Insomnia's banner

Insomnia

@insomnia_vip • 4,673 subscribers

26 / ai dev / building in public / ex @AIatMeta / dm open

Shorts

A CLUB JUST ASKED MY FLY TO DJ THEIR ENTIRE NIGHT Yesterday I got a message from a club called Neon Hive with a completely insane proposal: they wanted my FLY to run their next party from start to finish and choose everything by itself. They wanted no human controlling the set. So I gave them my code, connected the system to their DJ setup and let the fly take over the booth for the night. From that moment, every decision was hers. Fly picked what played next, controlled the decks and kept the party moving while hundreds of people danced, sang and had absolutely no idea who was actually running the music. The weirdest part is watching the footage afterward. You see a packed nightclub going completely crazy while this tiny fly is sitting behind the decks with its headphones on, casually running the entire night like this is what it was born to do. Nobody in the crowd knows their DJ is an insect. And somehow, the party just keeps going. Neon Hive might have been the first club brave enough to hire a FLY

A CLUB JUST ASKED MY FLY TO DJ THEIR ENTIRE NIGHT Yesterday I got a message from a club called Neon Hive with a completely insane proposal: they wanted my FLY to run their next party from start to finish and choose everything by itself. They wanted no human controlling the set. So I gave them my code, connected the system to their DJ setup and let the fly take over the booth for the night. From that moment, every decision was hers. Fly picked what played next, controlled the decks and kept the party moving while hundreds of people danced, sang and had absolutely no idea who was actually running the music. The weirdest part is watching the footage afterward. You see a packed nightclub going completely crazy while this tiny fly is sitting behind the decks with its headphones on, casually running the entire night like this is what it was born to do. Nobody in the crowd knows their DJ is an insect. And somehow, the party just keeps going. Neon Hive might have been the first club brave enough to hire a FLY

114,059 views

I BUILT A FLY A FULL BREWERY AND IT STARTED MAKING BEER It starts with a fly dragging a pallet jack loaded with seven bags of milled malt into an industrial brewing system. Then it gets to work. It opens the boiler valves, starts the transfer pump, moves mash between three massive tanks and climbs the metal stairs carrying a 33 kg bag of Pilsner malt like this is just another day at the brewery. The entire process is being controlled step by step. The fly decides when to open the steam, when to transfer the mash, when to start cooling and when to move on to the next stage. Every action appears in the log while a neural network with 393 nodes and 472 connections keeps track of the decisions happening behind the scenes. Then comes the part I really didn't expect to see. The fly adds its own hops. FLYHOPS goes into the kettle, the wort starts moving through the system and the little brewer goes back to work. A few moments later, it opens the cooling water and finally hovers over the last tank to add FLYMENTIS yeast. > Malt > Hops > Yeast Three giant tanks. One very determined fly. I think I accidentally gave an insect everything it needs to start its own brewery

I BUILT A FLY A FULL BREWERY AND IT STARTED MAKING BEER It starts with a fly dragging a pallet jack loaded with seven bags of milled malt into an industrial brewing system. Then it gets to work. It opens the boiler valves, starts the transfer pump, moves mash between three massive tanks and climbs the metal stairs carrying a 33 kg bag of Pilsner malt like this is just another day at the brewery. The entire process is being controlled step by step. The fly decides when to open the steam, when to transfer the mash, when to start cooling and when to move on to the next stage. Every action appears in the log while a neural network with 393 nodes and 472 connections keeps track of the decisions happening behind the scenes. Then comes the part I really didn't expect to see. The fly adds its own hops. FLYHOPS goes into the kettle, the wort starts moving through the system and the little brewer goes back to work. A few moments later, it opens the cooling water and finally hovers over the last tank to add FLYMENTIS yeast. > Malt > Hops > Yeast Three giant tanks. One very determined fly. I think I accidentally gave an insect everything it needs to start its own brewery

43,518 views

A 24-year-old generated an AI girl on his laptop and built a page pulling millions of views every month No camera No real person No photoshoots Just AI-generated models, fast edits and captions engineered to keep people watching One clip took 11 minutes to make It crossed 8,700,000 views in 5 days ChatGPT wrote the scripts Flux generated the model CapCut edited everything automatically Now the page runs 24/7 without him even touching the account Most people still think AI tools are “just for fun” Meanwhile anonymous creators are building entire audiences from people who don’t even exist

A 24-year-old generated an AI girl on his laptop and built a page pulling millions of views every month No camera No real person No photoshoots Just AI-generated models, fast edits and captions engineered to keep people watching One clip took 11 minutes to make It crossed 8,700,000 views in 5 days ChatGPT wrote the scripts Flux generated the model CapCut edited everything automatically Now the page runs 24/7 without him even touching the account Most people still think AI tools are “just for fun” Meanwhile anonymous creators are building entire audiences from people who don’t even exist

112,141 views

A FLY IS NOW WRITING PYTHON CODE WITH ITS OWN NERVOUS SYSTEM I genuinely didn't expect a fly to end up behind a code editor. The setup is strangely simple. A fly sits on a keyboard while its neural activity is captured in real time. On the other side of the screen, Python code starts appearing line by line inside wordle_solver_v1.py, with the neural visualization changing as the input comes through. The model contains 100,010 flyemg neurons, with 88,172 represented in the current neural model. As activity changes, the system translates those signals into keyboard input and the cursor keeps moving through the code. Then you start looking at what is actually being typed. import random import unicodedata def quitar_acentos(palabra): It's not just random characters filling the editor. The video shows an actual Python function being constructed while the neural activity and keyboard input remain synchronized beside it. Brain activity → keyboard → code. A tiny biological nervous system is being used as the interface between a fly and a programming environment. The fly isn't sitting there watching someone code. It's sitting on the keyboard while the system turns its neural activity into the next input. And somehow the first thing it ended up building was a Wordle solver

A FLY IS NOW WRITING PYTHON CODE WITH ITS OWN NERVOUS SYSTEM I genuinely didn't expect a fly to end up behind a code editor. The setup is strangely simple. A fly sits on a keyboard while its neural activity is captured in real time. On the other side of the screen, Python code starts appearing line by line inside wordle_solver_v1.py, with the neural visualization changing as the input comes through. The model contains 100,010 flyemg neurons, with 88,172 represented in the current neural model. As activity changes, the system translates those signals into keyboard input and the cursor keeps moving through the code. Then you start looking at what is actually being typed. import random import unicodedata def quitar_acentos(palabra): It's not just random characters filling the editor. The video shows an actual Python function being constructed while the neural activity and keyboard input remain synchronized beside it. Brain activity → keyboard → code. A tiny biological nervous system is being used as the interface between a fly and a programming environment. The fly isn't sitting there watching someone code. It's sitting on the keyboard while the system turns its neural activity into the next input. And somehow the first thing it ended up building was a Wordle solver

12,054 views

A $20 APP CAN ORGANIZE WHAT MOST PEOPLE NEVER DO One creator got tired of scattered notes, forgotten ideas and unfinished projects, so he built his own operating system on top of Obsidian Powered by Claude, it transformed 378 notes, 1,816 connections and 53 knowledge clusters into a living dashboard that looked more like a sci-fi command center than a note-taking app He could see how ideas connected, which projects were active, what routines supported them and where every new piece of information belonged. The system evolved every day, becoming more valuable with every note he added instead of more chaotic In reality, spending $20 a month on the right AI tools and organizing your knowledge properly can be worth far more than buying another course you'll never finish

A $20 APP CAN ORGANIZE WHAT MOST PEOPLE NEVER DO One creator got tired of scattered notes, forgotten ideas and unfinished projects, so he built his own operating system on top of Obsidian Powered by Claude, it transformed 378 notes, 1,816 connections and 53 knowledge clusters into a living dashboard that looked more like a sci-fi command center than a note-taking app He could see how ideas connected, which projects were active, what routines supported them and where every new piece of information belonged. The system evolved every day, becoming more valuable with every note he added instead of more chaotic In reality, spending $20 a month on the right AI tools and organizing your knowledge properly can be worth far more than buying another course you'll never finish

11,660 views

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

A FLY JUST BECAME A DJ AND IT ACTUALLY LOOKS LIKE IT KNOWS WHAT ITS DOING There is something completely ridiculous about seeing a tiny fly sitting behind a professional DJ setup with headphones on like it has a residency at some underground club. Its legs move across the turntables while the mixer sits directly in front of it, and the whole performance is being captured from multiple angles. It doesn't look like an insect accidentally landed on the equipment. It looks like someone gave the smallest creature imaginable an actual DJ booth. But underneath the performance, something much stranger is happening. The system is tracking the fly's neural activity in real time. A glowing neural structure pulses beside a live spike raster as electrical activity changes inside its nervous system, turning the tiny movements of the insect into something that can be watched alongside the performance. The fly keeps moving, the neural signals keep firing and the decks keep responding. For a moment, it genuinely looks like we're watching an insect discover a completely new career. Forget human DJs. We might have just created the world's smallest one

A FLY JUST BECAME A DJ AND IT ACTUALLY LOOKS LIKE IT KNOWS WHAT ITS DOING There is something completely ridiculous about seeing a tiny fly sitting behind a professional DJ setup with headphones on like it has a residency at some underground club. Its legs move across the turntables while the mixer sits directly in front of it, and the whole performance is being captured from multiple angles. It doesn't look like an insect accidentally landed on the equipment. It looks like someone gave the smallest creature imaginable an actual DJ booth. But underneath the performance, something much stranger is happening. The system is tracking the fly's neural activity in real time. A glowing neural structure pulses beside a live spike raster as electrical activity changes inside its nervous system, turning the tiny movements of the insect into something that can be watched alongside the performance. The fly keeps moving, the neural signals keep firing and the decks keep responding. For a moment, it genuinely looks like we're watching an insect discover a completely new career. Forget human DJs. We might have just created the world's smallest one

265,800 views • 25 days ago

I BUILT A ROBINHOOD CHAIN TERMINAL THAT TELLS ME WHAT MY BAG CAN ACTUALLY EXIT FOR BEFORE I TRUST THE NUMBER ON THE SCREEN Meet HOP OUT Give it a Pons V2 token and an amount or public wallet: → reads Robinhood RPC, Pons API and DexScreener → finds the position → calculates four exit estimates → applies pool depth, fees and a visible haircut → shows estimated ETH/USD returned and retained percentage → prints a timestamped receipt in text, JSON or Markdown A bag can look huge on a chart and still have a very small exit. HOP OUT doesn’t take the spot price, multiply it by your balance and call that liquidity. It estimates what the position could actually exit for through the published pool depth, then shows the assumptions beside the result. The whole loop is: TOKEN / WALLET → PUBLIC SOURCES → FOUR EXITS → RECEIPT There’s also a browser terminal, local CLI, offline demo, provider doctor and read-only API. No private key. No signer. No transaction path. No buy button. No automatic sell. Just public data, explicit assumptions and a receipt you can keep. BIG BAG. SMALL DOOR. HOP OUT Open source repo in the first reply ↓

I BUILT A ROBINHOOD CHAIN TERMINAL THAT TELLS ME WHAT MY BAG CAN ACTUALLY EXIT FOR BEFORE I TRUST THE NUMBER ON THE SCREEN Meet HOP OUT Give it a Pons V2 token and an amount or public wallet: → reads Robinhood RPC, Pons API and DexScreener → finds the position → calculates four exit estimates → applies pool depth, fees and a visible haircut → shows estimated ETH/USD returned and retained percentage → prints a timestamped receipt in text, JSON or Markdown A bag can look huge on a chart and still have a very small exit. HOP OUT doesn’t take the spot price, multiply it by your balance and call that liquidity. It estimates what the position could actually exit for through the published pool depth, then shows the assumptions beside the result. The whole loop is: TOKEN / WALLET → PUBLIC SOURCES → FOUR EXITS → RECEIPT There’s also a browser terminal, local CLI, offline demo, provider doctor and read-only API. No private key. No signer. No transaction path. No buy button. No automatic sell. Just public data, explicit assumptions and a receipt you can keep. BIG BAG. SMALL DOOR. HOP OUT Open source repo in the first reply ↓

65,011 views • 1 month ago

MY ROBIN GROK MADE $9,041.64 IN LESS THAN A WEEK I gave Robin Grok a simple objective: take a small trading bankroll and keep finding opportunities in the Robinhood meme-coin market without me sitting in front of the screen all day. It started with just $1,047.28. From there, Grok began scanning price action, liquidity, volume and wallet activity across the market, constantly looking for where capital was moving next. The system tracks everything from market pressure to whale activity and even flags abnormal volume migration across connected venues. The interesting part is that Grok isn't just looking for coins that are going up. It watches BUY → ROUTE → EXIT signals, measures the pressure behind the move, and decides when the position is no longer worth holding. One trade can be opened while another is already being routed out, keeping the capital moving instead of letting it sit idle. The dashboard is now showing: $1,047.28 → $9,041.64 That's roughly an 8.6x increase in less than a week. And the numbers behind the system are running live: 72% buy pressure, 94.8% AI confidence, 248/250 nodes online, 99.98% data integrity, with the signal engine continuously processing market activity. The part I like most is that I don't have to tell it what to trade every five minutes. Robin Grok watches the market, detects the signal, routes the capital, updates the position and waits for the next opportunity

MY ROBIN GROK MADE $9,041.64 IN LESS THAN A WEEK I gave Robin Grok a simple objective: take a small trading bankroll and keep finding opportunities in the Robinhood meme-coin market without me sitting in front of the screen all day. It started with just $1,047.28. From there, Grok began scanning price action, liquidity, volume and wallet activity across the market, constantly looking for where capital was moving next. The system tracks everything from market pressure to whale activity and even flags abnormal volume migration across connected venues. The interesting part is that Grok isn't just looking for coins that are going up. It watches BUY → ROUTE → EXIT signals, measures the pressure behind the move, and decides when the position is no longer worth holding. One trade can be opened while another is already being routed out, keeping the capital moving instead of letting it sit idle. The dashboard is now showing: $1,047.28 → $9,041.64 That's roughly an 8.6x increase in less than a week. And the numbers behind the system are running live: 72% buy pressure, 94.8% AI confidence, 248/250 nodes online, 99.98% data integrity, with the signal engine continuously processing market activity. The part I like most is that I don't have to tell it what to trade every five minutes. Robin Grok watches the market, detects the signal, routes the capital, updates the position and waits for the next opportunity

17,561 views • 1 month ago

How girls in Miami make thousands of dollars without having a job No office No 9-to-5 She built multiple income streams straight from her phone The foundation is TikTok Live and Snapchat monetization But here’s where it gets crazy Agencies pay her a flat fee just for access to her Instagram DMs She is literally renting out her profile But the wildest passive income is random cash drops Guys send her $100 on Venmo just for posting a basic selfie She makes hundreds every month purely from stranger tips While you grind 40 hours a week, someone is out here monetizing their mere existence

53,133 views • 4 months ago

NVIDIA's new $1,999 box isn't interesting because of the hardware It's interesting because of what it does to trust Most people look at a machine like DGX Spark and immediately focus on the specs > More memory > More compute > Bigger models > Wrong angle The real value shows up during the sales call When you're pitching private AI, local inference or on-prem deployments, the hardest part is usually convincing buyers that your setup is legitimate That's why third-party validation matters The moment serious infrastructure reviewers start testing the hardware, the conversation changes Now you're no longer asking a client to trust your opinion You're pointing to a platform that has already been examined, benchmarked and discussed by people their IT team already follows And that's the part most builders are overlooking Everyone is talking about what the box can run Very few are talking about what the box can help sell

NVIDIA's new $1,999 box isn't interesting because of the hardware It's interesting because of what it does to trust Most people look at a machine like DGX Spark and immediately focus on the specs > More memory > More compute > Bigger models > Wrong angle The real value shows up during the sales call When you're pitching private AI, local inference or on-prem deployments, the hardest part is usually convincing buyers that your setup is legitimate That's why third-party validation matters The moment serious infrastructure reviewers start testing the hardware, the conversation changes Now you're no longer asking a client to trust your opinion You're pointing to a platform that has already been examined, benchmarked and discussed by people their IT team already follows And that's the part most builders are overlooking Everyone is talking about what the box can run Very few are talking about what the box can help sell

35,353 views • 4 months ago

A 23-year-old launched a fake AI influencer and pulled $45,000 in a single month Zero photoshoots Zero models hired Zero locations booked Most AI creators fail because their images look like plastic The secret isn't the AI model itself. It is the master prompt He stopped generating basic images and started engineering reality Adding the realism layer changed everything -> iPhone camera framing -> Uneven skin texture -> Visible pores -> Subtle blemishes and natural light Details designed to completely trick the human eye It takes minutes to generate a post that looks 100% human Free tools. One laptop. An infinite content machine While people complain about AI taking over Brands are just paying for the engagement

A 23-year-old launched a fake AI influencer and pulled $45,000 in a single month Zero photoshoots Zero models hired Zero locations booked Most AI creators fail because their images look like plastic The secret isn't the AI model itself. It is the master prompt He stopped generating basic images and started engineering reality Adding the realism layer changed everything -> iPhone camera framing -> Uneven skin texture -> Visible pores -> Subtle blemishes and natural light Details designed to completely trick the human eye It takes minutes to generate a post that looks 100% human Free tools. One laptop. An infinite content machine While people complain about AI taking over Brands are just paying for the engagement

25,393 views • 4 months ago

ONE PROMPT AND CLAUDE FABLE 5 CLONED AN AWARD WINNING WEBSITE AGENCIES CHARGE $10,000 Higgsfield MCP + Claude Code + One prompt. That's the entire stack. No Figma, no manual video generation, no stitching tools together The MCP handles everything automatically. Nano Banana Pro for images, Seedance 2.0 for video, scroll frame slicing for the cinematic effect The total cost per full website generation with custom images and videos is around $1 to $2. Not $1,000. Not $5,000. One to two dollars Full walkthrough is above

ONE PROMPT AND CLAUDE FABLE 5 CLONED AN AWARD WINNING WEBSITE AGENCIES CHARGE $10,000 Higgsfield MCP + Claude Code + One prompt. That's the entire stack. No Figma, no manual video generation, no stitching tools together The MCP handles everything automatically. Nano Banana Pro for images, Seedance 2.0 for video, scroll frame slicing for the cinematic effect The total cost per full website generation with custom images and videos is around $1 to $2. Not $1,000. Not $5,000. One to two dollars Full walkthrough is above

18,133 views • 3 months ago

CREATORS ARE STARTING TO REVERSE ENGINEER VIRAL CONTENT WITH CLAUDE Instead of guessing why a video reached millions of people they are using AI to break down hooks retention patterns storytelling structures and recurring formats that consistently outperform everything else Hours of scrolling through TikTok YouTube or Instagram are slowly being replaced by workflows that can map an entire niche highlight outliers and reveal opportunities most creators would never notice on their own Winning attention is becoming less about inspiration and more about understanding patterns before everyone else does The next viral idea might come from analysis instead of intuition

CREATORS ARE STARTING TO REVERSE ENGINEER VIRAL CONTENT WITH CLAUDE Instead of guessing why a video reached millions of people they are using AI to break down hooks retention patterns storytelling structures and recurring formats that consistently outperform everything else Hours of scrolling through TikTok YouTube or Instagram are slowly being replaced by workflows that can map an entire niche highlight outliers and reveal opportunities most creators would never notice on their own Winning attention is becoming less about inspiration and more about understanding patterns before everyone else does The next viral idea might come from analysis instead of intuition

12,018 views • 2 months ago

A 27-YEAR-OLD CREATOR REPLACED HIS ENTIRE VIDEO TEAM WITH THREE AI TOOLS He stopped paying editors, designers and voice actors after building a workflow powered by Claude, Flux and Kling A single idea now turns into a finished short-form video in under 20 minutes, cutting production costs by more than 90% while publishing up to 8 videos every day The biggest surprise wasn't the quality, it was the testing speed Instead of spending a week perfecting one upload, he launched dozens of variations, doubled his average reach and discovered winning formats before larger creators even finished editing The future won't belong to creators with the biggest teams It will belong to creators with the fastest AI workflows Bookmark this

A 27-YEAR-OLD CREATOR REPLACED HIS ENTIRE VIDEO TEAM WITH THREE AI TOOLS He stopped paying editors, designers and voice actors after building a workflow powered by Claude, Flux and Kling A single idea now turns into a finished short-form video in under 20 minutes, cutting production costs by more than 90% while publishing up to 8 videos every day The biggest surprise wasn't the quality, it was the testing speed Instead of spending a week perfecting one upload, he launched dozens of variations, doubled his average reach and discovered winning formats before larger creators even finished editing The future won't belong to creators with the biggest teams It will belong to creators with the fastest AI workflows Bookmark this

12,329 views • 3 months ago

$30,000+/month. $40 cost. 500M views. 75,000% margin faceless creators. viral arbitrage. Claude writes scripts. "Make" automates. AI renders. YouTube scales 1 video: $0.40 → 30,000% ROI > zero filming > zero editing > just Claude and a "Make" scenario zero manual production. Claude automates the entire content factory the new creator economy runs on automation, not manpower

$30,000+/month. $40 cost. 500M views. 75,000% margin faceless creators. viral arbitrage. Claude writes scripts. "Make" automates. AI renders. YouTube scales 1 video: $0.40 → 30,000% ROI > zero filming > zero editing > just Claude and a "Make" scenario zero manual production. Claude automates the entire content factory the new creator economy runs on automation, not manpower

14,314 views • 5 months ago

A 27-year-old built a roster of models that don’t exist and signed $60,000 in brand deals in 30 days No drama No travel expenses No human error Most AI avatars fail because they look like dead-eyed robots The lips don’t sync, the movements are stiff and they can’t show real emotion He fixed that in three simple steps, taking under five minutes per video It starts with the base. He uses Nano Banana Pro to generate hyper-realistic skin texture and natural lighting. Then, he builds an asset library One fake character, dozens of outfits and angles. Infinite content from zero photoshoots But the secret is the emotion He plugs the static images into VEO 3.1. Suddenly, they aren’t just talking. They can laugh. They can cry. They can look confused. The lip-sync is flawless. "I'm not real but you can't tell, can you?" Millions of views are pouring in from audiences who have absolutely no idea they are watching lines of code The future of influence isn't human. It's perfectly engineered avatars printing money on autopilot

12,514 views • 4 months ago

No more content to load