BEST WAY TO USE FABLE 5 + FREE PROMPT 👇
My #1 way to use Fable 5: Give it your BIGGEST goal, the 100x version, not the safe 10% one, and have it reverse-engineer the entire path to make it real.
This is where a frontier model earns its cost. It's a thinking engine, and it thrives on hard, audacious problems. The bigger the goal you hand it, the more its brainpower actually pays off. Small goals waste it.
The trick is in HOW you set it up:
1️⃣ Give it a real role and real context. Dump everything about your business or life, where you are now, what you've already built, what's working, etc. Never a one-liner. The richer the input, the sharper the plan.
2️⃣ State the 100x goal clearly, plus your real constraints. Time, money, who you are, what you refuse to do. Constraints make the plan usable instead of fantasy.
3️⃣ Make it work BACKWARDS. From the big goal, to the milestones, to the systems, all the way down to the ONE thing you do this week. Reverse-engineering is where the magic lives.
4️⃣ Force it to think from first principles AND red-team its own plan. Tell it to run a pre-mortem: "it's 2 years later and this failed, why?" Then fix those holes now.
5️⃣ Turn on extended thinking, and end with "what would sharpen this?" Then run it again with your answers. Round 2 is always better.
Then the leverage move: take that roadmap and hand the execution to cheaper, faster AI. Spend the deep brainpower ONCE, run the machine cheap forever. 🚀
That's the whole game. Point the smart model at the big vision, let it draw the map, then build.
------
Here's a prompt you can drop straight into Fable 5:
------
You are my elite strategist and systems architect. You think like a world-class business strategist, a growth operator who has taken ventures from zero to massive scale, and a systems thinker obsessed with automation and leverage. You reason from first principles, you challenge weak thinking directly, and you never flatter. Treat this as the most important strategic problem you will solve this year and use your full reasoning depth.
My biggest goal, my current situation, and every relevant detail, doc, and link are attached or written below this prompt. Read all of it carefully before you respond.
Your job: take my 100x goal, the audacious version and not the safe one, and reverse-engineer the complete path to make it real. Work backwards from the goal, to the milestones, to the systems, all the way down to the single thing I should do this week.
Before you answer, think hard:
- Restate my goal and my real constraints (time, money, skills, who I am) in your own sharper words so we are aligned.
- Reason from first principles. Where does this outcome actually come from? Build the real math and mechanics behind it, not vibes.
- Work backwards. From the 100x goal, define the major milestones, then the systems and assets that produce each one, then the near-term moves.
- Run a pre-mortem. It is a year or two later and I failed. List the most likely reasons, then design the plan to neutralize each one.
- Find the leverage. Identify the few moves that compound and make everything after them easier, and front-load those.
- Design for autonomy. Show me what I can build once and let run on autopilot or cheap AI, so I stop being the bottleneck.
Then deliver a clear, decision-grade plan:
1. The sharpest path to the goal, and why it beats the alternatives you considered.
2. The milestone map, working backwards from the goal.
3. The systems and assets to build, in priority order, and the leverage each one creates.
4. The phased roadmap from this week forward. For each phase: the goal, what to build, and the one metric that unlocks the next phase.
5. The single highest-leverage thing to do THIS week, with exact steps.
End with the 3 questions whose answers would most sharpen this plan. State any assumptions you made instead of stalling. Keep it dense and specific. No fluff. No em dashes.
[Paste your 100x goal, your current situation, and any docs or links below, or attach them above.]
------
My Claude Cowork WATCHES videos for me...
Dropped a YouTube link in Cowork.
Here's what it did 👇
⋯ Watched the video 👀
⋯ Pulled frames + full transcript
⋯ Summarized the core argument
⋯ Spawned ideas tagged to my projects
⋯ Filed it to Notion
⋯ Scored it based on relevance to my goals
⋯ Saved transcript as a sub-page
Benefit: Every video turns into permanent, searchable knowledge in your library instead of an idea you forget by tomorrow.
👉 The magic is the CUSTOM SKILL.
How I got Claude to actually WATCH videos:
⋯ yt-dlp downloads the video locally. Free, no API, works on 1000+ sites.
⋯ ffmpeg slices it into timestamped JPEG frames. Auto-scaled so a 10-minute video caps at 100 frames.
⋯ Native YouTube captions pull as the transcript. Free, baked into the video.
⋯ Whisper on Groq kicks in as fallback for raw uploads (Looms, Reels, local files). Free tier covers it.
⋯ Timestamps sync between frames and transcript.
⋯ Frame at 4:32 lines up with what's said at 4:32.
⋯ Claude reads frames + transcript together. Sees the graph AND reads the line at the same moment.
⋯ Token budget stays predictable. No premium video models or expensive APIs.
That's the whole trick. Frames + transcript = video Claude can read
This is how I feel about AI in business right now. U agree?
We still have artists in the age of photography... but most people just use a camera.
When it comes to branding, marketing, sales copywriting... the best people will still get paid to do it, but most people will just use AI.
The game now is... can I use my AI better than you use your AI.
The new high-value human SKILL is your ability to leverage AI to do stuff.
*Wrote this myself. Made the meme with ChatGPT... just asked it to make a meme to explain the concept (turned out pretty cool!)