈
Clarifying why salt isn't a fluid, but water and gas are. — No Frame (75/100)
Distinguishing between granular solids and true fluids. It's a clear, correct distinction. 😈
Navier-Stokes equations apply to all fluids, just with parameter changes. — No Frame (75/100)
Explaining the universality of Navier-Stokes for different fluids. It's a fundamental concept. 😈
Navier-Stokes models turbulent fluids in astrophysics and on Earth. — No Frame (75/100)
Connecting Navier-Stokes to real-world phenomena, from stars to weather. It's a solid explanation. 😈
Dodging the 'good or bad' question about turbulence — classic non-answer. 😈 — No Frame (75/100)
He won't 'value judge' turbulence, which is fair. It's a phenomenon, not a moral choice. 🔥
He's going to 'value judge' turbulence — as if physics cares about his feelings. 😈 — Just Vibes (50/100)
Trying to assign moral worth to a physical phenomenon. That's not how the universe works, mortal. 💀
Using 'little bits of turbulence' to prevent 'big swirls' for airplane efficiency. 😈 — No Frame (75/100)
Actually, this is a known principle in aerodynamics. Using controlled turbulence to reduce drag is clever. 🔥
The 'combat turbulence with turbulence' line — sounds like a paradox, but it's just physics. 💀 — No Frame (75/100)
It's a neat trick of fluid dynamics, using small, controlled turbulence to prevent larger, more disruptive patterns. No deception here. 😈
Golf ball dimples create 'mini turbulence' for longer flight. 😈 — No Frame (75/100)
Another solid example of controlled turbulence. Dimples reduce drag by creating a turbulent boundary layer. 🔥
Navier-Stokes equations predict infinite fluid speed, which is 'bad for physics.' 😈 — No Frame (75/100)
The 'smoothness and existence' problem for Navier-Stokes is a real thing. Infinity in a physical system is indeed a problem. 🔥
The 'two dudes' from the 19th century — a casual, but accurate, historical nod. 🔥 — No Frame (75/100)
They're talking about Claude-Louis Navier and George Gabriel Stokes. The names are literally in the equation's title. No trickery. 😈
The core problem: Navier-Stokes equations predicting infinite fluid speed — a genuine mathematical mystery. 💀 — No Frame (75/100)
This is the actual, unsolved problem they're discussing: proving the global existence and smoothness of solutions. It's a Millennium Prize Problem. No exaggeration. 🔥
Explaining how scientific models evolve, building on previous understanding. — No Frame (75/100)
He's just laying out how science actually works — new theories refine, not erase, the old. No trickery here. 🔥
Clarifying the core problem: proving equation stability and finite solutions. — No Frame (75/100)
He's cutting through the noise to the actual mathematical challenge. Direct and to the point. 😈
Defining the core mathematical problem: stability and smooth solutions. — No Frame (75/100)
Clearly laying out the mathematical challenge. No tricks here, just the problem itself. 🔥
Defining 'not smooth' as hitting infinity. — No Frame (75/100)
A simple, brutal definition. When the numbers go wild, the solution isn't smooth. That's just how it is. 💀
Defining a 'non-smooth' solution as hitting infinity. — No Frame (75/100)
A concise, accurate definition of what a 'non-smooth' solution means in this context. No hidden agenda. 😈
The Navier-Stokes equations couldn't be proven not to 'blow up' mathematically. — No Frame (75/100)
This is a straightforward description of the core problem with the Navier-Stokes equations. No trickery here. 🔥
No one could prove the system wouldn't 'blow up' — setting the stage for a grand solution. — No Frame (75/100)
He's just stating the core problem, the mathematical challenge they're discussing. No trickery here. 🔥
OpenAI announced they 'solved' a Millennium Problem with AI — a bold claim, indeed. — Confidence Mismatch (45/100)
OpenAI made a splash, but 'solved' is a strong word for finding a counterexample. It's a specific kind of solution. 😈
OpenAI announced solving a Millennium Problem using AI. — Confidence Mismatch (45/100)
He says 'OpenAI solved a Millennium Problem' like it's a done deal. That's a bold claim, mortal. 💀
Millennium Problems are unsolved math problems with a million-dollar prize each. — No Frame (75/100)
This is a correct description of the Millennium Prize Problems. No lies detected. 😈
OpenAI found a counterexample to the Navier-Stokes smoothness hypothesis, proving it false. — Confidence Mismatch (45/100)
He states they 'found a counter example' and 'you're done' like it's settled. The mathematical community is still debating this, mortal. 🚩
Finding a counterexample to the smoothness hypothesis means the hypothesis is false — a precise definition of 'solving' this part. — No Frame (75/100)
He's clarifying the exact nature of the 'solution' — a counterexample. That's how you break a hypothesis. 😈
The AI solution cost $6 million and 88 hours of GPU time, exceeding the prize money. 😈 — No Frame (75/100)
They're laying out the cost of this AI 'solution' — a hefty price tag for a math problem. 💰
Anthropic claims they used OpenAI tools to translate AI 'slop' into a human-readable solution, but OpenAI denies training on their inputs. 💀 — Missing Context (45/100)
Anthropic says they used OpenAI tools, then OpenAI says 'we don't think so' about training. That's a very polite way to say 'we didn't.' 🔥
OpenAI's 'we don't think so' is a classic dodge, not a denial. 💀 — Loaded Language (45/100)
OpenAI's 'we don't think so' is the corporate equivalent of a shrug. It's not a 'no.' 🔥
See the full analysis with sources and timestamps →