---
title: "How to Spot a Fake AI Builder: Does \"Show Me Where It Breaks\" Work?"
date: "2026-09-21"
video_id: "7688131254121729293"
tiktok_url: "https://www.tiktok.com/@fractionalctomatt/video/7688131254121729293"
creator: "@fractionalctomatt"
category: "AI"
tags: ["ai automation", "n8n", "zapier", "fmea", "hiring ai builders"]
excerpt: "Asking an AI builder where their system breaks is a fast version of failure mode analysis. It can be gamed, and Zapier and n8n stacks are not the red flag."
---
A TikTok from @fractionalctomatt says you can spot a fake AI builder with one question: "Show me where it can break." The question is a compressed version of failure modes and effects analysis, and it holds up better than the three red flags the video lists before it. It can still be gamed by anyone who has read a public list of AI risks, so the follow-up question matters more.

## What does "show me where it can break" actually test?

It tests whether the builder has a stored list of ways their system fails. That list only exists if they have run the system in production or deliberately analyzed it for failure before shipping.

Engineering has formal names for this:

- **Software FMEA.** NASA's Software Engineering Handbook defines it as a hardware FMEA with software components substituted in, and says the analysis "begins looking for potential system problems while the project is still in the design phase."
- **The premortem.** Gary Klein's 2007 Harvard Business Review article describes a session where "team members assume that the project they are planning has just failed" and list the likely reasons.
- **Chaos engineering.** The Principles of Chaos Engineering define it as "the discipline of experimenting on a system in order to build confidence in the system's capability to withstand turbulent conditions in production." Netflix's Chaos Monkey, an Apache-2.0 tool, randomly terminates production instances to put this into practice.
- **Postmortems.** Google's SRE book lists triggers for a postmortem, including "data loss of any kind" and "a monitoring failure (which usually implies manual incident discovery)."

The video's line that real builders "already know because they built it into the last one" matches the SRE view. Failure knowledge comes from having operated something.

## Why the question separates real builders from fake ones

Specific failure stories are hard to fake. Someone who has run a workflow against a live API can name what broke and what they added to stop it recurring: rate limit backoff, an upstream schema change, a token that expired overnight, a retry loop that charged a customer twice. Those details belong to one system, not a category.

The Google SRE trigger list shows what an operator's answer sounds like. Operators talk about thresholds, rollbacks and manual interventions. People who have only assembled a demo talk about capabilities.

## Can a fake AI builder pass the test?

Yes, with a little preparation. The vocabulary of AI failure is public. OWASP's Gen AI Security Project lists ten named LLM risks for 2025, from LLM01 Prompt Injection to LLM10 Unbounded Consumption. Reciting "prompt injection" and "hallucination" shows only that someone has read the list.

The video says the question works in five seconds. Five seconds shows whether someone hesitates, but judging the answer takes longer, and the video gives no way to score it. A stronger follow-up is to ask what they changed after the last failure and when it happened. Generic knowledge does not hold up when asked for a date.

The question also produces false negatives. A competent engineer who just inherited someone else's system may freeze for honest reasons, while a confident bluffer will not.

## Are Zapier and n8n stacks a red flag?

No. The video's first red flag is a stack of apps "wired together from a tutorial" that is "really just five tools talking to each other." Both major platforms ship the failure handling the video's own question looks for.

n8n documents:

- An error workflow, set per workflow in Workflow Settings, that "runs if an execution fails."
- An Error Trigger node, which an error workflow must start with.
- A Stop And Error node to "force executions to fail under your chosen circumstances."

Zapier documents:

- Autoreplay, an account-wide setting on Professional plans and higher that "works to retry any failed steps in a Zap run."
- Custom error handlers on Professional, Team and Enterprise. When you add one to a step, "the Zap will run the error handler as an alternative workflow." Every step except triggers and Paths can have one.

What to check is whether the builder turned these features on, which is what the "where does it break" question already checks. A hand-written Python service with no retries, no dead-letter queue and no alerting is more fragile than a five-node n8n workflow with an error workflow attached, even though it looks more serious.

## What the video leaves out

The clip names no tool, company, product, person or number, so viewers get a test without a way to grade it.

The second red flag, "the same setup they sold the last 10 guys and there's nothing built around it," mixes two complaints. Reusing a template across clients is normal for delivery firms. The real problem is the second half, that nothing was adapted to the buyer's business.

The video also shifts its claim. The description says the question reveals "whether someone actually knows how to build with AI," while the spoken hook says it reveals "whether someone knows how they're using AI or not." The question tests building much better than using.

Buyers who want a formal version can use the NIST AI Risk Management Framework, released January 26, 2023. It organizes the same idea into four functions (Govern, Map, Measure and Manage) and is voluntary and free.

## Bottom line

"Show me where it can break" is a sound screening question because failure knowledge comes from running a system. Follow it by asking what they changed after it broke, and when, since the risk lists are public. Judge Zapier or n8n builds by whether error handling is turned on, not by how many tools they connect.

## Sources

- [NASA Software Engineering Handbook 8.5, SW FMEA](https://swehb.nasa.gov/display/SWEHBVC/8.5+-+SW+Failure+Modes+and+Effects+Analysis): software failure modes and effects analysis.
- [Performing a Project Premortem, Harvard Business Review](https://hbr.org/2007/09/performing-a-project-premortem): Gary Klein's premortem technique.
- [Principles of Chaos Engineering](https://principlesofchaos.org/): the definition of chaos engineering.
- [Google SRE Book, Postmortem Culture](https://sre.google/sre-book/postmortem-culture/): blameless postmortems and their triggers.
- [Netflix Chaos Monkey](https://github.com/Netflix/chaosmonkey): a tool that terminates production instances on purpose.
- [OWASP Top 10 for LLM Applications](https://genai.owasp.org/llm-top-10/): the ten named 2025 LLM risks.
- [n8n: Handle errors gracefully](https://docs.n8n.io/build/flow-logic/handle-errors-gracefully): error workflows and related nodes.
- [Zapier: Error handling advanced settings](https://help.zapier.com/hc/en-us/articles/14167175792909-Decide-how-your-Zap-handles-errors-with-advanced-settings): Autoreplay.
- [Zapier: Set up custom error handling](https://help.zapier.com/hc/en-us/articles/22495436062605-Set-up-custom-error-handling): per-step error handlers.
- [NIST AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework): AI RMF 1.0.


## Video description and verification notes

**The one question in this video is a compressed version of failure modes and effects analysis, a design-phase method NASA has published guidance on since at least 2000, and it holds up better than the three red flags that precede it.** Fractional CTO Matt posted the 48-second clip on 2026-09-21. He speaks straight into a front-facing camera from a mesh office chair, with a Blue Snowball microphone pushed into the bottom of frame and neon question marks glowing on the wall behind him. A dark title card reading "How to spot a FAKE AI builder" sits over his head for the opening seconds, with "FAKE AI" highlighted in orange. Word-by-word captions run across his chest for the entire clip. Three numbered flag icons sit along the bottom and switch from black to glowing red as he counts off each red flag. Two stretches cut to an orange chroma-key background with an "AI" chip icon and person silhouettes, one cuts to stock footage of a man at a desk under the word "SYSTEM" in quotation marks, and the last third layers in mock iMessage bubbles plus stock portraits of a wide-eyed man in plaid and a smiling bearded man in a suit.

## The question, and what it actually measures

The ask is four words on screen and in the audio: "Show me where it can break." The video frames it as a five-second lie detector. What it really samples is whether the person has a stored inventory of failure modes for the thing they built, which is only possible if they have either run it in production or deliberately analyzed it for failure before shipping.

That inventory has formal names in engineering. NASA's Software Engineering Handbook defines a software FMEA as a hardware FMEA with software components substituted in, and notes the analysis "begins looking for potential system problems while the project is still in the design phase." Gary Klein's 2007 Harvard Business Review piece describes the premortem, where "team members assume that the project they are planning has just failed" and then generate plausible reasons why. The Principles of Chaos Engineering page defines the practice as "the discipline of experimenting on a system in order to build confidence in the system's capability to withstand turbulent conditions in production." Google's SRE book treats the same knowledge as a post-incident artifact and lists concrete postmortem triggers, including "data loss of any kind" and "a monitoring failure (which usually implies manual incident discovery)."

Matt's line that the real ones "already know because they built it into the last one" lines up with the SRE framing. The knowledge is a byproduct of having operated something, not a thing you produce on demand.

## Why the test has genuine discriminating power

The asymmetry is real. Somebody who has run a workflow against a live API for six months can name the specific thing that broke, when, and what they put in to stop it recurring. Rate limit backoff, a schema change upstream, a token that expired at 3am, a retry loop that double-charged a customer. That detail is expensive to fabricate because it is indexed to a specific system rather than to a category.

The Google SRE trigger list is a useful proxy for what a lived answer sounds like. Real operators describe thresholds, rollbacks, and manual interventions. People who have only assembled a demo describe capabilities.

## Where the test itself breaks

The video does not address the obvious counter. The vocabulary of AI failure is published, free, and short enough to memorize in an afternoon. OWASP's Gen AI Security Project lists exactly ten named LLM risks for 2025, starting with LLM01 Prompt Injection and ending with LLM10 Unbounded Consumption. Reciting "prompt injection" and "hallucination" when asked where a system breaks proves nothing except that someone has read a list. A fake with moderate preparation passes this question.

The five-second claim is also doing more work than it can carry. Five seconds is long enough to see whether someone hesitates. It is not long enough to evaluate the answer, and the answer is the part that matters. The video offers no rubric for telling a memorized answer from a lived one. My read is that the follow-up does the real work: ask what they changed after it broke, and ask for the timestamp. Generic knowledge collapses under a request for a date.

There is also a false-negative problem the video skips. A competent engineer who has just been handed a system someone else built will freeze on this question for honest reasons, and a confident bluffer will not.

## The first red flag argues against his own conclusion

Red flag one is a stack of apps "wired together from a tutorial" that is "really just five tools talking to each other." The premise is that glue-code automation stacks are inherently unserious. The platforms themselves contradict that.

n8n documents an error workflow you set per workflow in Workflow Settings that "runs if an execution fails," an Error Trigger node that the error workflow must start with, and a Stop And Error node to "force executions to fail under your chosen circumstances." Zapier documents Autoreplay, an account-wide setting on Professional plans and higher that "works to retry any failed steps in a Zap run," and a separate custom error handler feature on Professional, Team, and Enterprise where adding an error handler to a step means "the Zap will run the error handler as an alternative workflow." Each Zap step other than triggers and Paths can carry one.

So the number of tools in the stack is not the signal. Whether the builder turned any of that on is the signal, which is the same thing the video's actual question tests. The red flag is redundant with the punchline and weaker than it.

The reverse case holds too. A hand-rolled Python service with no retries, no dead-letter queue, and no alerting is more fragile than a five-node n8n workflow with an error workflow attached, and it looks more serious from the outside.

## What the video leaves out

No tool, company, product, person, or number appears anywhere in the clip, spoken or on screen. That is unusual for the genre and it makes the advice hard to act on. A viewer who runs this test gets a yes or no with no way to score the answer.

Red flag two, "the same setup they sold the last 10 guys and there's nothing built around it," conflates two separate things. Reusing a template across clients is how every delivery firm achieves margin and is normally a mark of competence. The actual complaint is in the second half, that nothing was adapted to the buyer's operation, and that deserved its own sentence.

The framing also shifts between the caption and the audio. The TikTok description says the question reveals "whether someone actually knows how to build with AI." The spoken hook says it reveals "whether someone knows how they're using AI or not." Building and using are different claims, and the question tests the first one much better than the second.

For a buyer who wants the formal version of this, the NIST AI Risk Management Framework, released January 26, 2023, organizes the same instinct into four functions: Govern, Map, Measure, and Manage. It is voluntary and free.

## Key Takeaways

- The question is "Show me where it can break," and it works because failure-mode knowledge is a byproduct of operating a system rather than something produced on demand. NASA's FMEA guidance, Klein's premortem, and the chaos engineering principles all formalize the same instinct.
- The test is gameable. OWASP publishes ten named LLM risks for free, so a prepared bluffer can recite plausible failure modes. Asking what they changed after the last failure, and when, is the harder follow-up the video never mentions.
- Red flag one contradicts the conclusion. n8n documents error workflows, the Error Trigger node, and Stop And Error; Zapier documents Autoreplay and per-step error handlers. A no-code stack with those enabled is more robust than custom code without them.
- Transcript quality: Whisper's output matches the burned-in captions closely across all 48 seconds and does not cut off, ending at 47.92 seconds. It captures none of the on-screen material, which is roughly half the video. The title card, the three-flag counter, the mock iMessage exchange where "How it works?" is answered with a blurred link and "Show me where it can break" is answered with "aaaaaaaa," and the stock reaction portraits are all invisible in the text.
- Unverified: the "five seconds" figure and the "last 10 guys" detail are rhetorical and have no source. The URL in the mock text bubble is deliberately blurred and unreadable. The account's follower count and bio could not be confirmed because TikTok profile pages render client-side and returned no data. The display name "Fractional CTO Matt" and all engagement counts come from the downloaded metadata rather than a live page load, and I did not independently establish who operates the account.

## Resources

- [How to spot a fake AI builder](https://www.tiktok.com/@fractionalctomatt/video/7688131254121729293) - the source video, 48 seconds, 295 views, 8 likes, 1 comment, 0 reposts at capture
- [NASA Software Engineering Handbook 8.5, SW Failure Modes and Effects Analysis](https://swehb.nasa.gov/display/SWEHBVC/8.5+-+SW+Failure+Modes+and+Effects+Analysis) - defines software FMEA and states the analysis begins during the design phase, the formal analogue of the video's question
- [Performing a Project Premortem, Gary Klein, Harvard Business Review, September 2007](https://hbr.org/2007/09/performing-a-project-premortem) - the premortem technique, where a team assumes the project has already failed and lists the reasons
- [Principles of Chaos Engineering](https://principlesofchaos.org/) - the definition of chaos engineering and its five principles, including minimizing blast radius
- [Google SRE Book, Postmortem Culture](https://sre.google/sre-book/postmortem-culture/) - blameless postmortems and the concrete trigger list that shows what an operator's failure vocabulary sounds like
- [Netflix Chaos Monkey](https://github.com/Netflix/chaosmonkey) - Apache-2.0 tool that randomly terminates production instances, integrated with Spinnaker, a working example of building failure into the last one
- [OWASP Top 10 for LLM Applications](https://genai.owasp.org/llm-top-10/) - the ten named 2025 risks from LLM01 Prompt Injection to LLM10 Unbounded Consumption, free and public, which is why reciting them proves little
- [n8n, Handle errors gracefully](https://docs.n8n.io/build/flow-logic/handle-errors-gracefully) - error workflows, the Error Trigger node, and the Stop And Error node
- [Zapier, Decide how your Zap handles errors with advanced settings](https://help.zapier.com/hc/en-us/articles/14167175792909-Decide-how-your-Zap-handles-errors-with-advanced-settings) - Autoreplay retries failed steps, available account-wide on Professional plans and higher
- [Zapier, Set up custom error handling](https://help.zapier.com/hc/en-us/articles/22495436062605-Set-up-custom-error-handling) - per-step error handlers that run an alternative workflow on failure, on Professional, Team, and Enterprise
- [NIST AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework) - AI RMF 1.0, released January 26, 2023, with the Govern, Map, Measure, and Manage functions
