Pure retweeted
🚨 ETH COULD HAVE ONE MORE FLUSH BEFORE THE REVERSAL
ETH lost the $2,435 support, so I don’t think the downside is finished yet
My base case is a short relief bounce into that level, followed by another move lower
The $2,345 area stands out to me because it lines up with the previous lows and the projected end of the correction
If we sweep that liquidity and reclaim $2,435, I’d start looking for a much stronger recovery
That’s the setup I’m watching - one more shakeout before ETH can turn higher
I'll warn you
FOLLOW + NOTIFS ON!
🚨 ETH PERFECTLY MATCHES THE WYCKOFF PATTERN
Look at the structure:
ETH has already printed the same sequence of swings that preceded the major expansion in the previous cycle
We got the shakeout
We got the recovery
Now we’re sitting at the point where the real move can begin (point 7)
From here, I’m watching the same progression:
1. $3K-$3.2K - first major breakout
2. $4K-$5K - expansion phase
3. New ATH - price discovery
4. $8K+ - potential final leg
What would invalidate this idea?
A sustained break back below the current macro structure
Until then, I’m expect move higher
FOLLOW + NOTIFS ON!
Pure retweeted
🚨 BTC HAS ONLY ONE MAJOR RESISTANCE LEFT 🚨
I think the market has already done most of the hard work on the way up
First, BTC reclaimed the 200W MA at $65.4K, which had been acting as the main long-term barrier
Then $68.8K, the 2022 ATH, was taken back
After that, $71.4K turned into support instead of resistance
And now even $78.6K has been broken
That leaves $82.2K as the last major level standing between BTC and a much larger move
That matters because we’re no longer trying to recover the old structure
We’re rebuilding above it
The only major resistance I see left is $82.5K
Break and hold above it, and $88K becomes the next target
The level I’m watching underneath is still $78.5K. If it holds, the bullish structure stays intact. If it fails, $71.4K is the first place I’d expect buyers to step back in
So for me, the roadmap is pretty simple:
$78.5K = support
$82.5K = breakout
$88K = next target
And honestly, after everything BTC has already reclaimed, I think the bulls have earned the right to be taken seriously here
FOLLOW + NOTIFS ON!
🚨 BTC BULL TRAP IS IN 🚨
BTC broke above $82K, but failed to hold the breakout
That leaves a lot of late longs trapped above the old range
Below price, there’s still an unfilled FVG around $69K-$72K and a cluster of liquidity around the previous lows near $58K
That gives sellers a clear path to push price lower while forcing trapped longs to unwind
And the bottom is where the setup gets interesting
A sweep of those lows would clear the remaining liquidity and shake out the late sellers
Once that happens, there’s much less supply left above the market
Think about it!
I called this structure before the breakout
Now I’m tracking where the reset ends
Be sure to follow if you want the next move before it becomes obvious
Pure retweeted
🚨 BTC IS VERY CLOSE TO A CORRECTION 🚨
Here’s how I’m looking at the next few weeks
For this week, I expect Bitcoin to push above the current highs
That move should squeeze shorts and trap late longs, but I don’t expect BTC to break cleanly above $82K
From there, I’m expecting a rejection and our first major correction
The area I’m watching for the final retest is $68K-$71K, likely around the end of September or early October
If the bottom is already in, this could be the last major dip before the next leg of the bull run
That’s the playbook I’m working with for now
Testing a new approach
Let’s see how it plays out
Make sure you follow me so you'll be the first to know if anything changes
🤝 Paid partnership
Pure retweeted
🚨 THIS BTC BREAKOUT IS ANOTHER TRAP 🚨
I know that's the opposite of what everyone's feeling right now
Bitcoin filled the $72K FVG
After that pushed straight into $78k - yes, that’s very unusual
But despite the size of this move, we're still following the exact same script:
FVG fill -> bull trap -> dump
The fact that we ran further than usual isn't bullish. It's the manipulation
Running higher than the last three traps doesn't break the pattern - it just means this one needed a bigger crowd to fall for it
We've seen this setup before at $97K FVG and $82K FVG
Both times the crowd got confident right before the reversal - and they were wrong
My roadmap is pretty simple:
$78K -> $55K -> $48K-$44K (bottom) -> $100K+
I'm not bearish on Bitcoin long term. I'm bearish on buying it after the trap has already sprung
Don't forget - I've accurately predicted most of the major price movements this cycle, including the $126K ATH
Follow + Notifs on, you don't want to miss my next call
🚨 BITCOIN JUST ENTERED THE BULL TRAP ZONE 🚨
The move above $70K was exactly what the chart needed to pull liquidity into the market
What matters now is not whether BTC can push higher, but what happens to that liquidity once price reaches the upper boundary
This is where late longs start chasing momentum while trapped shorts are forced to cover
And that creates two possible paths
Scenario 1: $78K -> rejection -> $70K -> $60K -> $48K
Scenario 2: $78K -> $83K -> rejection -> $70K -> $48K
The destination stays the same
The only question is how much liquidity BTC takes before getting there
That’s what I’m watching now
To make sure you don't miss any important updates, be sure to follow me!
Pure retweeted
🚨 SPACEX IS NOT DONE - THE IPO UNWIND IS 🚨
The first months after the IPO followed a very specific sequence
1. THE INITIAL POP
IPO launch: $135
Peak close: $201.80
The first move was pure demand expansion
2. DISTRIBUTION
Within five weeks, $SPCX had flushed almost 50% and traded below its IPO price
That strength faded quickly
3. THE SUPPLY SHOCK
Then came the Aug. 6 (Unlocks):
-911.5M shares
- ~$98.7B of potential supply
- 143% of the existing float
That sounds like the perfect setup for another collapse
Instead, $SPCX closed the day +6.1%
4. THE SHORT SQUEEZE
Short interest then collapsed:
34% -> 11%
The stock ripped back toward $141
But this is where people can misread the chart
Short covering creates buying pressure
It doesn't automatically mean new long-term demand has arrived.
5. THE REAL TEST COMES NEXT
There are still eight unlock tranches scheduled through December 8
By then, as much as ~40% of the company could become tradable.+
So I don't expect the market to immediately go vertical from here
The more important question is whether each new wave of supply can still push price lower
Because we've already seen something interesting
July 7: Nasdaq-100 inclusion brought massive passive demand - and $SPCX still sold off
August 6: 911M shares became available - and $SPCX still closed green
Good news couldn't save the price
Bad news couldn't break it
That is the kind of reaction I want to see near a potential bottom
And there's a historical parallel worth watching
For example:
Facebook went public at $38 in 2012, then dropped roughly 54% as lock-up fears hit
The November unlock brought another huge supply event
Facebook closed that day +13%
It eventually became the beginning of a much larger accumulation process
$SPCX just printed a remarkably similar reaction
But there’s one important difference:
A squeeze is not a bottom
At roughly 88x sales, this still needs time to digest the IPO excess
My read:
Remaining unlocks -> supply absorption -> $100-$120 base -> accumulation -> next expansion
The market may still look ugly while the actual bottom is being built
That’s usually how accumulation works
Be sure to follow me!
I'll warn you
🤝 Paid partnership
Pure retweeted
🚨 AI IPO COULD BECOME THE MARKET’S BIGGEST PROBLEM
Read this carefully before assuming the next IPO wave is bullish
SpaceX, OpenAI and Anthropic
All of them moving toward the public markets at the same time, with trillions in combined private valuations
And the interesting part isn’t the IPOs themselves
It’s where that capital is going to come from
The S&P 500 is already heavily dependent on a small group of mega-cap tech companies
Nvidia, Microsoft, Google, Amazon, Meta, Apple and Tesla are carrying a huge part of the index
Now imagine the biggest private AI companies entering the market and asking investors for hundreds of billions in fresh capital
That money doesn’t just appear
Funds have to rebalance(!!!), take profits and free up liquidity from somewhere
And where does that liquidity usually come from?
The stocks that already went up the most
That creates a very interesting chain reaction:
AI giants go public -> capital gets redirected -> mega-cap tech gets sold -> S&P loses its strongest support
And there’s another part people are ignoring
IPO markets usually become most aggressive when private investors have the most confidence in valuations
Early investors finally get liquidity
Employees can monetize equity
VC funds can realize years of gains
And retail gets access to companies that were previously impossible to buy
We’ve seen this movie before
During previous bubbles, investors kept concentrating into a handful of companies because they believed the leaders were simply too important to fall
Until liquidity started leaving
The same thing can happen again without AI being a scam
AI can completely transform the world and still be massively overvalued at the wrong price
That’s why I’m watching these IPOs differently
🤝 Paid partnership
Pure retweeted
🚨 BTC IS LOSING THE WEEKLY STRUCTURE 🚨
The rejection from the $65.3K weekly high is the first warning
Price tried to reclaim the range top, failed, and immediately got sold back down
Now $62.7K is the key pivot
If BTC fails to hold it, the next move likely targets the compression zone around $61K-$60K
That’s where I expect liquidity to start getting swept
A break below the compression would expose $59K-$60K as the next major downside area
For now, I’m not interested in chasing longs into resistance
The market needs to reclaim $65.3K before the bearish structure starts to invalidate
Pure retweeted
When one of the world's best AI engineers says he feels behind... you should probably pay attention.
OpenAI co-founder Andrej Karpathy recently admitted:
"I've never felt more behind as a programmer."
Not because he's getting worse.
Because AI is changing faster than anyone can comfortably keep up with.
The biggest shift isn't "vibe coding."
It's how software gets built.
Your experience.
Your judgment.
Your taste.
Those become the advantage.
AI agents handle more of the execution.
That's the direction Karpathy sees engineering moving.
Not humans vs. AI.
Humans designing systems while AI does more of the work.
The interview is worth watching.
The article below explains why the best AI engineers are already moving from prompts to loops.
When one of the world's best AI engineers says he feels behind... you should probably pay attention.
OpenAI co-founder Andrej Karpathy recently admitted:
"I've never felt more behind as a programmer."
Not because he's getting worse.
Because AI is changing faster than anyone can comfortably keep up with.
The biggest shift isn't "vibe coding."
It's how software gets built.
Your experience.
Your judgment.
Your taste.
Those become the advantage.
AI agents handle more of the execution.
That's the direction Karpathy sees engineering moving.
Not humans vs. AI.
Humans designing systems while AI does more of the work.
The interview is worth watching.
The article below explains why the best AI engineers are already moving from prompts to loops.
Pure retweeted
AI agents won't replace your work overnight.
They'll earn your trust one task at a time.
In a recent interview, Sundar Pichai explained why Google is introducing AI agents through products like Gmail and Calendar first.
Not because scheduling meetings is the ultimate goal.
Because that's where people learn to trust the system.
Once an agent reliably handles the small things, you'll naturally let it take on bigger ones.
That's exactly how every major technology gets adopted.
First it removes the boring work.
Then it becomes part of your workflow.
Finally, you stop thinking about it altogether.
We're already entering that transition.
The biggest shift isn't moving from prompts to better prompts.
It's moving from asking AI to do one task... to building systems that keep working long after you've stopped asking.
The interview explains where Google thinks this is going.
The article below explains how AI loops make it possible.
AI agents won't replace your work overnight.
They'll earn your trust one task at a time.
In a recent interview, Sundar Pichai explained why Google is introducing AI agents through products like Gmail and Calendar first.
Not because scheduling meetings is the ultimate goal.
Because that's where people learn to trust the system.
Once an agent reliably handles the small things, you'll naturally let it take on bigger ones.
That's exactly how every major technology gets adopted.
First it removes the boring work.
Then it becomes part of your workflow.
Finally, you stop thinking about it altogether.
We're already entering that transition.
The biggest shift isn't moving from prompts to better prompts.
It's moving from asking AI to do one task... to building systems that keep working long after you've stopped asking.
The interview explains where Google thinks this is going.
The article below explains how AI loops make it possible.
Pure retweeted
Right now, most developers are still doing the job AI should be doing.
Assign a task.
Wait.
Review the result.
Fix it.
Repeat.
You're not building with AI.
You're babysitting it.
The shift happening now is much bigger than better coding agents.
It's loop engineering.
Instead of managing every AI agent yourself, you build a system that manages them for you.
One loop assigns the work.
Another checks the output.
A third retries failures.
Everything keeps moving without waiting for your next prompt.
Your job doesn't disappear.
It changes.
You stop acting like a project manager...
...and start designing the system that manages the entire team.
That's exactly the transition I break down in the article below.
Right now, most developers are still doing the job AI should be doing.
Assign a task.
Wait.
Review the result.
Fix it.
Repeat.
You're not building with AI.
You're babysitting it.
The shift happening now is much bigger than better coding agents.
It's loop engineering.
Instead of managing every AI agent yourself, you build a system that manages them for you.
One loop assigns the work.
Another checks the output.
A third retries failures.
Everything keeps moving without waiting for your next prompt.
Your job doesn't disappear.
It changes.
You stop acting like a project manager...
...and start designing the system that manages the entire team.
That's exactly the transition I break down in the article below.
Pure retweeted
The fastest way to improve Claude isn't writing a better prompt.
It's giving it a better example.
Most people spend their time describing how they want AI to sound.
"Professional but friendly."
"Casual, but not too casual."
The result usually sounds like everyone else.
Instead, paste 2–3 paragraphs you've already written and tell Claude:
"Match this style."
Suddenly it picks up your rhythm.
Your vocabulary.
Your sentence structure.
The words you naturally use—and the ones you never would.
That's because AI learns patterns far better than abstract descriptions.
The same idea applies far beyond writing.
The best AI systems don't just follow instructions.
They learn from examples.
That's exactly the mindset behind the article below.
The fastest way to improve Claude isn't writing a better prompt.
It's giving it a better example.
Most people spend their time describing how they want AI to sound.
"Professional but friendly."
"Casual, but not too casual."
The result usually sounds like everyone else.
Instead, paste 2–3 paragraphs you've already written and tell Claude:
"Match this style."
Suddenly it picks up your rhythm.
Your vocabulary.
Your sentence structure.
The words you naturally use—and the ones you never would.
That's because AI learns patterns far better than abstract descriptions.
The same idea applies far beyond writing.
The best AI systems don't just follow instructions.
They learn from examples.
That's exactly the mindset behind the article below.
Pure retweeted
The people making the most with AI aren't doing more work.
They're building systems that do the work for them.
One creator generated $34,379 in just 3 months from a YouTube channel that takes about one hour a day to manage.
No filming.
No editing.
No sitting in ChatGPT writing prompts all day.
Instead, he built a workflow.
Claude researches the niche.
Gemini digs deeper.
Creators turn the ideas into videos.
His only job is making sure the system keeps moving.
That's the biggest shift.
You're no longer trying to become faster at every task.
You're designing loops that connect every task together.
The most valuable skill isn't creating content anymore.
It's building the system that creates it.
The video shows one real example.
The article below explains how to build the same kind of loops yourself.
The people making the most with AI aren't doing more work.
They're building systems that do the work for them.
One creator generated $34,379 in just 3 months from a YouTube channel that takes about one hour a day to manage.
No filming.
No editing.
No sitting in ChatGPT writing prompts all day.
Instead, he built a workflow.
Claude researches the niche.
Gemini digs deeper.
Creators turn the ideas into videos.
His only job is making sure the system keeps moving.
That's the biggest shift.
You're no longer trying to become faster at every task.
You're designing loops that connect every task together.
The most valuable skill isn't creating content anymore.
It's building the system that creates it.
The video shows one real example.
The article below explains how to build the same kind of loops yourself.
Pure retweeted
The biggest shift in AI isn't that it writes better code.
It's that you no longer have to ask it.
Boris Cherny and Cat Wu, the team behind Claude Code, recently revealed how they build software at Anthropic.
Instead of opening Claude and assigning every task manually, they let Claude Tag jump in automatically whenever it's needed.
It reviews code.
Fixes issues.
Runs for days.
And often opens the pull request before anyone even asks.
That's a completely different workflow.
AI isn't waiting for instructions anymore.
It's becoming part of the development process itself.
Anthropic says it now writes around 65% of their team's code.
The real breakthrough isn't the number.
It's moving from using AI... to working alongside AI.
The talk shows what that looks like.
The article below shows how to build the same kind of loops yourself.
The biggest shift in AI isn't that it writes better code.
It's that you no longer have to ask it.
Boris Cherny and Cat Wu, the team behind Claude Code, recently revealed how they build software at Anthropic.
Instead of opening Claude and assigning every task manually, they let Claude Tag jump in automatically whenever it's needed.
It reviews code.
Fixes issues.
Runs for days.
And often opens the pull request before anyone even asks.
That's a completely different workflow.
AI isn't waiting for instructions anymore.
It's becoming part of the development process itself.
Anthropic says it now writes around 65% of their team's code.
The real breakthrough isn't the number.
It's moving from using AI... to working alongside AI.
The talk shows what that looks like.
The article below shows how to build the same kind of loops yourself.
Pure retweeted
One developer spent 10 months building what most people are still trying to prompt.
After winning the Anthropic × Forum Ventures hackathon, Affaan Mustafa decided not to keep his workflow private.
He open-sourced the entire system for free.
Not just a collection of prompts.
A complete engineering setup.
Specialized AI agents.
Shared memory.
Built-in reviewers.
Planning.
Code review.
Long-term context.
Everything working together like an actual software team instead of one chatbot.
That's the biggest lesson from this project.
The advantage was never Claude itself.
It was the system built around Claude.
The video shows what that system looks like.
The article below explains why it works.















