THIS IS ACTUALLY INSANE.
A 19-year-old just revealed how she’s using AI to build a YouTube automation business generating $20K/month.
No course.
No team.
No complicated setup.
Just her phone and one AI tool.
And she breaks it all down in just 69 seconds.
Bookmark this. You’ll want to come back to it.
THIS IS FCK**** INSANE
MOST PEOPLE THINK THESE AI MOTION GRAPHICS ARE JUST ONE PROMPT
They’re not, It’s a studio you set up once and people are already paying $2,000 $3,000 for the videos it produces.
→ Build the headless render engine with Claude Opus 5.5 (extra high): Node.js + FFmpeg + Playwright
→ Claude writes code that paints every single frame (it doesn’t generate video)
→ Playwright captures every frame in Chrome
→ FFmpeg stitches the frames into the final MP4
→ Feed it real asset references so it stops defaulting to centered all caps text on gradients
→ Ban basic CSS and force closed form spring physics so the motion actually looks good
→ Synthesize the audio in code and lock every cut to the beat
→ Make Claude watch its own frames, critique them, hunt dead frames, and re render until every scene scores 8/10
Set it up once.
Then start selling the output.
Bookmark this for later.
Most people think building a company alone means doing everything yourself
This One-Person Company OS with Grok Bot takes a different approach
One founder. 5–6 AI teammates. A system that runs while you're away.
→ Chief of Staff coordinates your day and builds the task queue
→ Intel tracks competitors, trends, prices and news
→ Content turns research into posts, emails and media
→ Sales manages leads, replies and follow-ups
→ Ops tracks orders, invoices and payments
→ Build handles small features and documentation
But the real play isn't just assigning AI different jobs
It's giving every bot its own schedule, browser and dashboard, while keeping approval gates for anything that sends, posts, pays or deletes.
The article quoted breaks down the daily workflow a 7-day setup, copy-paste prompts and worked examples.
Bots prepare and you press the button
That's how you build a company around AI without handing over control.
THANI retweeted
Holy f*ck…...... Claude Duet can now handle payments and log into your accounts without exposing your actual password.
Turn on the sound for this one. It walks you through every screen and explains what's happening.
Anthropic has laid out its entire security framework in one place.
Here's what you need to know in 5 screens:
Any sensitive action, like making a payment, deleting something or sending information, gets paused for your approval.
You see exactly what's being done, where it's going and what will be sent.
Another security layer operates independently of the AI, so hidden instructions on websites cannot manipulate its approval decisions.
Your password is entered directly into the app, encrypted, and only accessed by a separate browser environment that handles the login.
Bank card details and government identification numbers are blocked from being shared through messages, even when requested.
Transferring funds between your accounts or trading securities is completely restricted, even if you approve it.
Watch this before giving any AI agent access to your accounts.
THANI retweeted
Do you have USDC sitting in your wallet doing absolutely nothing? 😭
I do too.
Which is why I deposited $30 USDC into Solflare Earn.
Solflare Earn lets you supply USDC to a DeFi lending market powered by @Kamino, all without leaving your wallet.
Your USDC can earn interest based on the market's current rate.
And the process is straightforward:
1. Open Solflare.
2. Find Earn.
3. Review the live rate, estimated yearly earnings, and risks.
4. Choose how much USDC to supply and confirm the transaction.
There are no fixed lockup periods, and withdrawals are generally available anytime.
Explore Solflare Earn and tell me your experience.
THIS IS F**KING INSANE!
One script. One photo. A finished presenter video.
A free tool just dropped that turns a script and a photo into a ready-to-post short, and it's reportedly already past 26K GitHub stars. It's called "lanshu-create-ai-presenter-video."
What it does in a single run:
Takes your script and a presenter photo you have permission to use
Generates the voiceover
Animates the presenter and syncs the lips to the audio
Adds subtitles and creates the cover
Checks the sync and hands you the video with a quality report
The output: 9:16, 1080x1920, 30fps, 45 to 75 seconds. It all runs locally on your own computer.
If you make short-form content, save this before you film your next video.
More free tools coming soon, so follow along.
THANI retweeted
this is f*cking insane.
Andrej Karpathy spent 8 years at OpenAI and Tesla.
last week, he dropped a FREE 2-hour lecture covering everything he learned.
Agents → Harness → Loops → Graphs
some bootcamps charge $15K to teach you less than this.
no time to watch 2 hours right now?
bookmark it, so you don’t lose it
then read the guide below and build your first agent harness ↓
THANI retweeted
THESE 2 FREE GITHUB REPOS CAN HELP YOU BUILD YOUR OWN AI-POWERED TRADING TERMINAL.
REA + Kronos + Claude Code can turn an ordinary computer into a Bloomberg-style market research workspace.
Professional traders spend thousands on market data and research tools, but these open-source projects give you the building blocks to create your own.
Here’s how it works:
REA helps AI agents understand existing software and recreate its functionality.
Kronos is a financial foundation model that treats candlestick data like tokens and predicts possible future price movements.
Combine them, and you can build a financial terminal with charts displaying AI-generated forecasts.
Here’s how to build it yourself for FREE:
Step 1: Connect REA to Claude Code using:
github.com/morluto/rea
Ask Claude to map out a professional market terminal with quotes, charts, screeners, news, and market dashboards.
Step 2: Have Claude build the interface with equities, forex, commodities, crypto, global indices, and financial news.
Step 3: Connect public market data APIs to populate the terminal with live prices, charts, and market information.
Step 4: Clone Kronos:
github.com/shiyu-coder/Kr…
Load Kronos-small, a 24.7M parameter financial forecasting model.
Step 5: Feed it the last 512 candlesticks and generate predictions for the next 24.
Step 6: Generate 100 possible future price paths and display them directly on your terminal charts.
Step 7: Analyze the forecast distribution, identify market regimes, and backtest your strategies with realistic trading fees before risking real money.
The result?
A Bloomberg style research terminal combining market data, financial news, charts, and AI-powered forecasting in one workspace.
Just imagine what you can build with these tools.
BOOKMARK THIS. You’ll want to come back to it.
THIS IS F**KING INSANE
Making the cartoon was never the hard part. Take a proven kids format, feed the idea to AI, and a polished 15-second clip is ready in minutes.
The hard part starts after you upload.
Where it gets tricky:
Kids content is usually marked Made for Kids, which limits personalized ads
Same views, potentially lower RPM
You can't ignore the rules, because they apply either way
YouTube says repetitive, mass-produced, template-based, or easily replicated content can run into monetization problems
What's true:
AI can generate the cartoons. The videos can get views. Production costs almost nothing.
What breaks the model:
Copying a winning channel and pumping out near-identical AI videos at scale.
AI made production cheap. YouTube decides whether cheap production becomes income.
Full breakdown in the article below.
THANI retweeted
OCTOBER RECAP:
my first month running a solo agency.
got no team, just me and a network of top AI creators.
here's what i got done:
> pioneered 2 major AI model launches
> paid out $60,600 to 400+ creators
> made $4,300 from paid partnership posts
> hit 28k followers
crazy to think this was just an idea a month ago.
now brands book campaigns with me directly, and i handle everything from sourcing the creators to getting the campaigns delivered.
Just getting started.
THIS IS F**KING INSANE!
One creator reports $10K+ from AI-made videos without ever opening a timeline. The tools matter less than the system behind them.
The loop:
Spot a video that's already performing
Break down why it works: hook, pacing, structure
Have Claude write a fresh script built on that structure
Generate the visuals, voiceover, and captions
Publish
Repeat
The mindset shift:
"How can AI speed up my editing?" still leaves you doing the work, just a bit quicker.
"How much of this pipeline can I remove myself from?" is the question that builds a business.
Saving 30 minutes a video is optimization. A system that turns out video after video without you touching the edit is an asset.
Save this one.
THANI retweeted
a 9B model just beat qwen's own 27B on 7 out of 7 benchmarks
and it's uncensored.
the stock version refused 100 out of 100 test prompts. this one refuses 6.
> 256K context window
> native vision
> 185 tokens/sec on a single GPU
> apache 2.0
check it out : huggingface.co/DavidAU/Qwen3…
I broke down every other free model worth running in my article 👇
This is gold.
A guy from Better Stack just showed how FreeLLMAPI turns the free tiers you already have into something actually usable
→ Setup takes under a minute
→ Add free API keys from Gemini, Groq, Mistral, OpenRouter, Cloudflare and more
→ Point your regular OpenAI client at a local endpoint
→ Automatically routes to models with available quota
→ Fails over when one hits its daily limit
No more switching dashboards mid-task
The 4B number is the combined capacity of separate free tiers, not one unlimited pool
Keys stay local and encrypted
He also compared it with OpenRouter and LiteLLM, making it clear this is for maximizing personal free-tier access, not a production gateway
If you already have free API keys sitting unused, this is a practical way to put them to work.
THIS IS F**KING INSANE
This guy claims he made $80,528 with AI YouTube videos.
The workflow for these 2D animated videos is ridiculously simple.
Instead of hiring animators and paying $100 to $200 per video, he uses AI to generate the entire thing.
Here’s the workflow:
→ Pick a topic with strong curiosity potential
→ Give the AI a simple video idea
→ Choose a 2D animation style
→ Select the video length
→ Pick an AI voice
→ Create a custom character
→ Let the AI write the script
→ Generate the visuals
→ Add captions, music and sound effects
→ Edit any scene you don’t like
→ Add motion or turn certain images into video clips
→ Generate thumbnails
→ Export and upload to YouTube
The entire video can start from one simple prompt.
The AI handles the script, voiceover, characters, animation, visuals, sound effects, captions, and thumbnails.
You no longer need to know how to animate to enter the 2D explainer niche.
Your job is finding topics people want to watch, creating a strong hook, and giving the AI good direction.
One channel mentioned got 2 million views in 11 days with this style of animated content.
The real opportunity is finding a content format that’s already getting attention, then using AI to produce your own version faster.
Bookmark this.































