I Think They Are Lying To You
Credibility score: 50/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "I Think They Are Lying To You" and rated it 50/100 for credibility (a BS score of 50/100 — mixed credibility), on 2026-06-12. Its weakest claim — "Two years of code shipped every quarter per employee" — scored 25/100 and was flagged as sketchy. 19 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 19 claims analyzed: 2 scored under 40, 16 between 40 and 69, and 1 at 70 or above.
Claims analyzed
Says coding is fully solved and now the easy part — Opinion (30/100)
At 0:30
Top comment already called it: they’re still hiring engineers to clean up the slop.
Why this score: Subjective framing — no metric defines when coding is "solved."
- Hiring data from Anthropic and peers shows continued demand for engineers fixing AI output.
- Infrastructure and user feedback being called the real problems is a convenient pivot away from current model limitations.
Original quote: “coding is mostly solved. Now, he just says coding is solved. And even more, he says coding is the easy part.”
Shipping 8x more code per employee in Q2 2026 vs pre-2025 — Unverifiable (50/100)
At 2:30
8x per employee with zero data, metrics, or company named — pure number, no receipt.
Why this score: No public data backs the 8x figure.
- Speaker presents it as a direct quote from unnamed source
- No company, team size, or measurement method given
- Q2 2026 is current quarter; impossible to verify externally yet
Original quote: “In fact, in this last quarter, Q2 2026, we are shipping 8x the amount of code per employee than we were on the average before 2025.”
Two years of code shipped every quarter per employee — Sketchy (25/100)
At 2:38
8x quarterly output equals 32x yearly — math only works if pre-2025 baseline was near zero.
Why this score: Simple math exposes the stretch.
- 8x per quarter = 32x annual output vs pre-2025
- That implies almost no code was written before 2025
- Real engineering teams don't go from near-zero to 32x without massive hiring or scope changes
Original quote: “In other words, they are shipping two years worth of code every single quarter for every single employee.”
Coding is solved — humans only prompt and review AI output — Opinion (50/100)
At 2:50
Bold take, but top comments already roasting the 'hiring engineers to fix slop' reality.
Why this score: Industry pushback is loud.
- Multiple top comments note teams still hire to clean up AI-generated code
- 'Prompt the prompters' framing ignores ongoing debugging and architecture work
- No evidence that review is fully automatic at scale
Original quote: “Coding is solved, bro. We don't write code no more. Okay? We prompt the prompters. All right? Big difference. We write loops. We do not code ourselves. Review code. It's all done automatically.”
AI hype is lying to and hurting developers — Opinion (50/100)
At 3:50
Personal take on industry messaging — frames it as actively harmful to devs' mental health.
Why this score: Strong opinion, not a fact claim.
- Based on 'lots of messages' from viewers (anecdotal)
- Speaker links productivity hype directly to burnout and job anxiety
- No specific studies or company policies cited as evidence
Original quote: “I think they are lying, and I think that they're hurting people. I think they're genuinely hurting people. First off, I get lots of messages about people genuinely worried about the future and hey, maybe there's no room for me left.”
Claude Code released February 2025 for researchers — Unverifiable (50/100)
At 5:02
No public record of that exact February 2025 researcher release — could be right, just nothing confirming it.
Why this score: Unverifiable release date.
- Web search shows no official Anthropic announcement matching February 2025 for a researcher-only Claude Code terminal.
- Absence of evidence here doesn't prove it didn't happen internally.
- Score stays neutral rather than calling it false.
Original quote: “Claude code was released for researchers in February 2025. Now, if you're unfamiliar, Claude Code is a little terminal application in which you can prompt Claude to go and do some angentic work and produce code.”
Speaker personally saw flickering bug in 2024, thinks Anthropic knew — Personal Story (60/100)
At 5:54
Personal experience claim — can't fact-check someone's own screen time, but the 2024 pre-release guess stays speculative.
Why this score: Anecdote, not evidence of corporate knowledge.
- Speaker's own usage memory is valid personal testimony.
- The leap to "they actively knew about it" in 2024 is an inference, not documented proof.
- Reasonable to be skeptical of the timeline guess without internal logs.
Original quote: “this bug most likely existed not even in 2025, potentially in 2024 while it was all being developed and they actively knew about it is my personal guess. If you used the product at all during that time period, it was so evident. I remember having my own experiences with it, being shocked people…”
Anthropic publicly claimed 85% flicker reduction on Dec 17, 2025 — Unverifiable (50/100)
At 6:19
No search results confirm that December 17, 2025 statement or the 85% figure — could exist internally.
Why this score: Quote presented as fact but unsupported by public records.
- Web context has zero hits on an Anthropic announcement matching that date or the 85% reduction claim.
- The speaker treats it as a direct quote, so the absence is notable.
- Stays at 50 rather than calling fabrication without positive contradictory evidence.
Original quote: “They finally respond to it publicly December 17th, 2025. And I quote, "We've rewritten Claude Code's terminal rendering system to reduce flickering by roughly 85%."”
Anthropic claims 85% flicker reduction fix — Dubious (45/100)
At 6:30
85% reduction on a pure software bug after nine months — why not 100%? That number feels like a press release, not engineering.
Why this score: Partial fix claim lacks context. Key issues:
- No independent verification of the 85% figure
- Speaker notes the problem persisted over a year, contradicting "largely solved" narrative
- Pure software issue with no hardware constraints mentioned
Original quote: “reduce flickering by roughly 85%.”
Speaker claims 85% bug fixes are unprecedented — Opinion (50/100)
At 6:35
Hyperbole — plenty of stochastic fixes exist in complex systems. The real point is the timeline, not the percentage.
Why this score: Personal incredulity presented as evidence.
- No data showing 85% fixes are rare
- Misses that partial rollouts are standard practice
- Stronger argument comes later with the December rollback
Original quote: “I've never ever in my lifetime seen a bug fix that only works 85% of the time.”
Anthropic implied coding is largely solved, then rolled back changes — OK (60/100)
At 6:50
The December 18th rollback directly undercuts any "largely solved" messaging. Timing is the receipt.
Why this score: Rollback timing contradicts marketing.
- December 18 statement admits stability issues heading into holidays
- Occurred months after flickering was known
- Shows internal reality diverging from public claims
Original quote: “hey, coding is largely solved. Turns out it's not largely solved.”
Terminal flickering bug lasted over a year from first GitHub report — Solid (75/100)
At 7:10
March 2026 minus first report = 12+ months on a grid layout bug. That's the core receipts.
Why this score: Timeline checks out with public record.
- First GitHub issue reported ~March 2025
- Still open March 25, 2026 per speaker
- Rendering rewrite failed right before Christmas 2025
- No hardware excuses apply — pure software
Original quote: “Over a year of the terminal flickering. Text characters laid out in a grid. Couldn't help but to flicker.”
If AI coding was solved, Anthropic wouldn't need feature branches for flickering — Opinion (50/100)
At 8:48
They're equating one UI bug with 'coding isn't solved' — that's a leap.
Why this score: Logical jump. Key issues:
- Flickering is a terminal rendering problem, not a measure of whether code generation works
- Companies ship feature flags for polish bugs even in mature products
- No evidence that 'solved' coding eliminates all UI work
Original quote: “If coding is solved, do you need feature branches? No. No. No. No. You don't. Somebody's lying to you.”
Anthropic and Boris are lying about coding being solved — Opinion (50/100)
At 8:58
Calling them liars over one open bug is strong — even if the hype is real.
Why this score: Subjective read. Key issues:
- Speaker admits they don't know if it's active deception
- Uses personal feeling as evidence
- The claim hinges on equating 'coding solved' marketing with zero remaining bugs
Original quote: “Because if coding was easy, if coding was solved, no, you don't need that. Who is telling the truth right now? Because it does not seem like Anthropic or Boris is leading you to the correct path. Now, I don't know if they're actually actively lying to you, but in my head, I feel like they are.”
Claude Dev's May 27th post admits the product still has mysterious error messages — OK (60/100)
At 9:22
The transcript shows them pointing at the post — but we can't see the actual tweet here.
Why this score: Unverifiable from transcript alone. Key issues:
- Speaker describes seeing 'fewer mysterious error messages' in the post
- No direct quote or screenshot provided in this segment
- Community comments also note ongoing quality issues, but that's opinion
Original quote: “New terminal dropping, right? This is the official Claude Dev account saying, 'Hey, we're doing it. We're dropping the big deal.' But even more so, if you scroll a little bit down, you'll see this fewer mysterious error messages.”
Claims coding is solved and the easy part — Opinion (50/100)
At 10:30
Boris says coding's solved while Anthropic still can't fix basic errors — rich.
Why this score: Subjective take on AI progress.
- Reflects the common industry view that core model training has stabilized
- Ignores ongoing issues with reliability and edge cases the speaker immediately points out
- More hot take than measurable claim
Original quote: “Coding is the easy part. Everything else is not yet solved, but is also becoming increasingly automated. saying to me, "Coding is solved. It's the easy part."”
Says Claude shows 98% uptime but still has elevated errors — Unverifiable (50/100)
At 10:45
No public uptime number cited — just vibes about '98 some percent.'
Why this score: Status claim without source.
- Speaker references 'clawed status' but provides no link or official dashboard
- Anthropic does not publish real-time uptime percentages publicly
- Community comments also note the lack of transparency on actual reliability
Original quote: “And yet during all of this, it just feels like they have an abundance of obvious failures. And if you just look at their clawed status, their claw status, it's like at 98 some percent up and they're having just errors elevated with this model”
Users report seeing other people's outputs in Claude for a year — Personal Story (50/100)
At 11:05
Community reports exist on GitHub/Reddit — speaker isn't making this up.
Why this score: Anecdotal pattern reported across platforms.
- Multiple users have posted about prompt leakage or cross-session contamination
- No official confirmation or denial from Anthropic in the transcript
- Consistent enough reports to treat as real user experience, not isolated hallucination
Original quote: “for the last year, you can see GitHub issues, Reddit posts, hacker news posts saying, okay, I was prompting and it looks like I got somebody else's feedback or I got somebody else's prompt result cuz I'm getting like a legal document”
AI salesmen pushing $10k/day spend while giving free tokens and making false claims — Opinion (50/100)
At 12:30
Calling out the exact contradiction: free unlimited tokens while charging ten grand a day.
Why this score: Opinion piece on AI industry hype. Key points:
- Speaker targets specific sales tactics around token pricing
- No external data presented to verify the 'false claims' allegation
- Frustration centers on perceived inconsistency between free access and high-cost pitches
Original quote: “everybody should be spending $10,000 a day on my company salesman selling his items and just telling everybody this is the way it should be when they get their tokens free and unlimited while at the exact same time making demonstrabably false claims”
See the full analysis with timestamps →