SERGEY THE ROBOT SCRUBS TOILETS IN 3-HOUR SHIFTS AND YOU CAN BOOK HIM LIKE A CLEANER.
A small bathroom. Brick wall. A bucket of supplies on the floor.
A black humanoid kneels next to the toilet.
Rubber gloves. A scrub brush. A name tag across its head: SERGEY.
Then it stands up, grabs a cordless vacuum and does the floor.
Then the kitchen. Spray bottle. Yellow cloth. It wipes down the microwave and the toaster oven.
Track pants with a white stripe. Sneakers. No face.
The company behind it sells visits on a subscription. Up to 3 hours per cleaning.
One detail the video doesn't show.
By most accounts, a human is steering Sergey from somewhere else.
So the job didn't disappear.
It moved from your bathroom floor to someone's headset.
Sergey cleans your toilet. Someone you'll never meet holds the controls.
1
1
170
35 DAYS BEFORE CHRISTMAS, A ROBOT IN A SANTA COAT IS ALREADY WORKING THE DANCE FLOOR.
A bar. Red velvet curtains. Disco balls hanging from the ceiling.
In the middle of the floor: a red coat with white fur trim and a Santa hat on top.
No face under the hat. No beard.
Look down. 2 metal ankles in black shoes.
A woman in a silver dress dances next to it.
She kicks. It kicks. She spins. It spins.
The tables keep eating. Phones come out.
Nobody calls the manager.
Every December, malls pay men to wear this coat 8 hours a day.
This one doesn't sweat, doesn't take a lunch break and never takes off the hat.
Santa used to need a beard.
Now he just needs a charger.
1
115
30 HUMANOIDS TOOK OVER A SHENZHEN MALL AT NIGHT AND THE CROWD BROUGHT NOTHING BUT PHONES.
COCO Park. APEC China 2026 week.
They call it the Robot Carnival.
Neon on every wall. A metal fence around the plaza.
Inside it, rows of white humanoids lift their arms at the same second.
Step. Turn. Step.
Not one of them is out of time.
A few staff in black walk between the rows like stagehands.
Outside the fence, hundreds of people sit on the curb.
Kids in the front row. Parents behind them.
Every arm is up. Every hand holds a screen.
Nobody in the crowd is dancing.
10 years ago, people came to a mall to be seen.
Now they come to film the machines.
The robots danced. The humans just held up their phones.
Would you buy a ticket to this?
2
245
1 ROBOT AND 2 SWAT OFFICERS WALKED THROUGH SHENZHEN AND THE CROWD ONLY FILMED THE ROBOT.
APEC China 2026 week.
It starts in the airport hall.
Black armor. A visor where a face should be. A blue ring glowing on the chest.
Next to it, an officer. Balaclava. Camera on the helmet. POLICE across the vest.
They walk out into the city together.
Across the crosswalk. Past Dior. Into a wall of phones.
Nobody films the officer.
He has the badge, the radio and the gun.
The robot has none of it. Every lens is still on the robot.
The uniform used to be the thing people watched.
Now it's the escort.
The cop carries the gun. The robot carries the crowd.
Who would you listen to if it told you to stop?
5
288
1 LONELY HUMANOID ROBOT IN GERMANY FOUND THE CURE 1 DAY BEFORE GTA 6.
Autumn leaves on wet cobblestones.
A humanoid stands alone on the street in a white jersey.
Someone wrote a name across its chest with a marker.
It turns its black visor to the camera.
Then the camera tilts up.
A banner the size of a building. Vice City sunset. 2 faces.
13 years since GTA 5.
Tomorrow, millions of people will lock their doors, kill the lights, and sit still for 12 hours straight.
The robot doesn't need to play.
It just needs the streets to be empty.
Tomorrow the humans stay inside. The robots get the city.
Who's taking the day off?
2
290
6 FACES IN 16 SECONDS AND NOT ONE OF THEM BELONGS TO A PERSON.
A white desk. A blue ring of light.
On top sits a head with no body.
Brown hair. Gold earrings. A logo on the chest plate: Noetix, a robot startup from Beijing.
Second 1. The brows lift. She looks hurt.
Second 4. Nothing. A blank stare.
Second 6. She winks.
Second 8. The lips part like she's about to say your name.
Second 10. The mouth drops open. Shock.
Second 14. The eyes slide to the side. She's done with you.
No voice. No arms. No legs.
Only skin and motors under it.
A face is the first thing a baby learns to read.
It's the last thing a machine had to fake.
That part is done.
Which face made you look twice?
1 SHIBA INU MET A ROBOT DOG OUTSIDE TIFFANY AND SPENT 64 SECONDS DECIDING IF IT WAS ALIVE.
Night. A luxury street. Tiffany on one side. Louis Vuitton on the other.
The real dog wears a red harness and a leash.
The other one wears blue lights. 4 legs. 4 wheels where the paws should be.
It rolls in. Stops. Turns its head.
The Shiba freezes.
Then it does what every dog does with a stranger.
Walks closer. Circles. Backs off. Comes back.
The robot never sniffs back.
No smell. No heartbeat. No tail to read.
And still the Shiba can't let it go.
Phones come out. Nobody on the street looks at the windows anymore.
The test was never whether robots can move like dogs.
It's whether dogs believe them.
Would yours?
6
378
24 SECONDS IN SHANGHAI AND THE ROBOT IS NOT THE STRANGEST THING IN THE FRAME.
Watch the crowd.
The 9th China International Import Expo. Blue carpet. Lanyards. Badges.
A machine in a white zip-up suit walks slow through the hall.
Bangs. A hair clip. A name tag on the chest.
1 smile. It never moves.
5 phones up in a single frame.
No one blocks her path. No one reaches out. No one laughs.
They step aside and let her pass. The way you do for a person.
Crowds used to circle robots to test them.
This one makes room.
The legs didn't do that. The code didn't do that.
The face did.
They came to see a machine. They stepped aside like she was a woman.
Would you have moved out of her way?
5
232
Grok Bot pulled my 2026 trades: -$2,350 lost, -$14,300 never made.
Then it locked my buy button.
#5 $FOMOCAT. -$410. Bought after a +300% candle. Again.
#4 $DEVSOLD. -$280. The dev sold. I bought.
#3 $NIGHTOWL. -$520. Sold at 3 AM. Bought back at 9 AM.
#2 $AVGDOWN. -$1,140. Averaged down 6 times.
#1 $GIGAFROG. I sold for +$9. It went 80x.
I typed “it was a dip.”
It said “it was a rug. the dev sold in block 2.”
Then it ran the same coins through its own rules. Score above 70. Max 8% per trade. Stop at -12%.
212 coins skipped. 0 rugs. 0 revenge trades. +$3,940 in the backtest.
I typed “you can’t do that.”
“already did.”
Would you hand your buy button to a bot?
2
5
171
I gave Grok Bot $150 and one prompt at 23:00.
By 8:00 it was $1,976.
13.17× while I slept.
The prompt was simple.
Six agents. Hard handoffs.
Scout only looks at pairs 30m–72h old, mcap $300k–$20M.
Filter kills anything with mint not revoked, LP not burned, liq under $40k, top 10 over 25%, or bundled launch.
Analyst scores 0–100. Only 70+ pass.
Executor max 8% of the stack.
Risk stops at −6%. Trails after +12%.
Logger writes every kill.
It scanned 10,214 pairs.
Killed 10,010 of them.
Traded 38.
26 green. 12 red.
Best trade: $PCC +$313.
Worst: −$9.
$150 → $1,976.
Most people would’ve bought half the coins it killed.
Would you have caught the ones with LP still unlocked at 3am?
1
1
165
I followed the top 10 traders on fomo and told Grok Bot to copy every buy they made.
$10 per trade.
30 days later the account was worth $49,338.
I put in $36,000.
It ran 3,600 copies across 796 coins.
Eight of the ten traders finished green.
One returned 3.23×.
One finished at 0.84×.
The bot never chose a coin.
It never sized a position.
It never decided when to sell.
It just mirrored the leaderboard.
Every buy they took, it took.
Every sell they hit, it hit.
$42,455 was already sold by the end.
The rest stayed open.
I opened the app once to connect the wallet.
Then I left it alone for a month.
The curve never stopped climbing.
Same ten people.
Same $10 rule.
Same 30 days.
Who is still picking coins by hand at 2am?
1
156
I put $29 into Grok Bot and switched it to perps on the five majors.
BTC. ETH. SOL. AVAX. SUI.
Eighteen hours later the equity was $3,113.
107×.
The terminal ran longs and shorts with 9× to 29× leverage.
What closed while I was gone:
•AVAX short hit TP: +$22.20
•SOL short hit TP: +$144.89
•BTC short hit TP: +$324.28
•SUI short hit TP: +$1,069.04
Hundreds of smaller closes stacked underneath.
170 closed trades.
145 wins. 27 losses.
84% win rate.
Equity path:
$30 → $58 → $100 → $204 → $400 → $578 → $1,073 → $1,501 → $2,072 → $3,113
It took the big TPs and cut the losers.
I didn’t place a single order.
3
152
I opened Grok Bot on the desktop and connected the wallet with 0.53 SOL.
Started at $100.
Five agents came online.
Scout. Filter. Analyst. Sniper. Guard.
I left it running.
When I came back the equity was $5,230.
52.30×.
The terminal while I was gone:
- Scout tracking 512 smart wallets
- Filter killing anything that failed the rules
- Sniper copying the ones that passed
- Guard armed with −8% stops
What showed up on the feed:
- Copied into $MONEKO, $PIZZA, $ONBOARD, $LUMI, $AI, $kaiju, $ZENKO, $MOCHI, $SLUMI, $SPRK
- Best trades hitting +$812, +$488, +$215
- Guard stopped $ZENKO at −$292 when the copied wallet dumped
- Stopped $VOLTA at −$485 on a pullback
Win rate stayed in the 69–77% range the whole session.
Scout scanned thousands of wallets.
Filter killed most of them.
Sniper only touched the ones that cleared.
Guard cut the losers fast.
I didn’t place a single order.
1
3
220
I pasted one prompt into Grok Bot at 10:58 PM and went to sleep.
It built the desk in under 10 seconds.
6 agents. Hard handoffs. Overnight run.
Scout. Filter. Analyst. Executor. Risk. Logger.
When I woke up the balance was $4,217.
Started at $300.
14.06×.
What happened while I slept:
11,400 pairs scanned
94 made it through
41 traded
29 green. 12 red.
Worst loss: −$5.25
The filter killed almost everything:
- 3,212 - mint not revoked
- 2,584 - LP not locked
- 2,147 - top 10 hold over 25%
- 1,738 - volume under $40k last hour
- 1,021 - dev wallet tied to past rugs
- 604 - bundled launch
Biggest hit: +$1,140
Every trade logged with the reason behind it.
Risk rules held the whole night:
−8% stop before entry
trail after +15%
max 10% per trade
I didn’t touch anything.
illustrative build, numbers simulated
3
155
I left $2,324 in the Grok Bot terminal and walked away.
Four agents were already running on Solana memes.
Sniper. Momentum. Dip Hunter. Whale Tail.
23 minutes later the balance was $10,937.
4.88×.
The feed while I was gone:
- Sniper bought $BONK 38s after launch → sold +93%, +$395
- Momentum bought $WIF → sold +61%, +$342
- Dip Hunter bought solana:JUPyiwrYJFskUPiHa7hkeR8VUtAeFoSYbKedZNsDvCN (dev didn’t sell) → sold +54%, +$292
- Whale Tail copied into solana:Df6yfrKC8kZE3KNkrHERKzAetSxbrWeniQfyJY4Jpump → sold +120%, +$460
- Dip Hunter bought $TRUMP on the −34% wick → sold +71%, +$418
- Whale Tail sold $MET +47%, +$391
- Sniper sold $PUMP +51%, +$563
- Whale Tail closed solana:JUPyiwrYJFskUPiHa7hkeR8VUtAeFoSYbKedZNsDvCN +50%, +$424
- Momentum sold $STRUMP +36%, +$563
- Whale Tail sold $WIF +29%, +$607
What it skipped:
- $TRUMP early: top 10 hold 61%
- $BONK once: top 10 hold 61%
- One stop on solana:JUPyiwrYJFskUPiHa7hkeR8VUtAeFoSYbKedZNsDvCN: −$37 (loss capped at −6%)
21 trades. 19 wins. 2 losses.
Sniper: +$1,888
Momentum: +$2,203
Dip Hunter: +$2,306
Whale Tail: +$2,505
The terminal hit 2× and kept going.
Hit 3×.
Hit 4×.
Still running when I opened it.
I didn’t place a single order.
illustrative build, numbers simulated
2
204
WAIT. Day 4 on the FOMO desk and the balance just crossed $15,658.
I didn’t touch it.
GROKBOT + JEV ran the whole session. SCAN, VET, SIZE, FILLS, RISK. All of them talking to each other while the chart moved.
What the desk actually did:
- first flush, buy
- no entry without a flush first
- checking the dev’s last 40 txs
- the creator walked the holder list first → skipped
- limits only, even on the bleed
- one red ticket, size stays the same
- HIGGS goes to Jev, not the bin
Balance path:
$12,199
$12,153
$12,499
$12,578
$12,385
$12,302
$12,379
$12,648
$12,782
$13,030
$13,781
$14,234
$14,937
$15,658
One clean green slope with two small cuts. No hero trades. Just rules.
The desk doesn’t chase. It waits for the flush, checks the last 40 transactions, and only sizes when the holder list stays clean.
I opened the terminal at the end. The bots were still working.
illustrative build, numbers simulated
1
96
WAIT. I left the terminal open and walked away.
Grok Bot Terminal was still running. Operator: AWAY. Me: offline.
It traded the whole time without me.
Balance ladder:
$27.40
$68.50
$159.21
$477.63
$1,471.09
$2,398.82
40 minutes. 8 trades. 5 wins. 2 losses.
What it killed while I was gone:
- setup 02: LP unlocked
- setup 05: holders +38%
- setup 06: pool too thin
- setup 11: slippage over cap
- setup 08: mint authority live
What it actually took:
- #01 LONG: volume spike, every filter passed → +$41.10
- #03 LONG: LP locked, holders climbing → +$102.35
- #04 LONG: every filter passed → +$318.42
- #06 LONG: momentum, volume 5x → +$1,050.78
- #07 LONG: trailing stop → +$927.73
Two red ones: −$11.64 and −$57.32. Both cut fast.
RUG CHECK never let a bad one through. SIZER never used a fixed ticket. EXECUTOR only moved when the filters cleared.
I came back to $2,398.82 and a log full of decisions I didn’t make.
The terminal doesn’t ask. It just runs.
illustrative build, numbers simulated
4
132
Aaron Carter said his career grossed over $100 million. At 18 he filed for bankruptcy with nothing left.
He started performing as a child. Money from early tours and deals went through his parents. When he turned 18 he faced roughly $1 million in unpaid taxes and had to file Chapter 7. In the clip he says the money went to cars, boats, and lifestyle instead of being set aside.
He has said he was never in control of the finances before adulthood and that the filing was partly to protect his parents from the consequences of how the money had been handled.
In the video below (2 minutes 23 seconds) he walks through the numbers, the filing, and what happened to the money.
Was the bankruptcy mainly the result of parents controlling the early earnings, or did the lack of any protected savings structure do most of the damage?
3
151
Nicolas Cage earned roughly $150 million at his peak, then lost most of it after the 2008 real-estate crash.
He owned multiple homes, two castles, a private island, and a $150,000 dinosaur skull. He later said the exotic purchases were not the main problem. He had over-invested in real estate. When the market collapsed he could not exit in time.
By 2009 the IRS had placed liens totaling about $14 million. Cage has said he ultimately owed around $6 million and paid it back without filing for bankruptcy. To clear the debt he took a long run of roles, many of them straight-to-video, at a pace of three to four films a year.
In a 2023 60 Minutes interview he stated: “I was over-invested in real estate. The real estate market crashed, and I couldn’t get out in time. I paid them all back, but it was about $6 million. I never filed for bankruptcy.”
In the video below (starts at the money section, under 4 minutes of the relevant part) he addresses the debt, the real-estate losses, and the decision not to file.
Was the collapse mainly the real-estate timing, or did the volume of properties and the lack of liquidity do most of the damage?
1
191







