Writing Better, Faster, More Freely
Writing is the creative application most people reach for first — and with good reason. Whether you write professionally, occasionally, or rarely, AI can help you do it better, faster, and with significantly less friction. The key is using AI in a way that amplifies your voice rather than replacing it.
Overcoming the blank page
Writer's block is almost always a starting problem, not a continuation problem. Once you have a sentence on the page, the next one follows. AI is extraordinarily good at breaking the starting impasse — not by writing the piece for you, but by giving you enough momentum to take over.
I'm trying to write [describe piece — essay, blog post, speech, report]. My main argument or point is [your core idea]. I'm stuck on how to start. Give me five different opening sentences or paragraphs — each approaching the opening from a completely different angle. I'll choose one and continue from there in my own voice.
Five options is the key. One option creates a take-it-or-leave-it binary. Five options almost always contains something usable, and comparing them helps you understand what you're actually trying to say.
From seed to outline in minutes
The most valuable writing use of AI is not writing the first draft — it's developing the structure before you write a word. A clear outline is the difference between a piece that writes itself and one you fight with for hours.
I have a rough idea for a [article/essay/report]: [describe your idea in a sentence or two]. The audience is [who will read this]. The publication or context is [where it's going]. The length will be approximately [word count].
Build a detailed outline that:
— Identifies the single most interesting angle on this topic for this audience
— Structures the argument so each section flows logically into the next
— Tells me, for each section, what it needs to do (not just what it covers)
— Includes a note on the tone and what makes this piece worth reading for this particular audience
The draft-and-edit workflow
The single most effective AI writing workflow is: let AI write the first draft, then make it yours through editing. This feels counterintuitive to many writers — but it's transformative once you try it.
The reason it works is that editing is cognitively easier than drafting. Looking at an existing piece and improving it draws on different (and in many ways more natural) mental resources than generating from nothing. You also edit more honestly than you draft — you'll cut things from AI's draft that you'd never cut from your own, which often makes the final piece stronger.
The secret to using AI writing help without losing your voice: edit aggressively. Replace any phrase that doesn't sound like you. Cut the things AI adds that you wouldn't. Add the things you know that AI doesn't. The first draft is scaffolding — the final piece should be entirely yours.
Editing and improving what you've written
Sometimes you don't want AI to draft — you want it to improve what you've already written. Here are the most useful editing prompts:
[Paste your text]
Edit this for clarity, flow, and impact — but preserve my voice exactly. Do not introduce new ideas or phrases I wouldn't use. Tell me: what you changed and why, what you left because it was already strong, and what you'd cut entirely if the piece needed to be 20% shorter.
[Paste your text]
Read this as a discerning reader who will stop reading the moment it gets boring. Mark the specific sentence or paragraph where you'd stop and explain exactly why — what went flat, what lost momentum, what felt like it was going through the motions. Then suggest how to fix just that moment.
Use AI to break the starting impasse by generating five different opening options. Develop structure before drafting — an outline built with AI often produces the best pieces. The draft-and-edit workflow is more efficient than drafting alone. Edit AI output aggressively to preserve your voice. The "where does the piece lose energy" prompt is one of the most useful editing tools available.
Generating and Developing Ideas
AI is the most powerful idea generation tool that has ever existed — but most people use it wrong. They ask for ideas and take the first thing that appears. This lesson covers the diverge-converge method: generating broadly without filtering, then selecting ruthlessly.
Why your first AI ideas are usually your worst
When you ask AI for ideas, it produces the most statistically likely suggestions — the ones that appear most frequently in its training data. These are the obvious, generic, already-done ideas. They are a starting point, not a destination.
The skill is using that generic starting point to get to something genuinely original. That requires diverging further before you converge on anything.
The diverge phase
In the diverge phase, your goal is quantity and variety — not quality. Do not evaluate ideas as they appear. Generate at least 20-30 before you consider any of them seriously.
I am working on [project/challenge]. Generate 25 completely different directions I could take this — include obvious approaches, unusual ones, approaches from adjacent fields, approaches that might seem wrong at first, and approaches that combine things in unexpected ways. Do not evaluate any of them. Just generate.
The converge phase
After generating broadly, you converge: read through everything, mark the ideas that create genuine energy (not just "that seems fine"), and identify which ones could be combined or developed into something stronger. You bring the judgement; AI brought the volume.
Stress-testing and lateral thinking
Once you have a promising direction, use AI to stress-test it. Ask what could go wrong, what you're assuming, and what someone who disagrees with you would say. Then use lateral thinking prompts to push even further from the obvious.
I am developing this idea: [your idea]. First, give me the three strongest arguments against it. Second, tell me what this idea assumes that might not be true. Third, give me five completely different ways to achieve the same underlying goal — approaches that have nothing in common with my current direction.
AI's first ideas are the most generic — use them as a starting point, not a destination. The diverge-converge method: generate 20+ ideas without filtering, then select ruthlessly. Stress-testing good ideas is as important as generating them.
AI for Side Projects and Personal Goals
A side project or personal goal that you've been thinking about for months — maybe years — can now become something you actually start, because the friction of the beginning has dropped dramatically. AI can help you validate the idea, identify the skill gaps, plan the work, learn what you need to learn, and produce the first concrete output — all in a single session. This lesson covers how.
Moving from idea to action
Most side projects die between the idea and the first concrete step. The gap is usually one of three things: not knowing if the idea is worth pursuing, not knowing where to start, or not having the skills the project seems to require. AI addresses all three.
I have an idea for a side project: [describe it]. I've been thinking about it for [how long] but haven't started. Help me move from idea to action by doing three things:
1. Validation check: what would tell me this is worth pursuing? What's the fastest way to find out if there's real demand or genuine value here? What should I look for and what would make me abandon this idea?
2. Skill gap analysis: what skills does this project actually require? Which do I probably already have? Which would I need to develop or find another way around?
3. First 30 days: what are the ten most important things to do in the first 30 days, in the right order? What's the single best thing I could do this week to make this more real?
Addressing skill gaps with AI as a teacher
One of the most common reasons side projects stall is encountering a skill gap — you need to know how to do something you've never done. Writing copy. Building a simple website. Understanding the legal basics. Managing money. AI can teach you the fundamentals of virtually any skill quickly enough to get started, and connect you with the right resources to go deeper.
For my side project [describe], I need to [specific skill or task] and I've never done it before. Give me:
1. The 20% of knowledge that covers 80% of what I need for this specific use case
2. The three most common mistakes beginners make doing this
3. The fastest path to being good enough — not expert, just competent for my purposes
4. One concrete thing I can do in the next two hours to start building this skill
AI for personal goal setting
AI is a surprisingly effective thinking partner for personal goals — not because it can tell you what your goals should be, but because it can help you clarify what you actually want, identify what's getting in the way, and design a realistic plan for getting there.
I want to [describe your goal] by [timeframe]. Help me think this through clearly by asking me three to five probing questions that would help you understand whether this goal is genuinely what I want, whether it's realistic given my circumstances, and whether my plan for getting there is sound. Ask them one at a time and wait for my response before moving on.
The single most useful question to ask AI about any side project: "What is the minimum viable version of this that I could complete this weekend — something real enough to learn from?" Getting to a real thing, however small, is worth more than months of planning.
The three killers of side projects — uncertainty about value, unclear starting point, skill gaps — are all addressable with AI. The kickstarter prompt covers all three in one session. Use AI to bridge skill gaps: ask for the 20% that covers 80%. For personal goals, AI works best as a probing question-asker rather than an advice-giver. The minimum viable start is always more valuable than a longer planning period.
Creative Constraints and AI
There is a paradox at the heart of creativity: more freedom often produces less interesting work. Constraints force creative choices. They eliminate the infinite and make your decisions meaningful. AI, which has essentially unlimited freedom by default, produces its most interesting creative work when you give it tight constraints — and the same is true of your collaboration with it.
The constraint paradox
Ask AI to "write something creative" and you'll get something generic — the average of what "creative writing" looks like in its training data. Ask AI to "write a 200-word poem about grief that doesn't mention death, loss, or sadness, and uses only images from the natural world" and you'll get something that surprises you.
The constraint is not a limitation — it is a creative direction. It eliminates the default choices and forces AI into territory it wouldn't otherwise explore. This is true whether you're writing, generating ideas, designing something, or making any other creative decision.
Write [creative piece] with the following constraints:
— [Constraint 1: something to include, e.g. "must include exactly one question"]
— [Constraint 2: something to exclude, e.g. "cannot use any abstract nouns"]
— [Constraint 3: a structural constraint, e.g. "each paragraph must be shorter than the last"]
— [Constraint 4: a tonal constraint, e.g. "must feel hopeful without using optimistic language"]
The constraint is the brief, not a suggestion. If you can't meet all four, tell me which you'd relax first and why.
Using style as a constraint
One of the most interesting creative uses of AI is asking it to adopt a specific style — not as a mimicry exercise, but as a constraint that forces distinctive choices. This works particularly well for written content where you want to avoid the generic AI voice.
Write [piece] in the style of [specific writer, era, or publication — e.g. "early Orwell essays", "The Economist's leaders section", "a Victorian naturalist's field notes"]. Don't mimic their subject matter — adopt their rhythms, their relationship with the reader, their level of formality, and their approach to evidence and argument. Then tell me the three most distinctive stylistic choices you made.
Remixing and combining
Another powerful constraint technique: combining two things that don't obviously belong together. The creative tension between incompatible elements forces genuine novelty — something neither element would produce alone.
Combine these two seemingly incompatible things: [Thing A] and [Thing B]. Find the genuine creative connection between them — not a forced metaphor, but a real insight about what they share. Then produce [creative output — a short piece, a concept, a design brief] that emerges from that connection.
Unconstrained AI produces generic output. Constraints produce distinctive work. Four-constraint creative briefs — include, exclude, structural, tonal — reliably produce more interesting output. Style constraints are particularly powerful for escaping the default AI voice. Remixing incompatible elements forces genuine novelty. The constraint is the creative direction, not a limitation.
Part 3 Exercise: Create Something Real
This is your Part 3 exercise lesson. You have covered writing with AI, idea generation, AI for side projects, and creative constraints. Now you produce something real.
Choose a personal creative project — something you have been meaning to start, or something you are already working on. Then:
1. Use the diverge-converge method to generate 20 directions for this project.
2. Select the most promising three and develop brief outlines for each.
3. Pick one and draft the first real piece of content — an opening paragraph, a first scene, a first section, whatever fits your project type.
4. Apply at least two creative constraints to the draft and revise.
5. Write a 150-word description of the project you could share with someone else.
The output should be something you are genuinely proud of — not perfect, but real. AI generated the volume; your taste and judgement shaped it into something yours.
Mark this lesson complete to unlock your Part 3 completion. Save your work — the draft, the constraints you applied, and the 150-word description. You will reference these in your Part 6 capstone.