Make your first useful AI prompt
Turn a short note into an accurate action list you can verify.
patrickz.aiLet’s talk ↗THE PRACTICAL AI LIBRARY
57 hands-on guides from real experiments. Pick one outcome, follow the steps, and check what you made.
New to AI? Follow the beginner path ↗Just need a prompt? Browse the prompt library ↗
57 guides to explore
No guides match .
Try one of these: video · website · language · offline · prompt
Browse all 15 field notes · Start with the beginner path · Tell Patrick what you were looking for
Turn a short note into an accurate action list you can verify.
See why a fluent answer can still be wrong, then check an AI’s claims outside the chat.
Write a one-page problem brief with an explicit finish line.
Build a small comparison sheet based on work you can judge.
Run five controlled attempts at one small task, in one sitting or one a day, and keep evidence of improvement.
Finish a tiny creative artifact and identify what each tool contributed.
Write a failure review that produces one smaller, testable next attempt.
Choose the right workspace and complete a harmless first task.
Create a small context pack that makes a weekly planning answer specific and checkable.
Prepare a short bilingual message with an explicit confirmation step.
Produce an evidence-based weekly review and one action for next week.
Replace an aspiration with a small experiment and a review rule.
Improve one paragraph while retaining your facts, point of view and distinctive wording.
Build a short role-play that tests meaning and keeps a text fallback.
Create a short study loop that tests understanding against trusted material.
Turn a small record of project decisions into a tentative pattern and a testable rule.
Create a plan and shopping list while keeping observed items separate from guesses.
Make a simple diagram whose labels and relationships can be traced to your notes.
Plan and assemble a short sequence with one emotional or educational point.
Design a short demonstration with a visible before, change and result.
Create a three-shot production pack with reusable character references and review gates.
Create two images for different uses while preserving a consistent visual identity.
Create and test one complete local interaction before choosing a production stack.
Prepare a reproducible bug report and verify one focused fix.
Plan a small communication board with multiple ways to select and a visible cancel action.
Save a definition-of-done checklist in your project, then use it to review one small change with a meaningful test and a rollback point.
Define and test a small accessible tool with an ordinary-device fallback.
Build a small playable loop and use playtesting to control scope.
Design a game interaction that checks understanding and still works without a microphone.
Plan a site release with validation, a preview and a rollback checkpoint.
Design one safe AI command for an imaginary lamp: the AI proposes, a simple rule decides, and failure is honest.
Try one AI task on your phone the way people really use it (distracted, on a bad connection, interrupted) and write three recovery rules.
Compare a local model with a known-answer task and document its limits.
Map a weekly task, test a draft-only version on three cases, then run it on a schedule you control, with a manual fallback.
Design a short workshop in which every participant completes and checks one task.
Create a one-page site plan that helps a specific visitor take one action.
Produce a prioritized repair list based on the visitor’s main task.
Create an operating brief another person can use without reading your whole chat.
Create three consistent marketing drafts from one factual source brief.
Produce a short case study that separates your contribution, evidence and remaining limits.
Bigger builds with a boundary and a test. See them grouped by type ↗
Finish a one-room game with save, reload and a deliberate offline test.
Create a Today view with a persistent task and a useful evening review.
Make one playable language encounter with a text fallback.
Build a focused viewer that makes a visual bug reproducible.
Create a mobile-friendly event page with one verified booking route.
Save, export and restore a small set of local records.
Make movement, one attack and restart work across keyboard and touch.
Publish two readable posts with a dependable index and safe rendering.
Create a service page whose claims, scope and contact path are clear.
Normalize and compare listings without silently inventing missing facts.
Produce an implementation plan with a testable first milestone.
Create a file-based shot register with versions, references and review status.
Create a small library of reviewed prompts people can search, copy and check against a rubric.
Finish a small match with deterministic scoring and restart.
Design and test one create-and-list flow with a server-side user boundary.
Separate public room state from private player information.
Retire connectivity uncertainty before adding an AI feature.
What a chat assistant is really doing when it answers you, why it can sound certain and still be wrong, and a simple habit for trusting it the right amount.
Learn what an AI can actually see when it answers, and how to hand it a small, dated context pack that turns a generic reply into one that fits your life.
Many disappointing AI results answer the wrong question. Here's how to pin down the real problem first, so the AI's speed works for you.
Why making more with AI won't improve your results, and a simple practice loop that trains the judgment that will.
Why AI rankings can't choose your tools, and how to test them on a small bench of your own tasks.
Learn the create, review, revise loop: AI writes better instructions, checks its draft against your checklist and fixes one thing at a time, and you make the call.
Learn how to let AI find patterns in your own notes, calendar or chats, make it show its evidence, and turn one finding into a small experiment you can test.
A getting-started guide and a five-minute video for Microsoft 365 Copilot Cowork: have it find a repetitive process in your email, map the decisions, turn them into a shareable skill, and run that skill on a schedule, with guardrails that keep you in charge.
Learn to tell an AI coding agent what “done” looks like before it builds, then ship a small app in thin slices you can prove actually work.
Learn six steps for finding out why something broke (a game, a spreadsheet, a prompt) so the AI works from evidence, not guesses.
Space wizards. Expiring credits. A production workflow that became almost as interesting as the film.
One hundred and four posts, two years of ideas and zero websites. Here is the exact workflow, the prompts and the guardrails that turned them into patrickz.ai, plus the instructional video I made about it.
A request to train 20 new team members became a practical test of GPT-6: could the workflow I used for anime music videos make useful training material while I slept?
In one day, parallel AI agents rebuilt Super Mario Bros. and The Legend of Zelda starring me, upscaled and over the top, playable on my phone. The real lesson was how to direct agents and let a bot judge their work.
Learn to start an AI project from one person's real needs, pick the easiest input that works for them, keep a non-AI fallback, and measure what matters to them.
Based on an archive of 104 LinkedIn posts, four newer posts, and 18 project write-ups, with additional foundation lessons.
See the sources and coverage ↗