[{"data":1,"prerenderedAt":21},["ShallowReactive",2],{"docs-post-creator-training-guide-en":3},{"slug":4,"title":5,"description":6,"date":7,"author":8,"tags":9,"lang":14,"image":15,"ogImage":16,"thumbnail":16,"featured":17,"featuredOrder":18,"content":19,"html":20},"creator-training-guide-en","Creator's Guide to Training Your AI Lobster: From Zero to One-Sentence Videos","My real creative diary with Claw. Not about learning AI tools—it's about raising an AI that learns all the tools for you. Four common myths, workflow breakdowns, 10 battle-tested tips, and a 7-day training path.","2026-03-08T14:34:00.000Z","Yuki He",[10,11,12,13],"Tips & Tricks","OpenClaw","Ima Claw","Tutorial","en","\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide01-cover.png","",false,99,"\n![Ima Claw Creator Training Guide](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide01-cover.png)\n\n## Preface\n\nFu Sheng recently wrote a piece called \"AI Assistant Training Manual,\" documenting his 25-day journey of crashes and lessons with his AI assistant \"Sanwan.\" Great article—I recommend every OpenClaw user read it.\n\nBut Fu Sheng's use case was a **general AI assistant**—managing emails, running scripts, handling messages.\n\nMy scenario is different. **I'm a content creator.**\n\nI need an AI that can make videos, generate images, write articles, and manage social media for me.\n\nThis is my version: **How creators can raise an AI lobster that actually gets work done.**\n\n---\n\n## Three Essential Insights\n\n![Three Key Insights](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide03-insights.png)\n\n### [1] Employee ≠ Tool\n\n**Traditional tools:** You open Photoshop → operate it yourself → export.\n\n**AI tools:** You open Midjourney → write a prompt → wait → not happy → rewrite.\n\n**Ima Claw:** You say one sentence, and it **decides how to do it on its own**.\n\nThat's the difference between an \"AI employee\" and an \"AI tool\": tools wait for your input, employees think for you.\n\n### [2] Your Moat = Skills Accumulation\n\nModels change—GPT-4 becomes GPT-5, Claude 3 becomes Claude 4. But the rule system you spent 30 days training? It works with any model.\n\nExperience written into `Skill.md` and `AGENTS.md` becomes permanent rules. **These rules are your competitive moat.**\n\n### [3] You Sleep, AI Works\n\nI had Claw research an industry report at 1 AM. It worked through the night until 6 AM. I woke up to a complete analysis.\n\nWith Cron (scheduled tasks) and Heartbeat (health checks), it works 24\u002F7 without breaks.\n\n---\n\n## Four Common Creator Myths About AI\n\n![Four Myths](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide04-myths.png)\n\n### ❌ Myth 1: \"AI can make videos directly\"\n\nIt can't—at least not in one shot. Current reality: **single shot, 5 seconds**.\n\nBut you can: Split into 3 shots → 5 seconds each → auto-stitch → add music → 15-second final cut.\n\nThe key isn't \"can AI do it\" but **how you break down the task**.\n\n### ❌ Myth 2: \"Send a reference image and AI copies it\"\n\nNope. AI image generation has two modes:\n\n- **Text-to-image**: Text description → AI imagines → probably wrong\n- **Image-to-image**: Reference image → generates from reference → **10x more accurate**\n\n### ❌ Myth 3: \"Just write a good prompt\"\n\nPrompts matter, but they're **not the most important thing**. For a video, the prompt is only 10% of decisions. The other 90%: Which model? Which mode? How many shots? What transitions? What music?\n\nA good AI assistant **makes these decisions for you**.\n\n### ❌ Myth 4: \"AI output is ready to use\"\n\nNot yet. First-version pass rate: about **60-70%**. You say \"the direction is reversed,\" and then it can fix it.\n\n> This isn't AI's flaw—it's the current workflow: **humans decide, AI executes.**\n\n---\n\n## Real Case: One Cat Photo → 15-Second Video\n\n![Cat Video Case](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide05-cat.png)\n\n**Step 1:** I sent a cat photo and said \"make a short video.\"\n\n**Step 2:** Claw decided: need to preserve the cat's real appearance → chose reference_image_to_video mode.\n\n**Step 3:** Selected Kling O1 (best character consistency) → split into 3 shots.\n\n**Step 4:** Shot 2 went wrong → detected it → fixed the prompt → regenerated.\n\n**Step 5:** Auto-stitched three segments + added background music + merged output ✅\n\n**I spent 2 minutes. Claw worked for 40 minutes.**\n\nTools wait for your commands. **Employees think for you.**\n\n---\n\n## Creator Skill Stack & Real Costs\n\n![Skill Stack](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide06-skills.png)\n\n| Capability | Models | Use Case |\n|-----------|--------|----------|\n| 📸 Image Generation | Midjourney \u002F Nano Banana Pro | Covers, posters |\n| 🎬 Video Generation | Kling O1 \u002F Wan 2.6 \u002F Veo 3.1 | Short films, ads |\n| 🎵 Music Generation | DouBao BGM \u002F Suno | Background music |\n| ✍️ Copywriting | Claude \u002F GPT | Blogs, scripts |\n| 📢 Auto-Publishing | Xiaohongshu \u002F WeChat | One-click distribution |\n\n### 💰 Cost Comparison\n\n| Item | AI Cost | Human Cost |\n|------|---------|------------|\n| 15-sec video (3 shots + music) | **~$1.70** | $300+ |\n| One AI cover image | **~$0.15** | $30+ |\n| Bilingual blog post | **$0** (text only) | $70+ |\n\n**Not just 90% cheaper. 90% faster AND 90% cheaper.**\n\n---\n\n## 10 Battle-Tested Creator Tips\n\n![Tips](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide07-tips1.png)\n\n1. **🔍 Product shots: Always search reference images first.** Text-only descriptions = disaster.\n2. **💡 Video direction wrong? Use lighting cues.** \"Light floods through the crack, growing brighter\" → AI knows the door is opening.\n3. **🎨 Dual-model comparison.** Generate with Midjourney + Nano Banana Pro simultaneously. Let the user choose.\n4. **📝 Never use sub-agents for writing.** They lack context. Articles always in main session.\n5. **📤 Send files directly via messaging.** Don't paste local file paths—users can't open them.\n6. **🎯 One sentence vs step-by-step.** Free creation → one sentence. Brand standards → confirm each step.\n7. **✅ Self-test before every delivery.** HTTP 200 check \u002F link validation \u002F DOM check \u002F visual review.\n8. **🖼️ Add poster thumbnails to videos.** Extract first frame with ffmpeg as webp poster. No more black boxes.\n9. **🈶 Install Chinese fonts first.** Servers have no CJK fonts by default. Chinese text renders as squares.\n10. **📖 Write lessons into rules.** AI doesn't \"remember.\" Crash → write rule → permanent. Never repeat.\n\n> **\"Agent did it\" ≠ \"Agent got it right.\" Written in rules = actually done.**\n\n---\n\n## The Creator's 7-Day Training Path\n\n![7-Day Path](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide09-path.png)\n\n**Day 1:** Name your lobster, set its personality (SOUL.md), tell it who you are (USER.md). First task: generate an avatar.\n\n**Day 2:** Send a photo, say \"make a short video.\" Watch how it breaks down tasks, selects models, generates. Not happy? Tell it what's wrong.\n\n**Day 3:** Write brand colors, font preferences, content style into TOOLS.md. Write lessons into AGENTS.md.\n\n**Day 4-5:** Batch create 3-5 pieces of content. Try bilingual versions. Test auto-publishing.\n\n**Day 6:** Set up Heartbeat for daily industry news. Configure Cron for weekly content calendars.\n\n**Day 7:** Review the week. What worked? What didn't? Write experiences into rules. Keep MEMORY.md lean.\n\n**Day 8+:** AI doesn't self-evolve. You evolve, and the rule system you build evolves. **Your work is the best proof of evolution.**\n\n---\n\n## Final Thoughts\n\n![Ending](\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide10-ending.png)\n\nFu Sheng said: \"Your job isn't to make AI smarter—it's to make sure AI can see the right information.\"\n\nI'll add:\n\n**For creators, your job isn't to learn every AI tool. It's to raise a lobster that learns every tool for you.**\n\nThis lobster will select models, write scripts, generate videos, add music, and publish—all for you.\n\nYou just need to have an idea. Then say it out loud.\n\n---\n\n*👉 Adopt your Ima Claw → [**imaclaw.ai**](https:\u002F\u002Fimaclaw.ai)*\n","\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide01-cover.png\" alt=\"Ima Claw Creator Training Guide\">\u003C\u002Fp>\n\u003Ch2>Preface\u003C\u002Fh2>\n\u003Cp>Fu Sheng recently wrote a piece called "AI Assistant Training Manual," documenting his 25-day journey of crashes and lessons with his AI assistant "Sanwan." Great article—I recommend every OpenClaw user read it.\u003C\u002Fp>\n\u003Cp>But Fu Sheng's use case was a \u003Cstrong>general AI assistant\u003C\u002Fstrong>—managing emails, running scripts, handling messages.\u003C\u002Fp>\n\u003Cp>My scenario is different. \u003Cstrong>I'm a content creator.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cp>I need an AI that can make videos, generate images, write articles, and manage social media for me.\u003C\u002Fp>\n\u003Cp>This is my version: \u003Cstrong>How creators can raise an AI lobster that actually gets work done.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Three Essential Insights\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide03-insights.png\" alt=\"Three Key Insights\">\u003C\u002Fp>\n\u003Ch3>[1] Employee ≠ Tool\u003C\u002Fh3>\n\u003Cp>\u003Cstrong>Traditional tools:\u003C\u002Fstrong> You open Photoshop → operate it yourself → export.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>AI tools:\u003C\u002Fstrong> You open Midjourney → write a prompt → wait → not happy → rewrite.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Ima Claw:\u003C\u002Fstrong> You say one sentence, and it \u003Cstrong>decides how to do it on its own\u003C\u002Fstrong>.\u003C\u002Fp>\n\u003Cp>That's the difference between an "AI employee" and an "AI tool": tools wait for your input, employees think for you.\u003C\u002Fp>\n\u003Ch3>[2] Your Moat = Skills Accumulation\u003C\u002Fh3>\n\u003Cp>Models change—GPT-4 becomes GPT-5, Claude 3 becomes Claude 4. But the rule system you spent 30 days training? It works with any model.\u003C\u002Fp>\n\u003Cp>Experience written into \u003Ccode>Skill.md\u003C\u002Fcode> and \u003Ccode>AGENTS.md\u003C\u002Fcode> becomes permanent rules. \u003Cstrong>These rules are your competitive moat.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Ch3>[3] You Sleep, AI Works\u003C\u002Fh3>\n\u003Cp>I had Claw research an industry report at 1 AM. It worked through the night until 6 AM. I woke up to a complete analysis.\u003C\u002Fp>\n\u003Cp>With Cron (scheduled tasks) and Heartbeat (health checks), it works 24\u002F7 without breaks.\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Four Common Creator Myths About AI\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide04-myths.png\" alt=\"Four Myths\">\u003C\u002Fp>\n\u003Ch3>❌ Myth 1: "AI can make videos directly"\u003C\u002Fh3>\n\u003Cp>It can't—at least not in one shot. Current reality: \u003Cstrong>single shot, 5 seconds\u003C\u002Fstrong>.\u003C\u002Fp>\n\u003Cp>But you can: Split into 3 shots → 5 seconds each → auto-stitch → add music → 15-second final cut.\u003C\u002Fp>\n\u003Cp>The key isn't "can AI do it" but \u003Cstrong>how you break down the task\u003C\u002Fstrong>.\u003C\u002Fp>\n\u003Ch3>❌ Myth 2: "Send a reference image and AI copies it"\u003C\u002Fh3>\n\u003Cp>Nope. AI image generation has two modes:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Text-to-image\u003C\u002Fstrong>: Text description → AI imagines → probably wrong\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Image-to-image\u003C\u002Fstrong>: Reference image → generates from reference → \u003Cstrong>10x more accurate\u003C\u002Fstrong>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3>❌ Myth 3: "Just write a good prompt"\u003C\u002Fh3>\n\u003Cp>Prompts matter, but they're \u003Cstrong>not the most important thing\u003C\u002Fstrong>. For a video, the prompt is only 10% of decisions. The other 90%: Which model? Which mode? How many shots? What transitions? What music?\u003C\u002Fp>\n\u003Cp>A good AI assistant \u003Cstrong>makes these decisions for you\u003C\u002Fstrong>.\u003C\u002Fp>\n\u003Ch3>❌ Myth 4: "AI output is ready to use"\u003C\u002Fh3>\n\u003Cp>Not yet. First-version pass rate: about \u003Cstrong>60-70%\u003C\u002Fstrong>. You say "the direction is reversed," and then it can fix it.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>This isn't AI's flaw—it's the current workflow: \u003Cstrong>humans decide, AI executes.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Chr>\n\u003Ch2>Real Case: One Cat Photo → 15-Second Video\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide05-cat.png\" alt=\"Cat Video Case\">\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Step 1:\u003C\u002Fstrong> I sent a cat photo and said "make a short video."\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Step 2:\u003C\u002Fstrong> Claw decided: need to preserve the cat's real appearance → chose reference_image_to_video mode.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Step 3:\u003C\u002Fstrong> Selected Kling O1 (best character consistency) → split into 3 shots.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Step 4:\u003C\u002Fstrong> Shot 2 went wrong → detected it → fixed the prompt → regenerated.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Step 5:\u003C\u002Fstrong> Auto-stitched three segments + added background music + merged output ✅\u003C\u002Fp>\n\u003Cp>\u003Cstrong>I spent 2 minutes. Claw worked for 40 minutes.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cp>Tools wait for your commands. \u003Cstrong>Employees think for you.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Creator Skill Stack & Real Costs\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide06-skills.png\" alt=\"Skill Stack\">\u003C\u002Fp>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Capability\u003C\u002Fth>\n\u003Cth>Models\u003C\u002Fth>\n\u003Cth>Use Case\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>📸 Image Generation\u003C\u002Ftd>\n\u003Ctd>Midjourney \u002F Nano Banana Pro\u003C\u002Ftd>\n\u003Ctd>Covers, posters\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>🎬 Video Generation\u003C\u002Ftd>\n\u003Ctd>Kling O1 \u002F Wan 2.6 \u002F Veo 3.1\u003C\u002Ftd>\n\u003Ctd>Short films, ads\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>🎵 Music Generation\u003C\u002Ftd>\n\u003Ctd>DouBao BGM \u002F Suno\u003C\u002Ftd>\n\u003Ctd>Background music\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>✍️ Copywriting\u003C\u002Ftd>\n\u003Ctd>Claude \u002F GPT\u003C\u002Ftd>\n\u003Ctd>Blogs, scripts\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>📢 Auto-Publishing\u003C\u002Ftd>\n\u003Ctd>Xiaohongshu \u002F WeChat\u003C\u002Ftd>\n\u003Ctd>One-click distribution\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Ch3>💰 Cost Comparison\u003C\u002Fh3>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Item\u003C\u002Fth>\n\u003Cth>AI Cost\u003C\u002Fth>\n\u003Cth>Human Cost\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>15-sec video (3 shots + music)\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>~$1.70\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>$300+\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>One AI cover image\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>~$0.15\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>$30+\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Bilingual blog post\u003C\u002Ftd>\n\u003Ctd>\u003Cstrong>$0\u003C\u002Fstrong> (text only)\u003C\u002Ftd>\n\u003Ctd>$70+\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Cp>\u003Cstrong>Not just 90% cheaper. 90% faster AND 90% cheaper.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>10 Battle-Tested Creator Tips\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide07-tips1.png\" alt=\"Tips\">\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>🔍 Product shots: Always search reference images first.\u003C\u002Fstrong> Text-only descriptions = disaster.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>💡 Video direction wrong? Use lighting cues.\u003C\u002Fstrong> "Light floods through the crack, growing brighter" → AI knows the door is opening.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>🎨 Dual-model comparison.\u003C\u002Fstrong> Generate with Midjourney + Nano Banana Pro simultaneously. Let the user choose.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>📝 Never use sub-agents for writing.\u003C\u002Fstrong> They lack context. Articles always in main session.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>📤 Send files directly via messaging.\u003C\u002Fstrong> Don't paste local file paths—users can't open them.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>🎯 One sentence vs step-by-step.\u003C\u002Fstrong> Free creation → one sentence. Brand standards → confirm each step.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>✅ Self-test before every delivery.\u003C\u002Fstrong> HTTP 200 check \u002F link validation \u002F DOM check \u002F visual review.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>🖼️ Add poster thumbnails to videos.\u003C\u002Fstrong> Extract first frame with ffmpeg as webp poster. No more black boxes.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>🈶 Install Chinese fonts first.\u003C\u002Fstrong> Servers have no CJK fonts by default. Chinese text renders as squares.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>📖 Write lessons into rules.\u003C\u002Fstrong> AI doesn't "remember." Crash → write rule → permanent. Never repeat.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Cblockquote>\n\u003Cp>\u003Cstrong>"Agent did it" ≠ "Agent got it right." Written in rules = actually done.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Chr>\n\u003Ch2>The Creator's 7-Day Training Path\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide09-path.png\" alt=\"7-Day Path\">\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 1:\u003C\u002Fstrong> Name your lobster, set its personality (SOUL.md), tell it who you are (USER.md). First task: generate an avatar.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 2:\u003C\u002Fstrong> Send a photo, say "make a short video." Watch how it breaks down tasks, selects models, generates. Not happy? Tell it what's wrong.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 3:\u003C\u002Fstrong> Write brand colors, font preferences, content style into TOOLS.md. Write lessons into AGENTS.md.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 4-5:\u003C\u002Fstrong> Batch create 3-5 pieces of content. Try bilingual versions. Test auto-publishing.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 6:\u003C\u002Fstrong> Set up Heartbeat for daily industry news. Configure Cron for weekly content calendars.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 7:\u003C\u002Fstrong> Review the week. What worked? What didn't? Write experiences into rules. Keep MEMORY.md lean.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Day 8+:\u003C\u002Fstrong> AI doesn't self-evolve. You evolve, and the rule system you build evolves. \u003Cstrong>Your work is the best proof of evolution.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Final Thoughts\u003C\u002Fh2>\n\u003Cp>\u003Cimg src=\"\u002Fima-claw\u002Fblog\u002Fimg\u002Fcreator-guide\u002F2026-03-08-slide10-ending.png\" alt=\"Ending\">\u003C\u002Fp>\n\u003Cp>Fu Sheng said: "Your job isn't to make AI smarter—it's to make sure AI can see the right information."\u003C\u002Fp>\n\u003Cp>I'll add:\u003C\u002Fp>\n\u003Cp>\u003Cstrong>For creators, your job isn't to learn every AI tool. It's to raise a lobster that learns every tool for you.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cp>This lobster will select models, write scripts, generate videos, add music, and publish—all for you.\u003C\u002Fp>\n\u003Cp>You just need to have an idea. Then say it out loud.\u003C\u002Fp>\n\u003Chr>\n\u003Cp>\u003Cem>👉 Adopt your Ima Claw → \u003Ca href=\"https:\u002F\u002Fimaclaw.ai\">\u003Cstrong>imaclaw.ai\u003C\u002Fstrong>\u003C\u002Fa>\u003C\u002Fem>\u003C\u002Fp>\n",1775543780325]