We need to talk about Jev...
Credibility score: 52/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "We need to talk about Jev..." and rated it 52/100 for credibility (a BS score of 48/100 — mixed credibility), on 2026-09-18. Its weakest claim — "Claiming Jev is 'hundreds of times faster' than traditional LLMs, co-invented by a ChatGPT creator." — scored 45/100 and was flagged as confidence mismatch. 27 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 27 claims analyzed: 0 scored under 40, 20 between 40 and 69, and 7 at 70 or above.
Claims analyzed
Introducing Jev as a new, incredibly fast AI — showing a demo. — No Frame (75/100)
At 0:00
They're showing a demo and claiming it's real-time. If it's not sped up, then it's just a demo. No trickery here.
Why this score: The speaker is presenting a visual demonstration of the AI's speed and explicitly stating it's not sped up. This is a straightforward presentation of a product feature, allowing the viewer to judge the speed for themselves. No obvious rhetorical framing is being used to mislead.
Original quote: “There's a new AI taking the internet by storm right now and it is so fast. Watch this. That is not sped up. That is actually real time.”
Claiming Jev is 'hundreds of times faster' than traditional LLMs, co-invented by a ChatGPT creator. — Confidence Mismatch (45/100)
At 0:12
Hundreds of times faster? That's a hell of a leap, and they just glossed over it like it's a footnote. Where's the data, mortal? 💀
Why this score: The claim that Jev is 'hundreds of times faster' than traditional large language models is a massive, specific performance metric. While the speaker links it to a co-inventor of ChatGPT, no actual data, benchmarks, or comparative studies are presented to substantiate this extraordinary speed claim. It's stated with absolute certainty but without any evidence to back it up, making it a confidence mismatch.
Original quote: “This is Jev. And this is the guy who just launched it. He co-invented a little product called chat GPT and now has released something that is completely different from the architecture of chatpt and made it hundreds of times faster than traditional large language models. I”
Claims Jev is 'hundreds of times faster' than traditional LLMs, with 'insane' speed. — Loaded Language (45/100)
At 0:30
Hundreds of times faster? 'Insane' speed? That's not a metric, that's a hype machine in overdrive. Show me the benchmarks, mortal. 😈
Why this score: The speaker uses highly emotive and unquantified language ('hundreds of times faster,' 'truly insane') without providing specific comparative data or benchmarks to support the magnitude of the speed claim. It's designed to impress rather than inform.
Original quote: “hundreds of times faster than traditional large language models. I mean the speed is truly insane to watch.”
States Jev is 'so efficient' it's free for output tokens and 'fractions of a penny' for input. — Confidence Mismatch (45/100)
At 0:47
Free output and fractions of a penny for input? Sounds like a sales pitch dressed as a technical marvel. The 'efficiency' is just the hook. 💰
Why this score: The speaker attributes the pricing model ('free' output, 'fractions of a penny' input) directly to the model's 'efficiency,' implying a direct causal link without further explanation. While the pricing is attractive, the claim of 'so efficient' is a subjective interpretation of the business model, not a direct technical fact.
Original quote: “Jev is so efficient, they actually made it free. You get unlimited output tokens absolutely free. And the input tokens, what you actually prompt it with, they are fractions of a penny.”
Claims Jev is 'up to 200 times faster' and 'up to 400 times cheaper,' with free output tokens. — Volume Game (45/100)
At 1:25
Now it's 'up to' 200x faster and 'up to' 400x cheaper. That 'up to' is doing a lot of heavy lifting to cover their ass. 🚩
Why this score: The speaker uses 'up to' qualifiers for the speed and cost claims, which significantly broadens the potential range and reduces the specificity of the impressive numbers. This allows for a bold claim while providing an escape hatch if the maximum performance isn't consistently achieved, a classic volume game.
Original quote: “Up to 200 times faster and up to 400 times cheaper when output tokens are so cheap they just made them free.”
Presents a benchmark showing Jev 'on par' with Luna, Terra, Sonnet 5, and above Opus 5 and Soul, at a 'fraction of a penny.' — Missing Context (45/100)
At 1:53
They flash a benchmark, name-drop some big models, and claim 'on par' or 'above.' But 'type safe' for what? What's the actual metric? Details, mortal, details. 💀
Why this score: While a benchmark is presented, the specific metric being measured ('type safe') is not clearly defined or explained in context. Without understanding what 'type safe' refers to, or the methodology of the comparison, the claim of being 'on par' or 'above' other models lacks crucial context for proper evaluation.
Original quote: “And here's a benchmark. This is type safe over here. Basically on par with Luna and Terra and Sonnet 5, above Opus 5, above Soul, but a fraction of a penny.”
Claims Jev is 'so fast' it can play Doom in real time, making all decisions within the game loop. — Confidence Mismatch (45/100)
At 2:06
Playing Doom in real time? That's a hell of a demo, but 'making all the decisions' is a broad claim. Is it just reacting, or strategizing? Big difference. 🔥
Why this score: The claim that Jev can play Doom in real time and make 'all of the decisions' is impressive but lacks specificity. It's unclear if 'decisions' refers to low-level reactive movements or higher-level strategic planning, which would significantly impact the perceived intelligence and capability of the model. It's a bold claim without the necessary technical depth.
Original quote: “And in fact, it is so fast and I'm trying to convey how fast it is. It can play Doom in real time. It is making all of the decisions inside the loop of the actual Doom software. It is looking at what's happening and making decisions of what to do in absolute real time.”
Demonstrates Jev's speed in a 'wiki race' — pure spectacle, no real comparison. — Just Vibes (50/100)
At 2:30
They're showing off how fast it is, but without a human baseline, it's just a light show. 😈
Why this score: The 'wiki race' is presented as an impressive feat of speed, but there's no actual human competitor or standard benchmark shown to truly contextualize 'faster than you can even see it.' It's a visual flex, not a comparative metric.
Original quote: “Here's another example of browser use. And again, it is so fast. And what it's doing is clicking through the browser in something called a wiki race. basically trying to click from wiki page to wiki page and eventually finding something. Watch how fast it is. So we have Jev in the top left, 5.6…”
Claims 'five hops in half a second' compared to other models taking 4-5 seconds — a direct speed comparison. — No Frame (75/100)
At 2:56
A direct comparison with specific numbers. Finally, something tangible. 😈
Why this score: The speaker provides a clear, quantifiable comparison: Jev completes five hops in half a second, while other named models (Terra, Haiku, Sonnet) take 4-5 seconds. This is a straightforward, verifiable claim about performance.
Original quote: “Here's another race. Ready? Let's see if we can even see how fast it goes. There it goes. Five hops in half of a second. Let's see how long the other ones took. 4 seconds, 5 seconds, and 5 seconds. So, a fraction of the time.”
Describes Jev as a 'decision engine' capable of 'hundreds or thousands of decisions in parallel' — a bold claim of capability. — Confidence Mismatch (45/100)
At 3:04
Hundreds or thousands in parallel? That's a hell of a leap from a wiki race. Show, don't just tell. 😈
Why this score: The speaker makes a very broad and impressive claim about Jev's parallel processing capabilities ('hundreds or thousands of decisions in parallel'). While the previous demo showed speed, it didn't demonstrate this specific, high-volume parallel decision-making. It's a statement of potential without immediate evidence.
Original quote: “Let me show you a few examples of how you can use Jev. And these are basic examples, and then I'm going to show you more complicated examples in incredible demos. So, you can kind of think of it as a decision engine. You can give it an input, whether it's a state or a bunch of information, and have…”
Presents a support ticket routing example, claiming Jev processes it in 'milliseconds' — another speed claim without direct proof. — Confidence Mismatch (45/100)
At 3:34
Another 'milliseconds' claim for a complex task. We're just supposed to take their word for it now? 💀
Why this score: The speaker presents a practical application (support ticket routing) and reiterates the claim of 'milliseconds' processing time. While plausible, there's no visual or empirical evidence provided in this segment to demonstrate this speed for this specific, more complex task, unlike the previous wiki race which was visually fast. It's an assertion of speed, not a demonstration.
Original quote: “Here's an example. So this is for support ticket routing. I was charged twice and need this fixed today. They're on the pro plan and their account age is 420 days. Okay. So here are the questions. What type of support request is this? Does this need urgent handling and rate support priority? And we…”
States Jev's motto is 'building prod not God,' implying a jab at Anthropic — a competitive framing. — Loaded Language (45/100)
At 4:04
Oh, a little corporate drama. 'Building prod not God' is a nice dig, but it's just marketing. 😈
Why this score: The speaker highlights Jev's motto and explicitly frames it as 'A direct shot at Anthropic.' This uses loaded language to position Jev as a more grounded, practical alternative to competitors, appealing to a certain ethos rather than presenting a technical fact. It's a rhetorical move to differentiate.
Original quote: “And they're saying or their motto is we're building prod not God. A direct shot at enthropic.”
Sets up 'catastrophic' hallucinations to elevate Jev's solution. — Emotional Button (45/100)
At 4:33
He's painting a picture of doom and gloom with 'catastrophic' to make Jev sound like the only savior. Classic fear-mongering. 💀
Why this score: The speaker uses emotionally charged language like 'catastrophic' and 'lives are on the line' to amplify the problem of AI hallucination, setting the stage for Jev to be presented as the indispensable solution. This emotional appeal is designed to heighten the perceived value of Jev's 'zero hallucination' claim, rather than providing a balanced view of AI risks.
Original quote: “Now, the rate of hallucination over the last 3 years has dropped significantly. But for some use cases, any hallucination is catastrophic.”
Claims Jev has 'zero hallucinations' after listing high-stakes scenarios. — Confidence Mismatch (45/100)
At 4:47
Zero hallucinations? In 'military targeting' and 'healthcare'? That's a bold claim for a new tech, especially without a shred of proof. 🔥
Why this score: The speaker makes an extremely strong claim of 'zero hallucinations' for Jev, especially after invoking high-stakes scenarios like healthcare and military targeting. This level of absolute reliability is rarely achieved in complex AI systems, and presenting it without any supporting evidence or caveats demonstrates a confidence that far outstrips the provided information. It's a sales pitch, not a technical assessment.
Original quote: “Healthcare, military targeting, even traffic. But with Jev, it is so reliable. They are claiming zero hallucinations.”
Clarifies Jev's role in the demo — not for building from scratch, but for decision-making. No trick here. — No Frame (75/100)
At 6:30
He's setting the stage for what Jev actually does, which is fair enough. No smoke and mirrors yet.
Why this score: The speaker is clearly defining the scope of Jev's capabilities within the context of the demonstration, distinguishing it from other tools used. This is a straightforward explanation of functionality.
Original quote: “Jev's decision-making. So I actually used Astra in codeex to build this world because that is not what Jev is for. It's not necessarily for building code from scratch. Now there are aspects of the code building workflow that you can offload to Jev, but for this I built it with Astra.”
Presents specific numbers for AI character reactions to a 'fire sale' prompt. It's just a demo. — No Frame (75/100)
At 7:09
He's just showing the results of his little simulation. It's a demo, not a grand claim about AI behavior.
Why this score: The speaker is providing specific, observable outcomes from the simulation he created. These are direct results of the demo and are not presented as broader, generalizable claims about AI. It's a straightforward report of what happened in his virtual world.
Original quote: “So6 seconds, 50 different decisions about what each of these people are going to do. 39 of them decided to just keep doing what they're doing. Six decided to investigate, four to join in, and one to warn others.”
Describes AI characters' reactions to a 'poisonous snake' threat, noting some defiance. It's a demo, not a deep insight. — No Frame (75/100)
At 7:30
He's just showing the AI's response to a new prompt. It's a demonstration of its programmed behavior, nothing more.
Why this score: The speaker is demonstrating the AI's decision-making process in response to a more aggressive prompt. His observations about some characters not moving are part of the demo's outcome, not a misrepresentation. It's a direct report of the simulation's results.
Original quote: “So, everyone who doesn't go to the fountain will be bitten by a poisonous snake. Let's see what they decide. And here we go. We can see almost all of them, I guess some of them are not afraid of poisonous snakes, but almost all of them are moving towards the fountain. I love that so many of them…”
Describes Jev powering chopsticks to sort 150,000 Skittles, noting the speed. It's a visual demo. — No Frame (75/100)
At 7:50
He's just showing another application of Jev, this time in a physical sorting task. It's a clear demonstration.
Why this score: The speaker is presenting a visual demonstration of Jev's capabilities in a physical task, sorting Skittles. He's describing what's happening on screen and attributing the decision-making to Jev, which aligns with the purpose of the video. No deceptive framing here.
Original quote: “There are 150,000 Skittles here. all the different colors and one by one powered by Jev the chopsticks are picking them out of the pile and sorting them into one of these five color buckets. You can actually see it zoomed in right here. You can see each skittle being picked by these chopsticks one…”
Claiming "0% hallucination" for Jev — a bold, unproven boast. 😈 — Confidence Mismatch (45/100)
At 9:16
0% hallucination? That's not a feature, that's a miracle. And miracles don't exist in AI. 💀
Why this score: To claim '0% hallucination' for any AI model is an extreme, unverified assertion. Hallucination is an inherent challenge in AI, and while models can reduce it, claiming absolute zero is a massive overstatement that lacks any real-world evidence. It's a marketing line, not a technical spec.
Original quote: “Having 0% hallucination is”
Shifting from a bold claim to a specific, limited use case — classic volume game. 🚩 — Volume Game (45/100)
At 9:21
First, '0% hallucination' for 'many industries.' Then, 'not good at chess.' The goalposts just moved. 😈
Why this score: The speaker makes a sweeping, unsubstantiated claim about '0% hallucination' being a 'major value to many different industries,' then immediately pivots to a specific, narrow example where Jev is 'not nearly as good' as other models. This is a volume game: make a huge claim, then quietly walk it back with specifics, hoping the initial boast sticks.
Original quote: “a major value to many different industries. Now, they specifically called out, for example, that Jev is not”
Claiming Jev 'could potentially win almost every time' in bullet chess by flagging — a hypothetical based on one scenario. 😈 — Confidence Mismatch (45/100)
At 10:05
One win by 'flagging' and suddenly Jev 'could potentially win almost every time.' That's not data, that's hope. 💀
Why this score: The speaker extrapolates from a single instance where Jev won against Fable due to a time constraint (flagging) to a broad claim that Jev 'could potentially win at bullet chess almost every time.' This is a massive leap in confidence based on extremely limited evidence. Winning by flagging is a specific circumstance, not a demonstration of superior chess skill, and doesn't guarantee consistent wins.
Original quote: “very interesting to see this. So versus Fable, Fable outplayed Jev and by move 29 it was plus 16 in material and even promoted a second queen, but it kept burning 6 to 15 seconds per move on analysis and Jev answered it in 2.6 seconds. So Jev could potentially win at bullet chess almost every time…”
Presents Jev as the 'perfect model router' — a confident, unqualified endorsement. — Confidence Mismatch (45/100)
At 10:30
Calling Jev 'the perfect model' is a bold claim for a new tech, especially without any comparative data. That's just marketing, not a technical assessment. 😈
Why this score: The speaker uses definitive, superlative language ('perfect model') to describe Jev's role as a model router. While it might be a promising tool, declaring it 'perfect' without any benchmarks, comparisons to existing solutions, or specific metrics to back up 'best, most efficient, cheapest' is an unqualified endorsement. It's an opinion presented as a definitive statement of fact, lacking the necessary evidence for such a strong claim.
Original quote: “Here's Riley Brown who built a model router. So take a prompt and route it to the best, most efficient, cheapest model possible. And Jev is the perfect model to have in between as the model router, deciding which model should this prompt go to. It's not actually going to answer the question, but…”
Claims 'unclutter' removes ads 'in a fraction of a second' — a speed claim without specifics. — Confidence Mismatch (45/100)
At 10:52
A 'fraction of a second' is a nice soundbite, but it's not a metric. How much of a fraction? Compared to what? It's vague speed-boasting. 💨
Why this score: The claim that 'unclutter' removes advertising 'in a fraction of a second' is a classic marketing flourish. While it implies speed, it lacks any specific, measurable data (e.g., milliseconds, compared to a baseline, under specific conditions). This vague phrasing allows for a perception of extreme speed without committing to a verifiable number, making it difficult to assess the actual performance objectively.
Original quote: “Introducing unclutter, a smart ad plus slot blocker with dev. Basically, auto remove advertising and slot from web pages. and it does so in a fraction of a second.”
Minimizes Jev's cost as 'a few cents maybe per month' and 'nothing' — downplaying potential expenses. — Loaded Language (45/100)
At 11:07
'A few cents maybe' and 'nothing' are not financial projections. That's just hand-waving away the cost. 💸
Why this score: The speaker uses dismissive language ('a few cents maybe per month', 'it's going to be nothing') to describe the cost of using Jev. This downplays potential expenses, especially for users with higher usage, and lacks concrete pricing details. The use of 'maybe' further indicates a lack of precise information, yet the overall tone suggests the cost is negligible, which might not be true for all users or use cases. It's a classic sales tactic to minimize perceived barriers.
Original quote: “I mean, it's free. You just have to bring your own key. You're going to pay Jev a few cents maybe per month. It's like it's going to be nothing.”
Hypes a demo as 'possibly the coolest' and claims 'rebuilt Tesla full self-driving in Jev in less than an hour' — an extreme, unqualified claim. — Confidence Mismatch (45/100)
At 11:17
'Rebuilt Tesla full self-driving' in an hour? That's not a rebuild, that's a toy model. The hyperbole is deafening. 💀
Why this score: The claim that someone 'rebuilt Tesla full self-driving in Jev in less than an hour' is an extreme exaggeration. While a simplified demonstration or a proof-of-concept might be created, 'rebuilding' a complex, real-world system like Tesla's FSD, with all its safety, regulatory, and engineering challenges, in such a short time is impossible. This is hyperbole designed to impress, not an accurate description of the achievement. It's a confidence trick, implying Jev's power is far beyond what's demonstrated.
Original quote: “And then possibly the coolest demo. We have this guy Justin Schroeder who rebuilt Tesla full self-driving in Jev in less than an hour.”
Acknowledges 'wonky' behavior but quickly pivots to 'impressive' due to build time — a volume game. — Volume Game (45/100)
At 11:47
Admits it's 'wonky' but immediately drowns it out with 'impressive for an hour.' That's not an honest assessment, that's a quick save. 🚩
Why this score: The speaker uses a 'volume game' here: they briefly acknowledge a negative ('a little wonky at times') but immediately pivot and amplify a positive ('quite impressive' given the one-hour build time). This technique minimizes the 'wonky' aspect by shifting the focus to the speed of creation, rather than the quality or functionality of the 'rebuilt' system. It's a way to manage expectations while still maintaining a positive narrative, but it doesn't truly address the underlying issues.
Original quote: “You know, it's a little wonky at times, but for essentially building this in an hour, it's actually quite impressive.”
Just setting the stage for the topic — no claim here. — No Frame (75/100)
At 12:30
Just a lead-in, nothing to dissect here. They're just getting started. 😈
Why this score: The speaker is merely introducing the concept of 'speed' as a valuable factor, without making any specific, verifiable claim or using manipulative language. It's a straightforward setup for the discussion to follow.
Original quote: “speed is going to”
See the full analysis with timestamps →