Videos

Skill Leap AI. Shows the specific pattern the article uses — letting a Custom GPT trigger Zapier actions to send emails, update sheets, or create calendar events from a chat. The setup details have shifted slightly since the video was made, but the model of "your AI assistant calls a no-code automation that does the actual work" is exactly right.

Tom Nassr | XRAY. Walks through what a Zap actually is, how triggers and actions fit together, and how to build a working Google Sheets → Slack automation from scratch — exactly the mental model you need before adding AI to the mix. Also tours Zapier Tables, Interfaces, and AI features so you have a sense of what's in the box.

IBM Technology. Sits below our usual 100K bar but earns the slot because it's the single best short explanation of why an employee using a personal ChatGPT account on work problems is the actual risk most companies face. Crume's "don't say no, say how" framing is the same posture the article takes — you're not trying to ban AI, you're trying to make safe use the easy default.

IBM Technology. Jeff Crume's lightboard explainer of the three places generative AI introduces risk — the data, the model, and the usage — and what good controls look like for each. Useful for the article's argument that "be careful" isn't enough; you need to think about which category of risk you're actually exposed to as an employee.

Tina Huang. Walks through the structured self-study system Huang used to learn SQL in 11 days for a Meta interview — active recall, spaced repetition, and using ChatGPT to generate practice problems and explain mistakes. Useful as the "how do I actually execute my plan" follow-up to the primary pick.

Tina Huang. Hands you a study-plan prompt template ("Act as a coding tutor that creates study plans…") that takes the student's goal, time commitment, and resource preferences, then generates a structured plan. The framing is coding, but the prompt pattern is exactly what the article applies to any 30-day learning sprint.

Ali Abdaal. Lays out the same kinds of patterns the article uses — challenge prompts ("give me the steelman counterargument," "give me a scathing critique"), explanatory theories, and the Solomon method (have the AI play your 90-year-old self) — and is unusually clear about where the AI stops being useful (mental health, accountability). Pairs well with the article's insistence that the value is in better questions and tradeoff-mapping, not "tell me what to do."

Tina Huang. The prompting chapter (02:30–09:20) lays out two stackable mnemonics — "tiny crabs ride enormous iguanas" (task, context, references, evaluate, iterate) and "ramen saves tragic idiots" (revisit, separate, try analogous, introduce constraints) — that map cleanly onto the iterate-and-refine patterns in the article. Useful when a single-shot formula isn't getting you there.

Jeff Su. Six-part formula — task, context, exemplars, persona, format, tone — explained with the same office-work examples the article uses (workout plans, resumes, internal emails). Watch this first; almost every pattern in the article is a focused application of one of these six components.

Jeff Su. The Claude Projects chapter (starts around 04:46) is the clearest short explanation on YouTube of project-level vs chat-level context — illustrated with a product marketing example where one project document feeds many chats with different briefs. Pairs naturally with the article's argument that Projects and Custom GPTs solve the same problem in slightly different shapes.

Skill Leap AI. Walks through the GPT builder exactly the way the article describes — name and description, conversation starters, instructions, knowledge files, capabilities, and access settings. Uses a real example (an AI knowledge assistant fed with the creator's own scripts) so you see what good knowledge-file content looks like in practice.

Kevin Stratvert. Calm, end-to-end walkthrough of the current Sora workflow: getting access, setting up a Cameo, prompting, choosing orientation, and using the desktop version. Useful as a "pick one tool and actually try it this week" follow-up to the comparison.

Versus. Pits Sora 2 Pro, Veo 3.1, Runway Gen 4, and Wan 2.5 against the same set of image-to-video and text-to-video prompts, then scores them on physics, lighting, motion, and prompt fidelity. It's the closest thing on YouTube to the article's "what's actually usable" framing — you watch the same idea handled four ways and start to see where each model breaks.

ElevenLabs. Official walkthrough that contrasts Instant Voice Cloning (a minute of audio, results in seconds) with Professional Voice Cloning (30 minutes to several hours of audio, much higher fidelity). The recording-quality guidance — mic, room, levels, pre-processing — is the part that's hardest to find elsewhere and matters most for getting a clone you'll actually use.

Alec Wilcock. Tour of the platform that anchors most of the article's examples — text-to-speech, speech-to-speech, voice design, and voice cloning, all in one screen-recorded walkthrough. Goes through the free-vs-paid limits and the controls that actually matter (stability, similarity, style) without overselling.

Future Tech Pilot. Once you've decided Midjourney is worth a try, this is the soup-to-nuts walkthrough — prompt structure, the stylize / chaos / weird parameters, image prompting, blending, zoom and pan, style tuning. The interface has continued to evolve, but everything here still applies and the prompting intuition transfers cleanly to v7.

Matt Wolfe. Matt runs the same prompt set through Midjourney, DALL·E, Adobe Firefly, Stable Diffusion, Leonardo, and others, and grades them on the dimensions the article cares about — accuracy, realism, illustration, logos, text, price. The model lineup has shifted since (Flux is now in the mix) but the methodology is what makes this still the best orientation video for the space.

Tina Huang. Tina puts Perplexity's Deep Research and ChatGPT/Claude analysis modes in the context of a working analyst's day — pulling together market signals, exploratory data analysis, and structured outputs. Useful as the "what else is this good for besides academic literature reviews" companion to the primary pick.

Andy Stapleton. Stapleton runs Deep Research against the exact use case it shines at — a literature review with a clear scope, time frame, and structure — and shows the clarifying-questions handshake, the 10-minute wait, the 40-source output, and where the citations and tone still need a human pass. Calls out the limits honestly (no BibTeX export, occasionally too generous in source choice).

Matt Wolfe. Once you understand the basics, this is the right second video. Matt runs through a long list of less obvious uses — turning a book into a study guide, briefing yourself on a competitor from their content, generating mind maps from a folder of PDFs — that help you see how flexible the "grounded notebook" frame actually is.

Paul J Lipsky. A clean three-step framing — curate sources, ask the right questions, produce final outputs — that maps almost one-to-one onto how the article tells you to think about NotebookLM. Recent enough (early 2026) to match the current UI, including Studio, audio overviews, and mind maps.

Leila Gharani. Older (early 2023) but still the cleanest worked example of pasting a real Excel question into ChatGPT, getting a SUMPRODUCT-or-conditional-formatting answer back, and then iterating with the model when the first formula doesn't quite fit. The model has improved a lot; the prompting workflow she shows is the same one you'll use today.

Leila Gharani. Live demo of the new `=COPILOT()` function on the kinds of messy real data the article describes — shift handover notes turned into a structured action table, free-text reviews tagged for sentiment, and inconsistent job titles normalized against a master list. Also calls out the limits clearly: don't use it for anything that has to be exact, freeze results as values when you're done.

David Perell. Perell teaches writing for a living and walks through how he actually drafts with LLMs without the output reading as AI — what he uses the model for (research, structuring, surfacing his own thinking), what he refuses to delegate (voice, taste, opinions), and the specific moves he uses to keep his prose his. Sits just under the 100K-view bar but earns its place because the topic is genuinely niche and this is the most thoughtful long-form take on it.