trippy

news aggregator

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

$4,780/month. Built by one person with Claude, Obsidian, and n8n.

No team. No expensive SaaS stack. Just a system that gets smarter every day. Every Claude conversation is saved to Obsidian, every useful prompt becomes a reusable asset, and nothing valuable is ever lost.

Instead of starting from scratch, n8n automatically tests workflows, archives the best outputs, and pushes proven prompts into production. The system improves itself while you focus on building.

How it works:

Write once.

Save everything.

Automate the repetitive work.

Let your knowledge compound instead of disappear.

Progression:

Day 1: Empty vault.

Week 1: 300+ prompts organized.

Month 1: First automated AI workflow.

Month 3: Multiple AI assets generating revenue.

Month 5: Crosses $4,780/month.

Most people use AI to get answers.

The people making money use AI to build systems that remember, improve, and scale without them. That’s what a real second brain looks like.

[Issa @issatheblows]

https://t.co/dxnjAqNRKi

[trippy @trippybrippy]

A guy from Indonesia built an AI workflow that now brings in over $7,745/month.

No agency. No expensive team. Just Claude, Obsidian, and a library of AI tools.

Obsidian became his second brain. Every prompt, workflow, content idea, and experiment is stored in one searchable vault instead of getting lost in chats.

Claude writes scripts, improves prompts, and plans content. AI tools generate the assets. Obsidian keeps everything organized so every project becomes faster than the last.

The system:

Week 1: Organized every idea.

Month 1: Built a repeatable AI workflow.

Month 3: Published consistently.

Month 6: Crossed $7,745/month.

The people making the most money with AI aren’t using more tools.

They’re building systems that compound.

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

Terence Tao — one of the greatest living mathematicians — recently said: nobody truly understands why modern AI works so well… and exactly where it breaks.

He shows how models can produce perfectly confident answers that are mathematically wrong. And most people can’t even spot the flaw.

This isn’t a bug. It’s a fundamental limitation.

We taught machines to imitate patterns, but we haven’t given them real understanding of structure and rigorous proof.

Right now, AI is still a “very convincing bullshitter” in math and science.
The real breakthrough will come when models don’t just predict — they derive and prove new theorems with guaranteed correctness.

Tao is a living reminder: mathematics isn’t just a tool. It’s a way of thinking that today’s LLMs still don’t have.

Who’s following Tao’s work? Drop your
favorite lectures or papers below

[the lich @thelichhh]

https://t.co/aqsY18NvVz

[trippy @trippybrippy]

A guy from Poland built a system for AI models that now brings in over $5,200/month.

No agency. No expensive team. Just Claude, Obsidian, and a stack of local storage.

Obsidian became his second brain. Every prompt, workflow, and winning idea lives in one searchable vault.

Claude writes scripts, captions, and refines prompts. Local drives archive every AI model, dataset, and asset without relying on the cloud.

The system

Week 1: Organized knowledge.
Month 1: Automated content pipeline.
Month 3: First AI models live.
Month 5: Consistent $5k/month.

The people making the most with AI aren’t using more tools.

They’re building better systems.

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

A guy is making $4,500+ every month with AI girls.

His entire setup? Claude, Obsidian, UGC workflows, and a laptop.

No camera crew. No expensive studio. No real model.

He uses Claude to generate character concepts, scenes, and prompts. Every prompt, reference, and workflow is organized in Obsidian, so every new AI girl stays visually consistent.

Then he turns those images into realistic UGC-style videos that look like they were filmed on an iPhone in a real apartment.

People don’t realize they’re watching AI.

They comment, follow, share, and ask where they can find her.

The result?

More than $4,500/month from short-form content powered by AI.

The biggest advantage isn’t the AI itself.

It’s having a repeatable system that lets you create faster than everyone else.

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

How one AI workflow generated $12,940.

Most people use AI to save time.

One creator used it to build a system.

Claude writes.
Obsidian stores every winning prompt.
n8n automates the workflow.

Videos created: 37
Actors hired: 0
Hours filming: 0

Every new project gets faster because nothing starts from scratch.

The biggest advantage in AI isn’t the model you use.

It’s the system you build around it.

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[trippy @trippybrippy]

A guy created this girl with Claude — and is already making more than $3,500 a month with her.
Real apartment. Real lighting. Real atmosphere.

But the girl herself? Never existed.
He didn’t shoot for hours. Didn’t hire a model.

He took a real photo of his apartment, generated the girl in Claude, then brought her to life with video tools. One consistent seed so her face stays the same.

She dances, smiles, looks at the camera — everything feels completely real.
People in the comments write “she’s so beautiful”, ask where it was filmed, and spam hearts.

He just uploads short clips and watches the engagement (and money) roll in.
No studio. No permissions. No awkward “can I get your number”. Just a laptop and smart prompts.

This isn’t the future. This is right now.
One guy, one AI girl, a real apartment in the background — and the algorithm does the rest.

[Kardinall @kardinall]

https://t.co/96ym5qiHOW

[trippy @trippybrippy]

A builder loaded an ACASIS 8-bay beast with mixed drives and built a massive local second brain that powers his AI and video projects to real money.

Zero cloud dependency. Full ownership.

Obsidian runs as the knowledge core — everything is linked in one giant graph. n8n automations constantly ingest Claude outputs, test agents, archive patterns and trigger local workflows — all running 24/7 on raw local storage.

The hardware edge
Eight hot-swap bays accept any mix of HDDs and SSDs. Tool-less trays, rock-solid build, one power cable and it runs almost completely silent.

The entire vault sits right next to his Mac Mini/Studio. Every experiment, dataset, revenue log and idea stays instantly searchable in one beautiful Obsidian graph. n8n workflows pull data, process it and push results without ever leaving the desk.

The results
Week 1: dozens of drives online, hundreds of terabytes ready.
Month 1: Automated pipelines and RAID fully live.
Month 3: Multiple AI streams and local agents generating income.

Old drives get new life. Hot-swap design protects the brain. The feather test proves it — dead silent even under heavy load.

Local-first storage turned scattered ideas into a self-running system.

The second brain doesn't just remember. It compounds.

If this was useful — follow for more local
AI setups.

[trippy @trippybrippy]

https://t.co/SzufV6IbtO

[boozie @soboozie]

99% of Claude Code users build an entire agent team for a job 1 sub agent finishes in 2 minutes.

That's the mistake burning your context window right now.

SUB AGENTS are contractors. Hand over 1 job. They disappear into their own context window, finish it, come back with the result. Nothing left cluttering your main orchestrator.

Their value is parallelism. 3 independent tasks, 3 sub agents, running at once instead of one after another. Research a library, refactor a module, write tests. None of them need to know the others exist.

AGENT TEAMS are employees. They stay on the payroll for the whole project, sometimes for weeks, building context on top of context every day.

A real team has structure: 1 lead agent, 2 or 3 working agents, a shared task state everyone reads. When one finishes a piece, the others see it and adjust. Sub agents can't do that. No cross-talk, no shared state, no real-time reaction.

1 rule decides which setup you need.

Work is isolated: SUB AGENT. Work requires collaboration: AGENT TEAM.
A team for a 2-minute job is just overhead: more coordination, more state to manage, more places for agents to step on each other.

Save the team for jobs that need a lead making calls and workers reacting live. Everything else is a contractor's job.

Contractors don't need a desk. Employees do.

[trippy @trippybrippy]

A guy earns over 5K$ a month on AI videos with hot girls — while sitting at home.
It’s real in 2026. Here’s a short step-by-step guide how to do it.

1. Create your AI girl
Choose a style and niche. Generate a consistent character (same face and body) using modern AI tools for photos and videos.

2. Build content library
Create 20-30 photos + 5-10 short videos (5-20 seconds).
Split into teasers for attraction and hot PPV content for sales.

3. Launch your profile
Set up an account on an adult content platform.
Disclose that it’s AI-generated. Set subscription at $10-20 + individual videos at $5-15.

4. Drive traffic
Post teasers on social media with link in bio.
Post 4-7 times a week and reply to DMs.

5. Monetize and scale
Offer custom videos on request.
Consistency and engagement = growth to 5K+ per month.
The key is quality and consistency. AI does the heavy lifting, you control the process.

[Bober_smart @Bober_smart]

https://t.co/p4teS18nnJ

[trippy @trippybrippy]

A 25-YEAR-OLD GUY CREATED AI MODELS AND STARTED POSTING THEM ON ONLY FANS AND EARNING BIG MONEY FROM IT

25-year-old guy from Japan quits his office job after discovering AI models for OnlyFans. In his very first month he made over $11,000.

He started learning AI out of curiosity, went all in, and built everything with neural networks. No face on camera, no crew, just consistent content from generated models.

Now he runs it as a scalable business while traveling.

Sometimes one new skill completely changes your life.
Who else is building with AI? Share your story below.

#AI #SideHustle

[Issa @issatheblows]

https://t.co/eGvzuk22SJ

[trippy @trippybrippy]

How to make over $3000 a month with AI model videos — without showing your face or saying a single word.

Simple process:

1. Find a viral girl’s video on TikTok
2. Download it
3. Take a screenshot of the first second
4. Throw it into ChatGPT to get a detailed JSON prompt
5. Generate your own AI girl in the same look and style
6. Upload the original video + your AI image into Kling AI
7. Get an almost identical video with your model

Post it on Instagram Reels — and the views and money start coming in.
No editing, no face on camera, no voice. Just AI.

[BLAZT @BLAZT_AI]

https://t.co/Qj7J9qCS6R

[trippy @trippybrippy]

Within a few years, half of commercial ads will be created without a camera. Most people won’t even notice.

I made a 10-second cinematic ad that looks like a luxury perfume campaign: dramatic lighting, atmospheric haze, glossy textures, premium aesthetics. The entire production was built with AI.

My workflow is simple:
→ Claude develops the concept, writes the script, and generates production-ready prompts.
→ AI image and video models turn those prompts into high-end cinematic visuals.
→ Every successful prompt, reference, style, and workflow goes into Obsidian.

That’s the real advantage.

Obsidian becomes your personal Prompt Factory—a system where every successful visual style is documented, reusable, and continuously improved. Instead of starting from scratch, you open a proven pipeline and recreate premium-quality content on demand.

Then you package it as a business.

Don’t sell “AI.”
Sell the outcome:
“Luxury-quality brand commercials in days—without expensive production crews, cameras, or studios.”

Companies don’t care how the video was made. They care whether it captures attention, elevates their brand, and drives sales.

The people building AI production systems today won’t just compete with freelancers tomorrow—they’ll compete with creative agencies.

If you’d like me to break down the exact workflow—from the first prompt to landing paying clients—send me a DM or follow along.

[HodlReaper @HodlReaper]

https://t.co/Ib16Z0xAt3

[trippy @trippybrippy]

A GUY IS RIDING A SKATEBOARD. NO, THIS IS NOT A NEW ROCK STAR GAME. IT'S VERY SIMPLE.

Everyone asks how creators make AI videos that look like GTA 6.

It’s not one AI. It’s a stack.

• ChatGPT or Claude → write the cinematic script and prompts.
• Midjourney or Flux → generate characters and environments.
• Veo 3, Kling or Runway → turn images into realistic cinematic videos.
• ElevenLabs → create natural voiceovers.
• n8n + Obsidian → automate your workflow, save prompts, and build a reusable content library.
• CapCut or DaVinci Resolve → add sound design, motion graphics, color grading, and the final polish.

The creators getting millions of views don’t have better prompts.

They have better systems.

AI is only 20%.

The workflow is the other 80%.

[Andy🤖 @andy_neon_]

https://t.co/tDSgXqxEUk

[trippy @trippybrippy]

I stopped using ChatGPT like a chatbot.
I built a system where every conversation, idea, meeting, and piece of research becomes part of a growing knowledge base. Instead of disappearing into chat history, everything is searchable, connected, and reusable.

ChatGPT helps me think. Obsidian remembers everything. n8n moves information automatically between tools. YouTube summaries, AI outputs, notes, documents, and project updates all end up in one place without copy-paste.

After a few months, the biggest change isn’t speed—it’s leverage. Old ideas resurface when they’re relevant, successful prompts become reusable templates, and every project makes the next one easier to build.

AI isn’t the advantage. A system that compounds your knowledge is. Most people start every project from scratch. I build on everything I’ve already learned.

[Moysei @0xMoysei]

https://t.co/YzCkLZEykT

[trippy @trippybrippy]

QUIZ: IS THIS REAL? NO , THIS 15-YEAR-OLD GUY MAKES $5,000 EVERY MONTH

Most creators still open CapCut and hope inspiration shows up.

That works when you’re making one video.

It breaks when you’re making twenty.

Instead of relying on inspiration, build a system.

CapCut edits the videos.

Claude writes the scripts.

Obsidian remembers everything you’ve ever learned.

Every winning hook.

Every client brief.

Every successful script.

Every idea worth reusing.

Nothing gets lost.

Before Claude writes a single word, it reads your knowledge base.

It knows what worked before.

It knows your style.

It knows your clients.

Every new script starts with experience instead of a blank page.

Then n8n takes over.

A new brief arrives.

It collects the context, sends it to Claude, and saves everything back to Obsidian automatically.

By the time you open CapCut, the hardest part is already done.

Month 1: Build the knowledge base and connect the tools.

Month 2: Every project takes less time because the system remembers what you don’t.

Month 3: You’re no longer creating every video from scratch.

You’re improving a machine that gets smarter with every project.

Creators chase inspiration.

Systems compound knowledge.

Build the vault once.

Let Obsidian remember.

Let Claude think.

Let CapCut create.

Repeat.

[HodlReaper @HodlReaper]

https://t.co/Ib16Z0xAt3

[trippy @trippybrippy]

$18,741 a month. And no, I don’t film girls, buy expensive cameras, or organize photoshoots.

I simply take photos of girls from public sources, turn them into realistic AI-generated videos using tools like Runway, Kling AI, Hailuo AI, Luma Dream Machine, and Pika, and then publish the finished content on Instagram, TikTok, and OnlyFans.

Today, this strategy generates $18,741 every month, and judging by the growth, it’s only the beginning.

[Kardinall @kardinall]

https://t.co/96ym5qiHOW

[Kardinall @kardinall]

https://t.co/96ym5qiHOW