Subham

Founder of The Psychepreneur — Pattern Recognition × AI Authority for digital identity, AI visibility, and LLM recognition.

X = My crypto journal so far

[Subham @iamsubham067]

From the 88-Day Hold Challenge to The Black Swan Event.

This is not financial advice.
This is a psychological autopsy.
I made this account public again because I want the receipts visible.

For 88 days, I documented what it feels like to hold through a full meme coin bleed.

Not the motivational version.

The real one.

The part where your conviction gets tested after the dopamine is gone.

The part where the timeline moves on.

The part where everyone who was loud during the pump disappears during the drawdown.

The part where you realize that “diamond hands” sounds fun until your portfolio becomes a graveyard.

I have been watching Solana meme coins closely since June 2024.

Not casually.

Obsessively.

I saw narratives form before they became obvious.

I watched attention rotate.

I watched communities build, collapse, revive, and vanish.

I saw how fast a coin could go from invisible to cult-like.

I saw how fast the same coin could go from cult-like to forgotten.

I caught early waves across names like:

$FWOG
$GIGA
$PORK
$LOOK
$MANYU
$CUPSEY
$TROLL
$NEET
$RETIRE
$67
Some gave upside.
Some taught lessons.
Some were pure chaos.
But all of them taught me one thing:
Markets are not just charts. Markets are people under pressure.

The 88-Day Conviction Challenge started because I wanted to test something.

Not just the coins.

Myself.

Could I hold when the thesis was no longer emotionally rewarding?

Could I stay with a position when the timeline was laughing at it?

Could I document the bleed instead of pretending I was unaffected?

Could I sit inside uncertainty without constantly needing a new dopamine hit?

That was the real game.

The coins were just the surface.

The main bags were:

$TROLL
$NEET
$CUPSEY
$RETIRE
These were not random tickers to me.
They came from months of watching meme coin behavior, community energy, narrative formation, timing, rotations, and market psychology.
I had seen the upside.

I had seen the early signs.

I had seen how fast attention could turn a joke into a movement.

So I held.

Through pumps.
Through dumps.
Through silence.
Through cope.
Through boredom.
Through humiliation.
Through the slow death of attention.

88 days.

Documented.

And the result?
Brutal.
The challenge ended deep in the red.
A full bleed.
No cinematic comeback.
No perfect exit.
No fake guru screenshot.

Just a hard lesson:

Being early is not enough.

You can be early and still lose.
You can read the trend and still mistime the cycle.
You can understand the narrative and still get punished by liquidity.
You can have conviction and still be wrong in execution.

That is the part most people never admit.

The market taught me something very expensive:

Conviction without risk management becomes ego.

That line changed everything for me.

Because in crypto, especially memes, conviction is praised when the chart goes up.

But when the chart bleeds, the same conviction starts asking you deeper questions:

Are you holding because the thesis is alive?

Or because your identity is now attached to being right?

Are you patient?
Or are you trapped?
Are you disciplined?
Or are you addicted to proving the crowd wrong?

That is where the real test begins.

The 88 days were not just about $TROLL, $NEET, $CUPSEY, and $RETIRE.

They were about the psychology of holding through uncertainty.

The loneliness of being early.

The danger of confusing signal with hope.

The cost of building conviction without an operating system.

The emotional violence of watching something you believed in slowly lose attention.

That is why I documented it.

Because most people only post the entry and the win.

Very few post the bleed.

Since June 2024, I have seen the entire meme cycle evolve.

I saw how fast Solana became the casino, the culture layer, the attention machine.

I watched builders, degens, influencers, founders, communities, insiders, outsiders, and tourists all collide in the same arena.

I watched memes become financial instruments.

I watched narratives become liquidity magnets.

I watched people confuse community with exit liquidity.

I watched myself do parts of the same.

That is why I don’t talk about this like a clean success story.

It wasn’t clean.

It was expensive.

But it was real.

And honestly, that is why I still respect the game.

Because meme coins are not just stupidity.

They are compressed human behavior.

Greed.
Fear.
Status.
Belief.
Timing.
Tribalism.
Attention.
Delusion.
Reflexivity.
Narrative warfare.

Everything happens faster there.

You see people become philosophers on the way down and geniuses on the way up.

You see conviction turn into religion.

You see liquidity expose character.

You see who was building, who was gambling, and who was only loud because the candle was green.

The 88-Day Hold Challenge was my way of saying:

I was here.

I saw it.

I held through it.

I didn’t just appear after the next hype cycle.

I took the hit in public.

And I learned the real lesson.
Not “never take risk.”
That is a coward’s lesson.
The real lesson is:
Risk needs structure. Conviction needs execution. Belief needs systems.
Without that, you are not investing.
You are just emotionally negotiating with volatility.
So this is the chapter I’m reposting.
Not because I’m proud of being down.
But because I’m proud I can look at it clearly.

I played the meme cycle.

I caught waves.

I held through the bleed.

I documented 88 days of conviction.

And I came out with a sharper understanding of markets, psychology, attention, and myself.

The next phase is not random.

The next phase is systems.

Product.
AI.
Automation.
Authority.
Execution.

Same conviction.

Better operating system.

If you were there during the 88 days, you know.

If you’re seeing this now, understand the context:

This was not a flex.

This was not a signal group.

This was not a guru arc.

This was a public record of what happens when conviction meets volatility.

The market can humble you.

But if you learn properly, it can also forge you.

From the bleed to the build.

From meme cycles to systems.

From conviction to execution.

LOCK IN.

[Subham @iamsubham067]

Day 1/88:
The explosive kickoff to the Solana meme coin redemption saga! 💥🪦

Wrapped Q4 2025 with the bag obliterated ~88% from those September highs, pure carnage from pump to dump.

No jeets, no mercy, just savage HODL wisdom from the cosmos:
- Until I abandon my old self...
- No crypto luck is gonna "retire" me.

Gotta rise up and grind it out for myself. Prices sitting at rock bottom:
- $retire $0.001608 (-12.33%)
- $67 $0.01 (+16.66%)
- $Cupsey $0.000051 (-9.80%)
- $NEET $0.008169 (+13.78%)
- $TROLL $0.02 (-4.80%)

Bags locked since Q4 '24 ($retire, $Cupsey), Q2 '25 ($TROLL, $NEET), and $67 fresh off the press.

We rode the highs, tanked the lows. Moon or zero? We hit zero this cycle, but with it already 88% nuked...

I don't give a fuck.

Locking in through Q1 2026, and this round, I'm sacrificing it all.

Every procrastination habit, every wait-for-a-miracle crutch, every black swan delusion.

Gone.

Who's still holding?

[Subham @iamsubham067]

At the time, my résumé needed credentials to validate work I wasn’t yet "paper-validated" enough to demonstrate publicly.

But now, that has changed.

Building in public has forced me to replace a lot of “trust me” with visible proof.

I've realised what kind of work gives me energy.

- Give me an ambiguous problem and enough room to think.
- Give me something that normally requires several functions to coordinate, and let me figure out how much of it can be compressed.
- Give me responsibility for an outcome, a product surface, or a business unit, and hold me accountable for its performance.

That is where I do my best work.

I also know the environments where I struggle.
- Environments where tools are rejected simply because they are new.
- Where using AI is treated as a shortcut instead of another capability to understand and govern properly.
- Where execution is prescribed so tightly that there is very little room left to investigate the actual problem.

I don’t need complete freedom.

Good product work requires constraints, disagreement, review, specialists and strong teams.
But I do need ownership.

I want to be around people who care less about how traditionally the work was done and more about whether the thinking was rigorous, the risks were understood, and the outcome moved.

I spent years trying to collect enough credentials to look ready. These days, I would rather build enough proof that the question becomes unnecessary.

[Subham @iamsubham067]

The last time I took a premium course I didn't need.

why because my resume needed paper validation for the work I wasn't ready to do in public?

[Subham @iamsubham067]

Trust me, another kinda God mode hits you when you wake up at 3am or early af in the morning.

[Subham @iamsubham067]

This thesis is built on experiments I’ve already shared publicly:

1. Subham Ojha entity-resolution experiment
Reducing entropy around my digital identity helped newer reasoning models map my name correctly to my real work and The Psychepreneur.
https://t.co/Xs2FSD44vs

2. The Psychepreneur experiment
I tested ChatGPT, Grok, Meta AI and Gemini to see whether they could resolve The Psychepreneur back to me: without PR, SEO agencies or a team.
https://t.co/mC0nyVFiDP

The content strategy itself was never “expert content.” It was an evergreen Psyche content machine built through consistency and repeated intellectual territory.

https://t.co/e4HiKSq9Ex

3. AI Authority Playbook
https://t.co/b8CWd6l7Lm

That thinking became Conviction OS: infrastructure for keeping multi-channel, multi-account positioning and identity consistent.
https://t.co/AVIyDghJ3u

If you want a Reputation Engineering analysis:
https://t.co/O7PbFbbYtE

The core idea: content creates signals, websites define entities, reviews and citations reinforce them, and AI systems connect the dots.

That is how a reputation graph starts forming.

[Subham @iamsubham067]

Reputation Engineering is the new infrastructure of marketing.

I’ve been saying versions of this since December 2025.
- I never followed a trend.
- I never had a GEO playbook.
- I never studied what marketers were saying about AI search.

I just noticed something much simpler:
AI systems had started citing me last year. Before this was even a wave, I was there. That was enough to make me investigate.

I started testing how different models understood my name, my work, my products, my services, and the companies and projects connected to me.
And that changed the way I thought about marketing.

Because underneath search, ads, content, social media and discovery, there is increasingly another layer:
- What does the machine believe about you?
That is reputation engineering.

I was early to this space because I was studying the behavior before I cared about the terminology.

Since late 2025, I’ve been consistently publishing, building, documenting, testing and attaching my real identity to real work.
- No backlink campaign.
- No PR machine.

Yet today, I can find myself, and my work represented or cited across 8+ LLM environments.

And one thing became increasingly obvious through my own experiments:
- Consistency is massively underrated.

Before a machine can trust you, it first needs enough coherent evidence to understand you.
- Who are you?
- What do you consistently talk about?
- What have you actually built?
- What problems have you solved?
- What entities are associated with you?
- Do the claims on your website match what exists elsewhere?
- Are there real names behind the expertise?
- Are there case studies?
- Products?
- Research?
- Reviews?
- Public evidence?

I learnt AI visibility is just a reputation system.

A reputation system has multiple layers:
- Content that establishes what you believe.
- A website that clearly defines who you are.
- Real people attached to real expertise.
- Products and services connected consistently to the same entity.
- Case studies
- Reviews
- Original research giving other people a reason to reference you.
- Third-party citations reinforcing the claims.
And consistency across all of it.

That last part matters more than most people realize.
Cz when hundreds of independent signals repeatedly tell machines the same story about an entity, something more durable starts forming.

A reputation graph.

That is what I’ve been unintentionally, and later intentionally, building since 2025.

I never optimized my work around whatever the marketing industry happened to call the trend that month.
I kept producing evidence.

And eventually the systems began connecting those dots.

The next evolution of marketing is larger than SEO.
"Reputation Engineering is the infrastructure underneath them."

That is the game I’ve been studying since 2025.
And I think we are still extremely early.

I’ll add some of my older posts and experiments in the comments.

#psychepreneur #subhamojha #aivisibility #reputationengineering

[Subham @iamsubham067]

and this time when life gives me a chance, I'll not rush.

i'll just do what I'm good at.

I'll not care about investing, I'll not care about wealth creation. I'll just keep doing what I love. And I'll see where I go.

[Subham @iamsubham067]

One thing I’ve learned from working across ecommerce, growth, product, automation and now AI:
You don’t need to know every stack deeply to build useful systems.

You need to get good at:
→ understanding the problem
→ structuring context
→ breaking it into components
→ defining constraints
→ testing aggressively
→ actually deploying

A few builds if useful:
- Build 3D urban digital twin → https://t.co/j83XmQjgei
- Ecommerce revenue reconciliation problem solved → https://t.co/SABFvTaQQY
- Deploy AI Agents that work while you sleep -> OpenClaw on Oracle Cloud → https://t.co/hdBxwTUQTs
- Predict Future outcomes using AI -> MiroFish self-hosted → https://t.co/iLGUxVMogp
- Automate social media without getting banned -> Multi-user n8n automation → https://t.co/UwNCpVyA9Z

Different problems. Different domains.
The transferable skill I’m increasingly interested in is learning how to enter an unfamiliar system, understand it fast enough, and ship something useful.

I’ve spent years moving across problems: 40+ ecommerce builds from 0, dashboards, APIs, AEO/SEO/growth systems, automation and enterprise products.

Recently, I’ve been deliberately testing the same thinking with AI-assisted engineering.0

[Subham @iamsubham067]

I'm working to be niche-agnostic.

Specialization does matter.

But
- A CFO will understand finance at a depth I probably won't.
- A CTO will understand technology at a depth that comes from decades of building.
- A CEO will have lived through business decisions I have only studied or observed or maybe won't know at this stage of my journey.
I respect that.

What interests me is something slightly different.
I want to become useful at the intersections.
- Technology × business.
- AI × human behavior.
- Product × distribution.
- Psychology × decision-making.
- Systems × execution.
The kind of problems that don't neatly belong to one department or one job title.

That's also how I'm trying to approach my work outside a conventional job description: learn enough to understand the system, ask better questions, build something tangible, and leave behind proof that the problem moved forward.

I don't know what the eventual label for that will be.
Maybe I don't need one.

For now, I'm simply trying to build a body of work where, irrespective of the domain, people can trust the quality of thinking and execution I bring to the table.

A long way to go.

But that's the direction I'm deliberately building towards.

It was supposed to become a way of thinking across them.
And today, I'm testing that philosophy in a domain: sanctions-screening infrastructure.

I'll be building and shipping ScreenPath next.
It's a happy Sunday. :)

Stay tuned for the demo.

#psychepreneur #subhamojha

[Subham @iamsubham067]

100% social media automation & monetization with AI idea is 100% lame.

Social media companies want real human on their apps, not your Hermes Agent.

Although you can sometimes schedule posts , and all, but not automate everything 100%, whoever buys to that idea. Plz dont

[Subham @iamsubham067]

One iPhone 17 Pro Max or 26 months of Claude + ChatGPT + Grok

[Subham @iamsubham067]

Courses won’t save you. Doing will.

To make the post concrete, here are five systems I’ve taken from idea → terminal execution → testing → deployment:

1. OpenClaw on Oracle Cloud
Server setup, configuration, deployment and operation:
https://t.co/hdBxwTUQTs⁠

2. MiroFish self-hosted deployment
From repository setup to a working deployment:
https://t.co/iLGUxVMogp⁠

3. Multi-user content automation with n8n
Onboarding, editing, approvals and automated publishing across multiple accounts and channels:
https://t.co/UwNCpVyA9Z⁠

4. Urban-city digital twin
Built using Fable 5 and Codex in the terminal, with agents supporting research, development, design and testing:
https://t.co/j83XmQjgei⁠

5. Ecommerce revenue-reconciliation system
Evidence-backed findings, audit trails and human-in-the-loop review:
https://t.co/SABFvTaQQY⁠

These projects taught me the same lesson:
The advantage is not generating more code.

It is giving AI the right product context, architecture, reusable workflows, constraints and verification standards, then pushing the result beyond localhost.

Comment SKILL for the starter repository structure.

Comment PROJECT for five practical AI projects you can start building this week.

[Subham @iamsubham067]

They are five attempts to understand AI by building, breaking, deploying and improving real systems.

Not by wasting time on 3-8 hours tutorials.

[Subham @iamsubham067]

How are you keeping up with AI?

I don't watch tutorials to do it. I do it by building.

Here's 5 things you can build and Deploy step by step without depending on tutorials EVER.

[Subham @iamsubham067]

5. Ecommerce revenue-reconciliation system with human-in-the-loop review - ⁠https://t.co/TVD9hww7iT

[Subham @iamsubham067]

Comment PROJECT and I’ll send you five practical AI project ideas you can start building this week.

5 things I have already built and deployed.
Need the complete deployment guide? Interact with this post and comment GUIDE. I’ll send it to you.

1. OpenClaw on Oracle Cloud — end-to-end deployment - https://t.co/hdBxwTUQTs

From creating the server and configuring the environment to deploying and operating the application.

⁠

2. MiroFish — complete self-hosted deployment - https://t.co/iLGUxVMogp

An end-to-end walkthrough covering the journey from repository to a working deployment.

3. Multi-channel content automation using n8n - https://t.co/UwNCpVyA9Z

A multi-user, multi-account workflow for onboarding, editing, approval and automated publishing across multiple social platforms.

4. Urban-city digital twin from idea to deployment - https://t.co/j83XmQjgei

Built using Fable 5 and Codex inside the terminal, with multiple agents orchestrated across research, development, design and testing.
⁠

5. Ecommerce revenue-reconciliation system with human-in-the-loop review - ⁠https://t.co/SABFvTaQQY

They are five attempts to understand AI by building, breaking, deploying and improving real systems.

[Subham @iamsubham067]

Want to keep up with AI? Stop watching tutorials all day. Start BUILDING.

When I hear about something like OpenClaw / Hermes / some repo going viral I do not immediately search for ten explanation videos.
I try to deploy it directly.

You will learn more from one failed deployment than from another three-hour “complete AI agents” tutorial.

( Complete Thread Below )

[Subham @iamsubham067]

4. Build with real-world constraints

Connect it to something real:
- A database.
- An inbox.
- A webhook.
- A document repository.
- A scheduled job.
- A public API.
- A human approval queue.

That's how you keep up.

#psychepreneur #subhamojha #ai #2026