I Don't Need Frontier Models Anymore (Qwen 3.8 27B + DeepSeek Harness)
Credibility score: 49/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "I Don't Need Frontier Models Anymore (Qwen 3.8 27B + DeepSeek Harness)" and rated it 49/100 for credibility (a BS score of 51/100 — mixed credibility), on 2026-09-17. Its weakest claim — "Citing specific numbers for 'massive' improvements — but only the good ones. 🍒" — scored 20/100 and was flagged as cherry-picked. 43 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 43 claims analyzed: 3 scored under 40, 33 between 40 and 69, and 7 at 70 or above.
Claims analyzed
Declares Qwen 3.8 27B the awaited model — pure enthusiasm, no data yet. — No Frame (75/100)
At 0:01
Just pure, unadulterated excitement for a new toy. Nothing to dissect here, just vibes. 😈
Why this score: The speaker is expressing personal anticipation and excitement for a new AI model release. This is a subjective statement of personal preference and expectation, not a factual claim about the model's performance or capabilities. It's a straightforward expression of enthusiasm.
Original quote: “when 3.8 27 billion is out. Okay, this is the model that I was waiting for.”
Claims 80% of work done by previous model — a personal usage stat with no context. — Confidence Mismatch (45/100)
At 0:16
He's throwing out an '80%' like it's a hard number, but it's just his personal feeling. That's not a metric, that's a vibe. 💀
Why this score: The speaker states that '80% of the work' is done with the previous model. While this is a personal anecdote, presenting it as a specific percentage implies a level of measurement or data that isn't provided. It's a confident, specific number based on subjective experience, not objective tracking, making it a confidence mismatch between the precision of the claim and the lack of underlying data.
Original quote: “I do 80% of the work with that model. I like it a lot.”
Comparing Qwen to Obus 4.6 — based on what, exactly? 💀 — Anonymous Authority (45/100)
At 0:46
He says 'a lot of people are comparing it' to Obus 4.6, but offers no source, no data, no names. Just 'a lot of people.'
Why this score: This is the classic 'everyone knows' move. It lends weight to a comparison without having to actually provide any evidence or specific examples of who 'a lot of people' are. It's a vibe, not a fact.
Original quote: “This is what Quen released and this is the reason why a lot of people are comparing it to Obus 4.6.”
Citing specific numbers for 'massive' improvements — but only the good ones. 🍒 — Cherry-Picked (20/100)
At 1:05
He's showing specific, impressive jumps in 'agentic coding' and 'software engineering' — but only after admitting 'in some scenarios the gap is not that big.'
Why this score: He's highlighting the most dramatic improvements while glossing over the less impressive ones he just mentioned. It's a classic move: show the best numbers to build excitement, then quietly move past the average ones. It's not a lie, but it's not the whole truth.
Original quote: “But other scenarios like here aentic coding the gap is huge from 13.3 to 42.2 is massive and software engineering from 49.3 to 79. This is another massive step forward on frontier agenting task from 10.6 to 20. This is 100% improvement.”
One personal test, one bug difference — then he stops. That's not data, that's a mood. 💀 — Confidence Mismatch (45/100)
At 1:51
He ran 'my test bug hunting,' found one bug difference, and then 'stopped the test' because he was 'disappointed.' That's not a rigorous comparison.
Why this score: Basing a 'disappointing' conclusion on a single, limited personal test with a one-bug difference is a leap. It's an anecdote, not a benchmark. His confidence in the disappointment doesn't match the extremely thin evidence.
Original quote: “In my test bug hunting, the Q6 found one more bug than the Q5, even though the Q5 was the newer version.”
From 'hard to tell' to '30% improvement' in one breath. That's not analysis, that's a guess with a number. 😈 — Confidence Mismatch (45/100)
At 2:21
He admits 'it's hard to tell' the difference, then immediately pulls 'a 30% improvement' out of thin air. No data, just a number.
Why this score: He's trying to quantify an improvement he just said was 'hard to tell.' The 'let's say' is the tell — it's an estimate presented as a concrete percentage, giving it an air of authority it hasn't earned. It's a guess dressed up as a stat.
Original quote: “it's hard to tell, but let's say a 30% improvement, which is noticeable, but not another level.”
Attributing high reasoning improvement to 'Luna itself' — vague cause. — Anonymous Authority (45/100)
At 2:30
He says 'Luna itself' did a 'great job' improving reasoning. What does that even mean? 🧐
Why this score: The statement attributes a 'great job' to 'Luna itself' without specifying what 'Luna' is or how it achieves this improvement. It's a vague assertion that lacks concrete details or evidence.
Original quote: “This is because Luna itself could do such a great job when improving reasoning.”
Describing a score increase as 'insane' and 'more than double' — using loaded language. — Loaded Language (45/100)
At 2:50
He calls a jump from 27 to 52 'insane' and 'more than double.' The math is right, the drama is extra. 😈
Why this score: While the numerical increase (27 to 52) is indeed more than double, describing it as 'insane' is loaded language. It's an emotional descriptor rather than a purely objective one, designed to amplify the perceived significance of the jump.
Original quote: “The jump is insane is more than double the capability.”
Hermes's subscription model is 'the wrong path' — framing a business decision as a moral failing. — Loaded Language (45/100)
At 4:30
A company charging for its service is a business model, not a 'wrong path.' He's just not a fan. 🙄
Why this score: The speaker frames Hermes introducing a subscription as 'taking the wrong path,' implying a moral or strategic error rather than a standard business decision. This uses loaded language to justify his move to build an alternative, rather than simply stating a preference for open-source or co-owned systems.
Original quote: “recently when they start pushing their own subscription I felt like yeah that is not cool anymore I felt like a client I'm a guest within their system and this is the reason why I'm building with my community Resonant OS so we can co-own it and it's ours. The moment they started pushing that…”
Declares DeepSeek Harness + Qwen 3.8 27B the 'best combo' with no evidence. — Confidence Mismatch (45/100)
At 5:47
He says 'I believe' and then declares it the 'best combo' like it's gospel. That's not belief, that's a sales pitch. 😈
Why this score: The speaker states his belief that this combination is the 'best combo' without offering any comparative data, benchmarks, or specific reasons beyond his personal experience, which he only vaguely alludes to later. This is a strong, confident claim based purely on personal opinion, presented as a definitive statement.
Original quote: “Now, I believe that this is the best combo for Quen 3.8 27 billion.”
Claims the new combo solves 'everything' he wanted to build, based on vague past issues. — Confidence Mismatch (45/100)
At 6:07
From 'sometimes found a problem' to 'everything I wanted to build it built' in one breath. That's not a metric, that's a miracle. 🙄
Why this score: The speaker makes a very broad claim that the new DeepSeek Harness + Qwen combo has built 'everything I wanted to build so far.' This is a subjective, anecdotal claim that lacks specific examples or quantifiable metrics. He contrasts it with vague past issues where Qwen 3.6 'sometimes' found 'unsolvable problems,' but provides no details to back up the dramatic improvement, making the confidence in 'everything' seem unearned.
Original quote: “I start using it together with this uh new version of Quen and everything I wanted to build so far it built. So in the past I used to use Quen 3.6 27 billion let's say 80% of the time. Sometimes it would find a problem that was not solvable. I had to change model and use one of those frontier…”
DeepSeek Harness and Qwen 3.8 27B mean no more frontier models. — No Frame (75/100)
At 6:30
He's laying out his main point: these smaller models are now good enough. Straightforward, for once.
Why this score: The speaker is introducing his core argument that the combination of DeepSeek Harness and Qwen 3.8 27B has eliminated his need for larger, 'frontier' AI models. This is a direct statement of his thesis, not a rhetorical trick.
Original quote: “model. Now with this deepseek harness”
Can't show it running due to 'memory management' and recording issues. — Confidence Mismatch (45/100)
At 7:23
He's hyping up this 'amazing' thing he can't actually show you working. Convenient, isn't it? 😈
Why this score: The speaker is making a strong claim about the amazing performance of the new setup but immediately follows it with an excuse for not being able to demonstrate it live. This creates a disconnect between the confidence of his praise and the lack of direct visual evidence, which is a classic move to avoid scrutiny.
Original quote: “I can't show you while it's running
because while it's running I can't
record the screen because memory
management of the file format that I'm
uh recording.”
It's 'amazing' but also 'not a finished product' and 'developer preview'. — Volume Game (45/100)
At 7:33
He calls it 'amazing' then immediately says it's unfinished and buggy. That's how you get to claim both sides. 💀
Why this score: The speaker uses highly positive, emotional language ('just amazing') to describe the product, then immediately downplays expectations by stating it's 'not a finished product' and a 'developer preview'. This allows him to generate excitement while simultaneously preemptively excusing any potential flaws or shortcomings, a classic volume game.
Original quote: “What I can tell you here
is that it's just amazing. It's not a
finished product. Okay. Okay.”
Admits it crashes but says it's 'nothing serious' and 'unbelievable' how it manages context. — Volume Game (45/100)
At 7:45
It crashes, but 'nothing serious' — just 'resume working.' Then back to 'unbelievable.' The whiplash is intentional. 🔥
Why this score: The speaker acknowledges significant flaws, like system crashes, but quickly dismisses them as 'nothing serious' because you can 'resume working.' He then pivots back to effusive praise, calling the context window management 'unbelievable.' This rapid shift between admitting serious issues and then immediately hyping the product is a classic volume game, designed to minimize the impact of negative information while maximizing the positive.
Original quote: “It's there is a lot of things
maybe are missing. Things are not
perfect. So sometimes it doesn't do
compaction in the right moment and
therefore it's a kind of crash the
system nothing serious because you just
say resume working and it start working
without any problems but nevertheless
it's still…”
Can't explain the 'unbelievable' context window because it's 'confusing' to him. — Confidence Mismatch (45/100)
At 8:27
He just called it 'unbelievable' and now he can't explain it because it's 'confusing.' That's not a feature, mortal, that's a dodge. 😈
Why this score: After calling the context window management 'unbelievable,' the speaker admits he can't explain it because it's 'confusing' to him. This creates a significant confidence mismatch. He's praising something he doesn't fully grasp, which undermines the authority of his 'amazing' claims and suggests a lack of deep understanding rather than genuine insight.
Original quote: “window. I can't explain because it's
confusing to me. So I need to read”
Can't explain the tech, but claims 131,000 tokens are available. Confidence Mismatch. — Confidence Mismatch (45/100)
At 8:30
He says he can't explain it, then confidently states a specific token count. That's not how 'knowing' works. 💀
Why this score: The speaker admits to not understanding the underlying mechanics ('confusing to me') but then immediately presents a precise technical detail (131,000 tokens) as a known fact. This is a classic confidence mismatch, where certainty is projected without demonstrated comprehension.
Original quote: “window. I can't explain because it's confusing to me. So I need to read understand it better. But what I know it's these 131,000 tokens that I have available for this model on my 5090.”
Claims 'compaction is solid' and system is 'still going strong' after 38 million tokens. Just Vibes. — Just Vibes (50/100)
At 8:39
He's just throwing out 'solid' and 'strong' like they're metrics. It's a feeling, not a fact. 😈
Why this score: The speaker uses subjective, qualitative terms like 'last longer,' 'so much more solid,' and 'still going strong' to describe the model's performance, without providing any objective data or benchmarks to support these claims. It's a personal impression, not a technical assessment.
Original quote: “They just last longer. The compaction is so much more solid. And this system is been chatting for so long, building for so long. That is incredible. So if you I don't know if you can see, but here it had an input of 38 million tokens, still going strong.”
Dismisses the quality of the AI's output because it lacked context. Missing Context. — Missing Context (45/100)
At 9:55
He's downplaying the results, saying 'it didn't have context.' That's not a flaw of the AI, it's a flaw of the prompt. 🚩
Why this score: The speaker attributes the potentially low value of the AI's generated ideas to a lack of context provided to the model. While true, this shifts the responsibility from the model's capability to the user's input, effectively providing an excuse for suboptimal results rather than a direct evaluation of the model's performance in a real-world, well-contextualized scenario.
Original quote: “Now, it doesn't matter the ideas because it's honestly I don't from what I see maybe are valuable, maybe not, but doesn't matter because it didn't have the instruction of what is this channel is about and so on. So, at the moment it doesn't have much context. It just went there and grabbed…”
Claims no need for bigger models, a shift from past struggles. — No Frame (75/100)
At 10:30
He's just stating his personal experience with these models. It's his journey, not a universal law. 😈
Why this score: The speaker is sharing a personal observation about his workflow and the capabilities of the models he's using. This is presented as his own experience, not a generalizable fact or a sales pitch. It's a straightforward personal testimony.
Original quote: “I never had the need to use a bigger model. Something that always happened before when was doing something as complex at this kind of tasks.”
Describes local AI as a 'journey' not for everyone, but improving. — No Frame (75/100)
At 11:07
He's acknowledging the difficulty and ongoing nature of local AI. A rare moment of honesty. 🔥
Why this score: The speaker is being realistic about the adoption of local AI, admitting it's a process and not universally accessible. This adds a layer of nuance and avoids overpromising, making it a straightforward, honest assessment.
Original quote: “This is the journey but it's not first of it's not for everybody but also is it's a journey is not a non-off what I can see with this deepseek harness that the journey is getting better and better”
Highlights MIT license for commercial use, then pivots to audience access. — No Frame (75/100)
At 11:19
He's just laying out the licensing terms and then making a reasonable assumption about his audience. Nothing tricky here. 😈
Why this score: The speaker is simply explaining the benefits of the MIT license for the DeepSeek harness, which is a factual statement. He then makes an assumption about his audience's access to the model, which is a logical inference rather than a manipulative claim.
Original quote: “because now we have this harness that has MIT license meaning you can use it in your project you can use it yourself implement change it do whatever and use for commercial so what I'm suggesting here if you are watching this video probably it's because you have access to this model.”
Suggests DeepSeek v4 Flash for owners of high-end hardware. — Plain Sales Pitch (45/100)
At 11:40
He's pushing a specific product combo for the 'lucky' few. That's a pitch, not just a suggestion. 💰
Why this score: While presented as a suggestion, the speaker is actively promoting a specific product (DeepSeek v4 Flash) and a specific hardware setup (two DJX Spark or equivalent) to a niche audience. This moves beyond a general recommendation into a direct product endorsement, even if it's for a very high-end user base.
Original quote: “But even if you don't try this uh DeepS harness, if you're one of those lucky people that own a two DJX Spark or equivalent, try DeepSk version 4 flash because can run on two of them with this harness that probably they could work amazingly together.”
Compares current AI excitement to the 'massive push forward' of OpenClaw. — Confidence Mismatch (45/100)
At 12:20
He's drawing a grand parallel to a moment he can't even fully recall. That's not history, that's just vibes. 💀
Why this score: The speaker is making a strong comparison to a past event (OpenClaw's release) that he admits he 'can't even remember anymore.' This undermines the authority of the comparison, as he's expressing high confidence in the significance of a moment he struggles to recall details about. It's an emotional appeal rather than a precise historical analogy.
Original quote: “So when it came out it just was a massive push forward because everybody finally had this agent that can do a lot take control of your computer work for”
DeepSeek's harness is a 'similar jump' — vague comparison with no specifics. — Loaded Language (45/100)
At 12:33
A 'similar jump' to what, exactly? That's not a metric, it's just hype. 😈
Why this score: The speaker uses 'similar jump' to imply a significant, comparable improvement without specifying what it's similar to, or providing any quantifiable metrics for this 'jump'. It's an emotional appeal to perceived progress rather than a factual comparison.
Original quote: “DeepSync did with their harness is a similar jump”
The harness allows the model to do 'so much more' — another vague, unquantified claim. — Loaded Language (45/100)
At 12:39
So much more? How much? What exactly? This isn't a claim, it's a feeling. 💀
Why this score: The phrase 'so much more' is subjective and lacks any specific examples or metrics. It's designed to create an impression of vast improvement without providing any verifiable details, leaving the audience to fill in the blanks with their own optimistic interpretations.
Original quote: “allows this model to do so much more”
Frames strong claims as 'what I feel' — personal opinion presented as a conclusion. — Personal Story (60/100)
At 12:50
Oh, 'what I feel.' So, not data, just vibes. Got it. 💀
Why this score: The speaker explicitly states that their strong conclusions are based on 'what I feel,' shifting the basis from objective evidence to personal sentiment. While valid as a personal experience, it's not a verifiable claim about the model's universal performance.
Original quote: “at moment this is what I feel.”
Compares DeepSeek to Hermes and OpenCode, but without specific metrics or context. — Missing Context (45/100)
At 13:00
Can't say the same? What exactly can't you say? Give us the damn details. 😈
Why this score: The speaker makes a comparative claim ('can't say the same') without providing any specific criteria, benchmarks, or examples of how Hermes or OpenCode fall short. This leaves the audience without the necessary context to evaluate the comparison.
Original quote: “something that I can't say the same with Hermes or open code.”
Claims 'the result is there' for increased thinking time, but offers no specific data. — Confidence Mismatch (45/100)
At 14:44
He says 'the result is there' like it's a given, but where's the actual proof? Just vibes and a vague direction. 💀
Why this score: The speaker asserts a clear 'result' and 'direction' regarding increased thinking time leading to better performance, but provides no concrete data, benchmarks, or studies to back this up. It's a statement of conviction without evidence.
Original quote: “the result is there. We know that the direction was increase the thinking time”
Minimizes a 0.2% loss from Q8 as 'pretty acceptable' without context. — Missing Context (45/100)
At 15:17
0.2% loss sounds tiny, but without knowing what that percentage *means* in performance, it's just a number. 😈
Why this score: The speaker presents a 0.2% performance loss as 'pretty acceptable' without explaining what that percentage represents in real-world terms (e.g., speed, accuracy, bug detection rate). This lack of context makes it difficult to assess the true impact of the compromise.
Original quote: “And look, I only lose 0.2% from the Q8. So the compromise is pretty acceptable.”
Highlights one more bug found by Q6 over Q5 as significant. — Cherry-Picked (20/100)
At 15:47
One bug. Just one. And suddenly it's 'annoying' enough to justify a whole tier. That's not data, that's a sales pitch. 🍒
Why this score: The speaker emphasizes that Q6 found 'one more bug' than Q5 in bug hunting. While this is a difference, presenting a single data point (one bug) as the primary differentiator for an 'annoying' performance gap is a classic cherry-pick, potentially overstating its significance without broader context of total bugs, types of bugs, or overall performance metrics.
Original quote: “the bug hunting, the Q6 could find one more bug than the Q5.”
Admits to not trying the models, then immediately speculates on their performance. — Confidence Mismatch (45/100)
At 16:32
Says he hasn't tried them, then pivots to 'you lose a lot' as if it's a known fact. That's not a disclaimer, that's a setup. 😈
Why this score: The speaker explicitly states he hasn't tried the models, which should lead to a lack of definitive statements. However, he then implies significant performance loss without direct experience, creating a mismatch between his stated lack of knowledge and his implied certainty.
Original quote: “So while I never tried these two models, okay, I never tried them. I can't say”
Suggests ideal VRAM is 24-48GB, framing it as a necessity for optimal performance. — Loaded Language (45/100)
At 16:41
He says 'ideally you want to be in the 24, 32, 48' like it's a casual suggestion, but it's a massive jump from the 'minimum 12GB.' That's not ideal, that's a flex. 💀
Why this score: The speaker presents a significant jump in VRAM requirements (from 12GB minimum to 24-48GB ideal) as a casual 'want to be.' This uses loaded language to imply that anything less than these high-end specs is suboptimal, potentially pushing viewers towards more expensive hardware without fully justifying the 'ideal' leap.
Original quote: “This is something that you need to test yourself. But knowing that the minimum minimum requirement is 12, ideally you want to be in the 24, 32, 48 if you can do it. So now another elements you need to consider is the side of the memory that is required to run this model.”
Claims the 5090 is 'the best one' with 'insane' bandwidth, linking it to high token generation. — No Frame (75/100)
At 17:16
He's just stating the specs and his experience. The 5090 *is* a beast, no argument there. 🔥
Why this score: The speaker is providing specific technical details (bandwidth, token per second) and his personal experience with the 5090. While 'best' and 'insane' are subjective, the underlying technical claims about the 5090's performance are generally accurate for a high-end GPU, and he's transparent about it being 'on my hardware.'
Original quote: “the best one is the 5090. Okay. This has a bandwidth that is insane. 1,792 GB per second. That means the this model is working at around 100 110 token per second on my hardware. It's doing insanely well.”
Compares AMD's 640 GB/s to NVIDIA's 1792 GB/s, highlighting it as 'almost a third of the speed.' — Cherry-Picked (20/100)
At 17:37
Picks the highest NVIDIA spec and compares it to a lower AMD one to make AMD look bad. That's not a comparison, that's a hit job. 🍒
Why this score: The speaker selects a specific AMD card's bandwidth (640 GB/s) and directly compares it to the top-tier NVIDIA 5090 (1792 GB/s), stating it's 'almost a third of the speed.' This comparison is cherry-picked to highlight a significant performance gap, potentially ignoring other AMD cards that might offer more competitive performance or different price-to-performance ratios, thus creating a biased perception.
Original quote: “one, you can see that the AMD is doing 640. This is almost a third of the speed. So you have an idea of what you can expect.”
Refers to the '1390' when discussing GPUs, likely a misstatement for 3090. — Just Vibes (50/100)
At 17:58
He means the 3090, obviously. A slip of the tongue, or maybe he's seen the future. 😈
Why this score: The speaker likely made a minor verbal error, saying '1390' instead of '3090.' This is a common slip in rapid speech and doesn't fundamentally alter the meaning or intent of his discussion about the GPU, which is clearly understood from context to be the RTX 3090.
Original quote: “The good thing with 1390 that you can find them”
Estimating context window size with 'probably I guess' — pure speculation presented as calculation. — Confidence Mismatch (45/100)
At 18:48
He's throwing out numbers with 'probably I guess.' That's not a calculation, it's a guess with a calculator nearby. 💀
Why this score: The speaker uses specific numbers (131,000, 260 something) but prefaces the larger number with 'probably I guess,' indicating a lack of certainty or actual data. This presents a speculative estimate as if it's a direct extrapolation, when it's just a hopeful guess.
Original quote: “Have in mind that you on my 32 GB of RAM I can have 131,000 token contest window with 48. You probably I guess you could have the the total 260 something context window which is a nice plus.”
Claiming 'usable' system at 30 tokens/second — subjective metric presented as objective. — Loaded Language (45/100)
At 19:11
He says 'usable' at 30 tokens/second like it's a universal truth. 'Usable' for *him* maybe. That's not a metric, it's a feeling. 😈
Why this score: The speaker states that 'at 30 tokens per second, in reality, you have a system that works, is usable.' While 30 tokens/second might be acceptable for some use cases, 'usable' is a subjective term that depends heavily on the user's expectations and application. Presenting it as an objective reality is loaded language.
Original quote: “bandwidth. It is something to consider at the moment. If you want to buy new, this is really good. If you're looking for secondhand, this is pretty good. One more thing, this model can run on a DGX Spark and it runs depending on the quantization at around 20 30 to per second. At 30 to per second,…”
Declaring no need for bigger models after only 24 hours — a tiny sample size for a grand conclusion. — Confidence Mismatch (45/100)
At 20:11
Twenty-four hours, and he's already swearing off bigger models forever. That's not a conclusion, that's a honeymoon phase with a thesis. 💀
Why this score: The speaker makes a strong claim about not needing bigger models, basing it on 'early days' and 'the last 24 hours' of use. While personal experience is valid, drawing such a definitive conclusion from such a short testing period, especially for complex AI models, is a significant leap of confidence. It's an anecdote presented as a universal finding.
Original quote: “It's early days to say, but so far in the last 24 hours, I use it a lot. And I never felt I needed to use a bigger model. Never once.”
Standard call to action for engagement and community — nothing tricky here. — No Frame (75/100)
At 20:35
Just a straightforward ask for likes, subs, and to join their Discord. Standard creator stuff. 😈
Why this score: This is a typical content creator's call to action, encouraging viewers to engage with the content and community. There's no deceptive framing or hidden agenda; it's a direct request for support and participation.
Original quote: “If you're interested in this kind of content and you want to join our community, first like and subscribe. In the description, you find the link to our discord channel and there is a deep dive article. I will go deeper in everything that I talk inside this uh video.”
Promising to share a browser extension, but with a big 'if possible' attached. — Volume Game (45/100)
At 20:45
He's hyping up this 'amazing' browser extension, but then immediately walks it back with 'if possible'. Classic bait and switch. 🚩
Why this score: The speaker introduces a browser extension, 'augumentor powered by DS DSH,' with strong positive language ('honestly is amazing'). However, this enthusiasm is immediately tempered by a significant caveat ('if possible'), which diminishes the certainty of its availability. This creates a 'volume game' where a bold claim is made, then quietly retracted or made conditional, managing expectations while still generating interest.
Original quote: “I will add all the links that I discuss over here and if possible I will see how it goes. But I might also share this augumentor powered by DS DSH. This is an extension for the browser and by adding it you can have this sidecar which honestly is amazing.”
Hypes an 'amazing' tool, then admits it's not ready for public use because it lacks basic user configuration. — Confidence Mismatch (45/100)
At 21:03
He calls it 'amazing' and 'actually amazing' — then explains it's missing the core feature needed for anyone to actually use it. That's not amazing, that's a beta. 💀
Why this score: The speaker repeatedly emphasizes the 'amazing' quality of the browser extension, yet immediately follows this with a detailed explanation of its current limitations. The crucial missing feature is a user-friendly configuration system for adding models, which means it's not ready for general public use. This creates a 'confidence mismatch' where the strong positive assertion about the tool's quality doesn't align with its admitted lack of fundamental functionality for users.
Original quote: “This is an extension for the browser and by adding it you can have this sidecar which honestly is amazing. It is actually amazing. I say if I can because what is missing before I can share it is a way for you to add your own model. At the moment there's no a configuration system where you can add…”
Promises to share code 'if in time,' otherwise 'in the near future' — a vague commitment. — Volume Game (45/100)
At 21:45
He's giving you a 'maybe now, maybe later' on the code. That's not a promise, that's a shrug with extra words. 🔥
Why this score: The speaker offers a conditional promise to share the code for the agent, stating 'if I can manage to do this in time.' If not, it will be shared 'in the near future.' This phrasing uses vague temporal language ('in time,' 'near future') to create a sense of commitment without a firm deadline. It's a 'volume game' because it sounds like a definite plan, but the actual delivery is left open-ended, allowing for indefinite delays without breaking a specific promise.
Original quote: “So if I can manage to do this in time, I will link inside the Substack article the code for this um agent. If not, will be shared in the in the near future.”
See the full analysis with timestamps →