BERDAU EMPIREAI for creators

The beginner path

Start here. Four weeks,
four real things.

Not a curriculum. Four sessions, each producing something you can publish. If you only finish week one, you'll still have more than most people get from a month of tutorials.

Beginner 4 weeks 11 hours total No tools required to read

What to ignore for 30 days

Before the plan: the things that will eat your four weeks and produce nothing.

  • New model launches. The model you have is better than your brief. That ratio won't flip in a month.
  • Tool ranking videos. They're monetised by novelty, not by outcomes.
  • Agent frameworks. Not yet. You need a manual loop before you automate one.
  • Other people's dashboards. Revenue screenshots are marketing.
  • "10x" anything. The work is the same speed it always was. The thinking got cheaper.

Thirty days of ignoring that list is worth more than any tool subscription you'll buy this year.

Week 1 — One image that doesn't look AI-generated

Goal: one image you'd actually use as a cover. Not a gallery. One.

  1. Pick a subject you already know — your desk, your street, your hands, your kitchen at 7am. Specificity is the whole trick, and you can't fake knowing something.
  2. Write five adjectives you could defend to another person. "Moody" is not one. "Underexposed, sodium-lit, slightly out of focus" is.
  3. Collect three reference images. Not to copy — to point at when words fail.
  4. Generate four images, four times. Sixteen total. Change one variable per round so you learn which one is doing the work.
  5. Pick one. Spend five minutes on it: crop, contrast, grain. Texture is what separates a generated image from a photo.

Done looks like: you can say in one sentence why you picked that one.

How it usually fails: you generate sixty images and pick the prettiest. Prettiest is not yours — it's the model's default taste. The brief is what makes it yours.

Time: about three hours, most of it looking at references.

If you only have one hour

Do steps 1–3 and stop. A written brief with references is the reusable part; the generation is the cheap part. Come back to the rest when you have an evening.

Week 2 — One 30-second video, end to end

Goal: exported and posted. Quality is next week's problem.

  1. Write six lines: hook, three beats, a turn, a close. If it doesn't fit in six, it's two videos.
  2. Record the voice on your phone, in a small room, close to the mic. Do it three times and keep the third.
  3. Plan six shots, three to five seconds each, one idea per shot. Write them down before you generate anything.
  4. Cut to the voice, not to a beat. The audio is the spine.
  5. Export. Watch it once. Post it. Do not re-edit it for two days.

Done looks like: it exists in public and you didn't die.

How it usually fails: people try to make week 2 good. The point of week 2 is to prove you can finish. Finishing is a separate skill from taste, and you need it before taste is worth anything.

Time: about four hours.

Week 3 — One idea, four platforms

Goal: four posts from one idea, none of them a copy-paste.

Take the idea from week 2 and re-argue it four ways:

  • Writing — 150 words, line breaks, first line must survive truncation.
  • One post — under 240 characters. No thread. If it needs a thread, the idea isn't sharp enough yet.
  • Carousel — seven slides. Slide one works with zero context.
  • Email — 90 words, one ask.

The rule: no sentence may appear in two versions. If you can't write four different sentences, you don't have four posts — you have one post and three crops.

How it usually fails: repurposing becomes reformatting. A caption pasted onto a template is not native to the platform, and the audience can tell within two seconds.

Time: about two hours.

Week 4 — The brief you'll reuse every week

Goal: next week's content decided in twenty minutes.

Write six fields and keep them in one file:

WHO IT'S FOR   one specific person, not a demographic
THE ONE IDEA   a sentence someone could disagree with
THE LOOK       five adjectives + two reference files
THE FORMAT     the container (30s video / 7-slide carousel / 150-word post)
THE NO-LIST    three things this never does
SUCCESS TEST   how you'll know it worked, before you make it

Done looks like: you open the file, fill five lines, and start making instead of deciding.

How it usually fails: nobody writes the brief, so every week restarts from zero and every piece is a different person's taste.

Time: about two hours, and it pays back on week five.

Where to go after

If the visual side is what stuck, take content & visual production — it's this month, done properly. If the ideas are there but the output is inconsistent, take brand building.

If you want the prompts used in these four weeks, they're in the prompt index — filter by discipline, fill the variables, copy once.