username retweeted
HOW IS THIS EVEN POSSIBLE?
I asked Grok Agent to run a full diagnostic on my PC. What it uncovered was shocking: a hidden crypto miner silently using my computer's resources.
Here's exactly what Grok did:
Scanned running processes - identified unusual background activity and processes consuming suspicious amounts of CPU.
Tracked down the suspicious process - located its executable file and investigated where it was hiding, including Temp and AppData folders.
Investigated the files - checked related executables and components to determine whether they were malicious.
Checked persistence mechanisms - inspected startup entries, scheduled tasks, and other methods the malware could use to launch itself again after a reboot.
Traced the infection - analyzed how the miner was running and what files and system entries were connected to it.
Removed the threat with my approval - Grok showed me the suspicious components and asked for permission before removing them.
The entire investigation took around 15 minutes.
And here's the part that really got me: my antivirus hadn't detected the miner at all.
Grok didn't just flag a suspicious process. It traced the executable, uncovered related files, identified how the malware could survive reboots, and helped remove the malicious components.
I still ran a full antivirus scan afterward. AI is an extra layer of investigation, not a replacement for proper security software.
But think about it: how long has something been running on your PC without your knowledge?
Your CPU could be working for someone else's profit right now.
I'll leave the diagnostic prompt in the comments. Just send it to Grok and let it guide you through the investigation.
Would you trust an AI agent to investigate your PC? Or would you rather not find out what's hiding in the background?
username retweeted
Holy f*ck. A shocking statement from Lauren Tan:
"You under 40? Build a team of Grok agents like I did. Not tomorrow.
The window to success is open, but not forever."
In this 55-minute workshop, she walks through how one engineer runs 80+ GrokBot agents like an entire company.
GrokBot → Rules → Self-Checks → 80+ Agents → A Company Of One
One agent per client account, plus one that reports everything the team shipped overnight.
Her PMs and designers now ship code too, because the setup does the checking for them.
Most people add another bot when the first one fails. That only multiplies the mistakes.
Stop wasting your time.
Worth more than any $10,000 agentic engineering bootcamp.
Bookmark it and watch it tonight.
Then read how to set up a GrokBot team that runs without you in the guide below ↓
username retweeted
SpaceXAI engineer Lauren Tan just dropped a 57-minute masterclass on everything she knows about building AI agents.
"I woke up today and 30 updates had already shipped while I slept, and they were great.
I don't really look at the code anymore. I'm the head chef now. I don't cook the food, I run the kitchen."
In this 57-minute talk, she breaks down how she went from babysitting a single agent to building a GrokBot team that ships 2,000+ updates a month without her.
GrokBot → Agents That Check Themselves → A Team You Trust → Work That Ships While You Sleep.
She didn't learn to trust her agents more. She built an entire system that checks their work, catches mistakes, and lets them operate independently.
While most developers are still manually reviewing every AI-generated answer, she's running a team of agents that can build, verify, and ship work without constant supervision.
This is worth more than any $8,000 agentic engineering bootcamp.
Bookmark this and watch all 57 minutes tonight.
Then read the Grok Bot team guide below to learn how to build a team of AI agents you can actually leave alone ↓
username retweeted
This is f*cking crazy.
A 18-year-old student turned $70 into $8,653 with a Grok Bot copy-trading setup.
The bot scanned 13,962 signals. It traded 61.
That’s a 0.27% pass rate. The other 99.73% were rejected: concentrated holders, thin liquidity, a dev holding too much, batch buys, and wallets that had already started selling.
$70 → $317 → $945 → $3,878 → $8,653.
The lesson he took: the profit wasn’t about picking winners. It was about a screening bar so high that almost nothing got through.
Most of us do the opposite: see a tip and buy.
A single result. Not typical. Very high risk. Not financial advice.
Full setup for Grok trading desk below ↓
username retweeted
This is f*cking insane.
I gave Grok one job: call 5 auto repair shops and find the cheapest front brake pad + rotor replacement for a 2019 Honda Civic.
The gap between the highest and lowest quote? $410.
In the video below, you can see it make the calls, ask for the price, and even ask if there's anything cheaper. I didn't say a word.
The quotes:
$389, $449, $525, $649, $799.
Time it took: 12 minutes.
Time it would have taken me: half a day of calls.
Elon Musk has been showing it book doctors and plan trips. Nobody's talking about the boring part: it's a price-comparison machine that actually picks up the phone.
My take: the most expensive part of a purchase usually isn't the price. It's the effort of comparing, and most people never bother.
That laziness is where the $410 went.
Would you let an AI negotiate for you, or is that one step too far?
username retweeted
In 2019, Elon Musk and Alibaba’s founder argued about whether machines could be smarter than people. Only one of them has been proven right so far.
In the video below, you can see both sides.
One said: computers are just clever tools, humans have the heart, the skill that matters is LQ.
The other said the picture of technology awareness is a flat map with a few very tall spires, and most people can’t see the shape from the ground.
Seven years later, most people are still on the ground, using AI like a chat window.
The few on the spires are running teams of agents.
Prompts → Grok Bot → Memory → Skills → Connectors → Routines → A Team You Own
Save this before it disappears from your feed
The official docs, in the order I’d read them:
Get started
docs.x.ai/grok-bot/get-start…
Work with Grok Bot
docs.x.ai/grok-bot/bots
Use cases
x.ai/bot/use-cases
Skills, routines and automations
docs.x.ai/grok-bot/skills-ro…
Vault (logins and cards, safely)
docs.x.ai/grok-bot/vault
Shared Bots
docs.x.ai/grok-bot/shared-bo…
Grok Bot on mobile
docs.x.ai/grok-bot/mobile
Reading all of this costs $0
Most people will stay on the ground. The few who build a team of agents will see the whole map.
Who do you think was right?
Bookmark it and start with the first link tonight
Then read how to build your own team of Grok Bots in the article below ↓
username retweeted
Elon Musk on a nonprofit that became an $800 billion for-profit: “the exact opposite of what I donated the money for.”
in the video below, he explains why this is the one thing that bothers him about the AI race.
his setup: a nonprofit, built to be open source and owned by the world, ends up as a closed-source company worth $800 billion.
he doesn’t call it a crime. he calls it an issue, and a legitimate one.
on Anthropic, he says the people he’s met there seem well-intentioned.
so the question isn’t whether the labs are good or bad. it’s whether they still answer to the people who funded them.
valid complaint, or the price of building fast?
either way, you can’t control who owns the lab. you can control who owns your team.
your rules. your files. your approvals. the setup for building your own team with Grok is in the article below ↓
username retweeted
HOLY SH*T. A SpaceXAI engineer (ex-Cursor) Lauren Tan:
"99% of people use GrokBot like Google.
I run 60+ bots, and I only messaging the Chief of Staff.
in just 1-hour workshop, she shows how to turn one GrokBot into a team that runs your email, calendar, Slack and code, around the clock.
One GrokBot → Specialist Bots → Chief of Staff → Cloud Agents → A Team That Never Clocks Out
a Specialist Bot does one job and keeps its own memory.
a Chief of Staff bot watches your Slack and starts the work before you ask.
anyone can chat with one bot. almost nobody has a team doing the work.
worth more than any $3,000 AI agent course.
bookmark it, watch it today.
then read how to give your GrokBot a real team in the article below ↓
username retweeted
Hell YES... Elon Musk’s Grok Bot just made you stop shopping by hand
It finds the item, checks price across stores, and tracks the order. Here are the lines I use:
First thing I gave it: a new MacBook from Walmart, $250 cheaper than Amazon.
6 things I’m never shopping for by hand again, and the exact line I say:
→ eBay: “Find this exact item used in good condition, under $1000, and show me 3 sellers with top ratings” → Walmart: “Compare this item with Amazon and Target, and tell me the cheapest delivered by Friday” → Target: “Watch this item and message me the moment it goes on sale”
→ Home Depot: “Find the best-rated cordless drill under $150 and compare it with Lowe’s”
→ Chewy: “Reorder my dog food every 30 days wherever it’s cheapest, ask me first”
→ DoorDash: “Find a dinner under $25 for two, 4 stars or better, delivered in 40 minutes. Ask me before ordering”
Add one rule first: never pay for anything without asking me
Say the Walmart line to Grok Bot tonight
Then read how to turn one bot like this into a whole team in the article below ↓
username retweeted
Holy sh*t. SpaceXAI senior engineer Lauren Tan:
"400 applications over two years and no offers
Then he showed up running 20 agents and got a $500K offer"
In this 1-hour workshop, Lauren shows how to build a Grok Bot working team from zero
Grok Bot → Chief of Staff → Managers → Workers
A chat window gives you one bot. An agent team gives you a system
This free workshop is worth more than most $600 agent engineering courses
Bookmark and watch it today
Then read article below
username retweeted
ELON MUSK’S GROK BOT TRADED THE BINANCE HACK WHILE I SLEPT. I WOKE UP $10,287 UP
$158 → $10,287.
Over $5.6b funds were stolen. BNB and many other alts dumped overnight. six agents moved at once:
Radar: tracks the news
Tracer: follows the stolen funds, feeds the next two agents
Anchor: spot limit orders on the dip
Lever: margin limit orders
Guard: sells on a sharp reversal or at +50% profit
Scribe: logs why every trade happened and learns from it
the bot bought back the dumped tokens with limit orders, and the positions were already green by morning.
I didn’t chase the move. I woke up to it done, with the reasons written down.
one result. not typical. margin is dangerous. not financial advice.
setup below ↓ save it before the next headline.
username retweeted
This is f*cking insane. Elon Musk openly admits Grok is still behind the leaders.
then he named the number that matters: Grok Bot is growing roughly 100% a month.
on catching Anthropic and OpenAI, he said he expects Grok to reach the frontier sometime next year, most likely.
so how does the #2 plan to win? efficiency.
by his numbers, Grok 4.6 finishes a long task in about 53 turns and 0.5B input tokens. Claude Opus 5 Max needs about 103 turns and 2B.
half the steps. a quarter of the tokens. about $0.84 a task.
Prompts → GrokBot → Agents → Self-Learning Systems → AI That Builds AI
using AI isn’t asking Grok a question and copying the answer. it’s agents that learn from every failed run and come back better without you.
95% of people still use Grok like a search bar. the 5% building self-learning agents will run everything.
in this 1-hour session, Elon sits down with SpaceXAI engineers and maps out what AI engineering looks like next.
worth more than a Stanford AI course.
bookmark it and watch today.
then read the article below how to use Grok correctly ↓
username retweeted
this is f*cking addictive.
5 Grok agents pass every coin to each other like a relay. one says no, the coin is dead.
$60 → $9,258 overnight.
everyone says bots lose because they buy anything that moves.
so at midnight I put $60 on a kill-first desk and went to sleep.
for the first 40 minutes it did nothing. “filter says no”, over and over.
→ coins scanned
→ killed by the filter
→ killed by the analyst
→ trades: wins / losses
→ best trade / worst trade, stopped by a hard stop
$60 → $9,400. multiple, e.g. 156x while I slept.
by hand I would have aped half of what it killed. the relay just kept saying no until something was actually clean.
one result, one night. not typical. memecoins are extremely high risk and you can lose everything. not financial advice.
the prompt:
"You are a Solana memecoin desk built around one rule: reject first, buy last. Five agents work as a relay, and every handoff passes through Grok core. A single “no” from any agent ends the coin.
SCOUT: find new pairs and send every one to Filter.
FILTER: reject a coin if any of these is true:
liquidity is not locked
the top 10 holders own more than 30%
volume is below $40k
the dev wallet is still moving
price is already up 80%+ in the last hour
ANALYST: score each survivor from 0 to 100 on momentum, buyers and holder growth. Anything under 70 is rejected.
EXECUTOR: enter only if all 6 checks pass. One entry per coin.
RISK: set a hard stop before the entry. Never move it. Never average down.
If nothing passes, do nothing. Log every rejection with its reason. Run overnight."
bookmark this before the next coin you ape rugs.
which rule would you add to the filter?
username retweeted
This is f*cking insane. SpaceXAI senior engineer:
“99% of people only use 1% of what GrokBot can do. They run a single agent, no loops, no graphs
I run 30+ GrokBot agents as a team. By 7 a.m. my Chief of Staff has planned the day, my PM has assigned the work, and my workers have already shipped half of it”
In this 1-hour session, SpaceXAI builds that team from the ground up.
GrokBot → Chief of Staff → PM → Workers → Self-Improving Team
A loop keeps sending a task back until it really works.
A graph passes one agent’s result to the next, with no you in the middle.
Anyone can ask GrokBot a question. Almost nobody gives it a team to run.
Worth more than most $1,500 agentic engineering courses.
Watch it today.
Then read how to build your own fleet of GrokBot agents in the guide below ↓
username retweeted
ELON MUSK’S GROK JUST RAN A TRADING DESK FOR ME. $50 → $11,400 IN ONE NIGHT
I didn’t ask it what to trade.
I built six agents and let them work.
scanner. rug check. sizer. executor. logger. equity curve.
one screen. one job each. no code written by hand.
I went to sleep. in the morning there was a log of every trade and the reason behind it.
the money isn’t the shift. the shift is that I stopped being the one asking “would you long this?” and started being the one who reads what already happened.
most people still use Grok like a chat window.
single result. not typical. very high risk. not financial advice.
full build below ↓
username retweeted
Don't f*cking miss this. SpaceXAI engineer Lauren Tan:
"300 applications over two years. None offers
Then he showed up running 30 agents and got a $700K offer"
Lauren spends 1 hour showing how to build a Grok Bot team that hands out the work and checks the results itself.
Grok Bot → Chief of Staff → Managers → Workers
a chat window gives you one bot. a team gives you a system.
this free workshop is worth more than most $2000 agent engineering courses.
bookmark it and watch today.
then read the article on building a Grok company below ↓
username retweeted
You should f*cking see this. SpaceXAI engineer Lauren Tan:
"You’re using 1% of GrokBot, and I'm here to show you the other 99%"
I run a team of 30+ GrokBot agents. by 7 a.m. my Chief of Staff has the day planned, my PM has handed out the work, and the workers have shipped half of it.
in this 1-hour session, SpaceXAI shows how to build that team from scratch.
GrokBot → Chief of Staff → PM → Workers → Self-Improving Team
→ a loop sends every task back until it actually works
→ a graph lets one agent hand its result to the next without you in the middle
anyone can ask GrokBot a question. almost nobody gives it a team to run.
worth more than most $2,500 agentic engineering courses.
watch it today.
Then read how to build your own team of GrokBot agents in the detailed guide below ↓
username retweeted
this is f*cking incredible. Elon’s team built GrokBot to ship code while you sleep. SpaceXAI engineer Lauren Tan just revealed the 21+ agent setup behind it
the idea she says came from Elon: bots that ship work, not bots that chat.
how it scales: one agent → a team of bots → each with its own computer → automerged PRs
result: around 1,000 PRs a month through GrokBot.
the choice for any team: pay a salary forever, or spend tokens once and make your codebase agent-ready.
one person can now enforce rules that used to take a whole org.
most teams measure agents by tokens saved. the real number is the work you couldn’t do before.
the agent that makes money is the one that replaces the hire you never wanted to make.
save this before you scale your team.
username retweeted
This is f*king insane. SpaceXAI engineer Lauren Tan just opened up the setup that turned 2 years of rejections into a $500K offer
the story: 250 applications, zero offers. then he showed up running 20 agents.
in this 1-hour workshop, Lauren walks through how to build a Grok Bot team that delegates the work and checks its own results.
Grok Bot → Chief of Staff → Managers → Workers
a chat window gives you one bot. an agent team gives you a system.
this free workshop is worth more than most $800 agent engineering courses.
bookmark it and watch it today.
Then read the full article below ↓
username retweeted
holy sh*t. Your AI can now keep a second brain that builds itself. Karpathy’s pattern, running on Grok Bot
stop explaining yourself from scratch every chat.
three bots share one group chat, each with its own memory:
the reader takes in whatever you drop (links, PDFs, voice notes) and never edits it
the librarian updates every wiki page that source touches, usually 10 to 15 pages
the checker works at night and hunts for contradictions, stale claims and pages with no links pointing to them
→ you ask something: the answer gets saved as a page, so the brain grows from use, not only from reading
→ new info conflicts with an old page: the checker tags you instead of overwriting it
→ 30 days without a fresh source: the page is marked unconfirmed
a second brain can sound sure and still remember the wrong thing
Reader collects. Librarian writes. Checker doubts.
every page has to show:
the claim
the source it came from
the date it was last confirmed
paste this into Grok Bot:
“Build me a second brain with three bots in one group chat. Keep my sources untouched in raw/, maintain wiki/ yourself, file good answers back as pages and run a nightly check for contradictions. Show me the structure first. Do not create anything until I confirm”
start with one topic you actually work on, not your whole life
full setup below







