Introducing ChatGPT Work, powered by Codex and GPT-5.6
Credibility score: 52/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "Introducing ChatGPT Work, powered by Codex and GPT-5.6" and rated it 52/100 for credibility (a BS score of 48/100 — mixed credibility), on 2026-07-10. Its weakest claim — "Claims 'outstanding out of the box' – but we only see one example. Cherry-picked? 🍒" — scored 20/100 and was flagged as cherry-picked. 24 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 24 claims analyzed: 1 scored under 40, 17 between 40 and 69, and 6 at 70 or above.
Claims analyzed
Speaker expresses excitement for new GPT models, GPT-5.6 — No Frame (75/100)
At 0:06
Just a standard intro, setting the stage for the big reveal. No tricks here, just good old fashioned product launch hype. 🎉
Why this score: The speaker is introducing new models and expressing enthusiasm, which is typical for a product launch. There's no specific claim being made that requires fact-checking or exposes manipulative framing.
Original quote: “Hi everyone. I am so excited to be here today. We are releasing our latest and most capable models GPD56”
Announces new GPT models Soul, Terra, and Luna, with availability details — No Frame (75/100)
At 0:16
They're just laying out the new product tiers and who gets what. Clear as day, no hidden agendas. 🚀
Why this score: This is a straightforward announcement of new product names and their availability to different user tiers (paid vs. free). It's factual information about their product rollout, not a persuasive claim with hidden framing.
Original quote: “Soul, Terra, and Luna. Soul will be available over the next 24 hours to all our paid plans. And Terra and Luna are also coming to all our free users.”
Claims almost a billion users use 'Cad' — Missing Context (45/100)
At 0:28
Almost a billion users for 'Cad'? What the hell is 'Cad'? Sounds like a typo or a massive leap. 🧐
Why this score: The speaker states 'almost a billion users use Cad' but 'Cad' is not a recognized product or platform in this context. It's likely a mispronunciation or an auto-translation error for 'Codex' or 'ChatGPT' given the video title. Without clarification, the claim is impossible to verify and lacks crucial context.
Original quote: “Today, almost a billion users use Cad”
Claims 'almost a billion users' for ChatGPT weekly, a huge number. — Confidence Mismatch (45/100)
At 0:30
A billion users weekly? That's a massive number for any product, especially one that's not a social media giant. Sounds like a stretch.
Why this score: The claim of 'almost a billion users' weekly for ChatGPT is incredibly high and lacks immediate, verifiable public data from OpenAI to support it. While ChatGPT is popular, this number would put it in the league of Facebook or WhatsApp, which seems unlikely for a generative AI tool at this stage. It's stated with high confidence but without any source or breakdown.
Original quote: “Today, almost a billion users use Cad GBT every week.”
Speaker claims 'really excited' while showing zero excitement. — Just Vibes (50/100)
At 1:50
Says 'really excited' with the energy of someone reading a grocery list. The vibes are not vibing with the words, chief. 😐
Why this score: The speaker's delivery of 'really excited' comes across as very flat and unenthusiastic, creating a comedic mismatch between the stated emotion and the actual presentation. It's a classic case of telling, not showing, when it comes to excitement.
Original quote: “But why I'm really excited to be here today is we're not only releasing the models, we're also announcing three major product updates.”
GPT-5.6 'soul' models are a 'breakthrough' for complex tasks. — Loaded Language (45/100)
At 2:11
Calling it 'GPT-5.6 soul' and a 'breakthrough' without showing anything yet. That's just marketing magic words. ✨
Why this score: The use of 'soul' for a model name and 'breakthrough' for its capabilities are highly emotive and positive terms designed to generate excitement. While the models might be good, these words are doing a lot of heavy lifting to persuade the audience before any actual demonstration or evidence is provided. It's a classic hype-building technique.
Original quote: “The models with GPD56 soul have really been a breakthrough in their ability to do things of complex nature for you.”
New desktop app works on local files, apps, and browser. — No Frame (75/100)
At 2:25
Okay, this is a clear feature announcement. No tricks, just stating what the app will do. ✅
Why this score: This is a straightforward announcement of new functionality for the desktop app. It's a clear, verifiable claim about product features that will either be true or false upon release, without any obvious rhetorical manipulation.
Original quote: “It goes even further and allows chatbt to work on your local files, your apps, and your browser.”
ChatGPT Work 'was able to pull' data and schedule meetings — Confidence Mismatch (45/100)
At 2:30
Claiming the AI 'was able to pull' and 'schedule' meetings without showing the actual output or process. Just trust us, bro. 🪄
Why this score: The speaker states that the AI 'was able to pull' information and 'schedule meetings' based on the previous prompt. However, the actual output, the sources it pulled from, or the scheduling process isn't shown, making it a claim of capability without direct visual evidence in the demo.
Original quote: “the results and chatbt work was able to
pull from the sources that I asked for
um find interesting people to talk to
and then schedule meetings with them
while I'm in town.”
Claiming a $2 million beat, then immediately saying it's fake numbers. 🤡 — Volume Game (45/100)
At 4:30
Announces a win, then immediately pulls the rug out with 'demo numbers.' The classic 'look, but don't touch' move. 🙄
Why this score: The speaker presents a specific achievement ($2 million beat) to create impact, but then quickly qualifies it as 'demo numbers.' This technique allows them to claim a positive outcome without having to back it up with real data, effectively playing both sides of the credibility coin.
Original quote: “or sorry, we just closed June, uh, and we beat our forecast by $2 million. Um, and I do need to caveat, these are demo numbers too.”
Hyping up the 'delight' of interactive websites without showing the actual delight 🙄 — Loaded Language (45/100)
At 6:30
Calling it 'such a delight' is pure emotional appeal — show, don't just tell me it's delightful! 🤩
Why this score: The speaker uses emotionally charged words ('delight', 'highly interactive', 'compelling') to describe the output of ChatGPT Work, but doesn't actually demonstrate the 'delight' or 'compelling' nature of the websites. It's all hype, no proof.
Original quote: “Yeah, it is such a delight to receive these highly interactive and compelling”
ChatGPT Work created a full presentation from a folder and open Chrome tabs in 90 seconds. — Confidence Mismatch (45/100)
At 8:30
Claiming it 'produced a fully ready slide deck' from a folder and open tabs in 90 seconds is a bold flex. We're just taking their word for it. 🪄
Why this score: The speaker describes a very specific and impressive capability of ChatGPT Work: synthesizing diverse information from local files and browser tabs to create a presentation in a company template, all within 90 seconds. While the claim itself is clear, the audience is asked to accept this without seeing the actual output or the complexity of the input materials. It's a 'trust me, bro' moment for a significant feature.
Original quote: “And so, what I'm going what I what I did this morning is I came in and I said, can you please look at the materials in this folder that everybody sent me? Um I also have three open tabs in Chrome that have content for this launch. Go look at it and make me a presentation that I can give to my team…”
Promoting "instant results" and "quick searches" as a core feature. — Plain Sales Pitch (45/100)
At 10:30
Talking up the app's speed and convenience like it's a new invention, not a basic search engine function. 🙄
Why this score: The speaker is presenting standard search engine capabilities (quick, instant results) as a unique, 'great feature' of their app, which is a common sales tactic to highlight basic functionalities as premium benefits.
Original quote: “we've got this we've got this really great feature in the app, um, to do quick searches and anything that you want to know really quickly and get instant results. My friend Tithy wants to know when the World Cup uh, game is so that he can watch it.”
Describing the visualization as 'rich' and 'on demand' without showing the actual data. — Loaded Language (45/100)
At 12:30
Calling it 'rich' and 'on demand' without showing the actual richness. Just tell me what it does, don't sell me on the adjectives. 💅
Why this score: The speaker uses positive, subjective adjectives ('rich,' 'on demand') to describe the visualization without providing concrete examples or a clear demonstration of what makes it 'rich.' This relies on loaded language to create a positive impression rather than objective features. The visual context shows a chart, but the 'richness' is still a subjective claim.
Original quote: “So here's the visualization I made from the spreadsheet. We've got um it's like rich. It's on demand. Uh, it uses the theme that we're that we're using in the app and it kind of went down and ranked like, hey, here are the buckets. Here's how important they are.”
Claims 'outstanding out of the box' – but we only see one example. Cherry-picked? 🍒 — Cherry-Picked (20/100)
At 14:30
Calling it 'outstanding out of the box' after showing one example. We need more than one cherry to make a pie, chief. 🥧
Why this score: The speaker makes a broad claim about the model being 'outstanding out of the box' based on a single demonstration. This could be considered cherry-picking, as a single impressive example doesn't necessarily represent the average or consistent performance 'out of the box' across various use cases.
Original quote: “But it's really outstanding out out of the box.”
Claims a 'night and day' transformation in 3 months, using 'literally' for emphasis. — Loaded Language (45/100)
At 16:30
Went from 'night and day' to 'literally 3 months ago' like that's a precise metric. The word 'literally' is doing overtime here 😭
Why this score: The speaker uses 'night and day' and 'literally' to emphasize a dramatic transformation, but 'literally' is often used hyperbolically. While the change might be significant, the language is designed to evoke a strong emotional response rather than provide a precise, measurable impact.
Original quote: “Yeah. And it's really been, you know, night and day. Literally 3 months ago, spreadsheet today. These like, you know, it's really like transformed the way that the whole company works.”
Claiming anyone can build anything with a few prompts, based on a single example. — Confidence Mismatch (45/100)
At 18:30
Went from a 'lighthearted example' to 'anyone can really build anything' in one breath. That's a leap of faith, not logic 🤸♀️
Why this score: The speaker uses a single, 'lighthearted' example to then make a sweeping generalization about the model's capabilities for 'anyone' to 'build anything' with 'a few prompts.' The confidence in the broad claim doesn't match the limited, specific evidence provided.
Original quote: “know, it can ride all sorts of uh crazy
things. It can ride another pelican. Um
so, this is a lightarted example, but
just like as a creative, it's just like
so amazing how much it's expanded
things. So, um I kicked off that prompt
uh this morning. So, I'm just showing uh
showing another example…”
Claiming Soul autonomously post-trained Luna from a short prompt, replacing senior researchers. — Confidence Mismatch (45/100)
At 20:30
They're showing a prompt that looks simple, then claiming it replaced a 'team of senior researchers.' That's a leap from 'short prompt' to 'automated researcher' 🚀.
Why this score: The speaker presents a short prompt and then immediately extrapolates that this capability means 'the automated researcher is pretty close' and that it previously required 'a team of senior researchers.' While the prompt might be efficient, the leap to replacing an entire team or being 'pretty close' to an automated researcher is a significant overstatement of the evidence presented. It's a confidence mismatch between the specific example and the broad conclusion.
Original quote: “Already soul has been transforming our research program. As one example, 5.6 Soul actually autonomously post-trained Luna. This is the actual codeex prompt that we a researcher on our team used to have soul kick off a post-raining job for Luna. You can see here that, you know, it's like asking to…”
Claims experts say it's three times faster and better than anything else, but names no experts. — Anonymous Authority (45/100)
At 22:30
Oh, 'experts in the field' say it's three times faster? Which experts, exactly? The ones in the room? 🕵️♀️
Why this score: The speaker attributes a strong claim about performance and superiority to unnamed 'experts in the field.' This uses the weight of authority without providing any verifiable sources or specific individuals, making the claim hard to scrutinize.
Original quote: “all of this is something that when we ask experts in the field, you know, how they're using this model in their everyday workflows, they say it's better than anything else they've ever used, anything else on the market while being three times as fast. So, we're really excited for you to finally try…”
GPT 5.6 Soul found vulnerabilities in every major browser and database. — Confidence Mismatch (45/100)
At 24:30
Claiming 'every major browser and database' is a huge flex, but without naming names or showing receipts, it's just big talk. 🤷♀️
Why this score: The speaker makes a very broad and confident claim about the model's capabilities without providing any specific examples or evidence. 'Every major browser and database' is an incredibly sweeping statement that would require extensive, verifiable proof to back up. It sounds impressive but lacks the detail to be truly convincing.
Original quote: “with GPT 5.6 soul researchers have already found vulnerabilities in every major browser and database.”
GPT 5.6 Soul found vulnerabilities in every major browser and database. — Confidence Mismatch (45/100)
At 24:30
Claiming 'every major browser and database' is a huge flex — but without naming names or showing proof, it's just vibes. 💅
Why this score: The speaker makes a very broad and absolute claim about finding vulnerabilities in 'every major browser and database' using GPT 5.6 Soul. This is a massive assertion that would require significant, specific evidence (like a list of browsers/databases, CVEs, or research papers) to back up. Without any of that, it sounds like an overconfident generalization.
Original quote: “with GPT 5.6 soul researchers have already found vulnerabilities in every major browser and database.”
Setting up ChatGPT as a live translator for a Japanese speaker. — No Frame (75/100)
At 28:30
They're just setting up the demo here, explaining how the live translation will work. Straightforward. 🤖🗣️
Why this score: This is a clear, direct setup for a demonstration. The speaker is explaining the functionality they are about to showcase, which is a live translation feature using ChatGPT. There's no hidden agenda or manipulative language; it's purely instructional for the audience.
Original quote: “Hey chat, I'm here with Hiokisan, a dear and esteemed friend who traveled all the way from Japan. I want you to act as my live translator. He's going to speak Japanese live translate to English to me and I'm going to be speaking English live translate back to Japanese.”
Speaker claims receiving messages from farmers globally and contact from Japan's Ministry of Agriculture. — No Frame (75/100)
At 30:30
Just sharing personal anecdotes about recent interactions — no grand claims, just vibes. 🌍✉️
Why this score: The speaker is recounting personal experiences and interactions, which are inherently subjective and not presented as verifiable facts. It's a personal story, not a claim requiring external evidence.
Original quote: “Lately, I've been receiving messages from farmers all over the world. >> And I've also been contacted by people from Japan's Ministry of Agriculture, Forestry, and Fisheries. All sorts of unexpected things have been happening.”
Claiming the advice is 'profound' and product updates are rolling out in 24 hours. — Confidence Mismatch (45/100)
At 32:30
Calling the advice 'profound' is a bit much, it's pretty standard 'start small' stuff. And 'rolling out over the next 24 hours'? That's a bold promise. 🚀
Why this score: While the advice itself is reasonable, labeling it 'profound' is an attempt to elevate its significance beyond what was actually presented. The claim about product updates rolling out in 'the next 24 hours' is a specific, time-bound statement that would require immediate verification, and while possible, it's delivered with a high degree of confidence that might not always align with the realities of software deployment.
Original quote: “Everything you just saw and the advice from Hioki is profound. It is about taking small steps and raising the level of ambition you have in order to have chips work get things done for you. What you saw and all of the product updates and the models are rolling out over the next 24 hours. Soul”
Announcing a bunch of new product rollouts and availability tiers, all happening 'over the next 24 hours' or 'next day'. — No Frame (75/100)
At 34:30
Just a straightforward product announcement and rollout schedule. No tricks, just telling us what's coming and when. 🚀
Why this score: The speaker is simply detailing the availability of new features (Soul, Chad work desktop app, Terran Luna) and their respective rollout timelines (next 24 hours, next day) for different user tiers (paid vs. free). This is a standard product update, not employing any specific framing techniques to mislead or persuade beyond the inherent excitement of a launch.
Original quote: “the product updates and the models are rolling out over the next 24 hours. Soul will be available for all our paid plans together with Chad work the desktop app and sites as well. The Chad desktop app is also available for free users. Soul is available for all our paid plans and Terran Luna for…”
See the full analysis with timestamps →