TenureOS

// 1984 — PRESENT  ·  TALLAHASSEE, FL

FROM
COMMODORE 64
TO AI

Four decades of curiosity, building, breaking — and building better. A personal journey through the most transformative era in human history.

SCROLL
1984 – 1985
01

THE MACHINE
THAT STARTED IT ALL

It was 1984. I was ten — maybe eleven — years old when a Commodore 64 came into the house. It was the size of a small suitcase, and I couldn't put it down. Not then. Not ever.

This wasn't the internet age. There were no tutorials, no Stack Overflow, no YouTube walkthroughs at two in the morning. But there was Telnet, there were early online forums, and somewhere in my city there was a community center computer club where a kid my age could walk in on a Saturday and trade 5.25" floppy disks with grown men twice his age who were just as obsessed.

"My parents would drop me off at the local computer club. A bunch of computer guys, much older than me. We'd share floppies, copy software, talk hardware. Nobody was teaching us. We were just… building."

64 KB of RAM
1 MHz Clock Speed
170 KB Floppy Capacity
17M Units Sold
LATE 1980s
02

WE BUILT THE
MACHINES

By high school, the floppy disk clubs were behind us. Now it was about building the hardware itself. Intel 8086 chips. Hard drives with maybe 4 MB of storage. Slow, clunky, unglamorous — and completely captivating.

We tried the C++ books. The code manuals. The dense, unforgiving textbooks that treated programming like a punishment. They never clicked. The itch to code was real, but the resources of the day weren't built for the way my brain worked. I dropped the class after four weeks — but I never lost the itch.

The OS Revolution — Witnessed in Real Time

From MSDOS prompts to Windows 95 arriving like a cultural event — graduating from FSU meant graduating from the command line into a GUI world. The terminals went dark. The icons appeared. Everything changed.

8086 Intel Processor
4 MB Hard Drive
640 KB RAM ("Enough")
LATE 1990s – 2000
03

THE GREAT
DIGITAL LEAP

First job out of college: real estate appraisal. Days spent appraising property. Nights and weekends spent as the unofficial network engineer for the entire firm. Nobody asked. Nobody needed to. It just needed doing, and I was the one who could do it.

This was one of the most extraordinary periods of productivity acceleration in history — and I had a front-row seat.

Before

🎞 Roll film → 1-hour photo lab for property shots
📋 Scratch-off pages to identify comps
🗂 Courthouse microfiche for deeds & mortgages
📼 Cassette tape server backups, rack by rack

After

📷 Digital cameras — implemented in a single week
🗺 Digital maps & first-in-industry comp database
💻 Windows 2000 server — built and maintained solo
🎮 After-hours Counter-Strike & Quake server

"We were so behind the times — and then, almost overnight, we weren't. That week I implemented digital cameras, everything we'd done before looked like archaeology."

MAY 4, 2000
04

THE DAY THE
INTERNET GOT SCARY

May 4, 2000. The ILOVEYOU worm launched from the Philippines and circled the globe in hours — one of the first true cyberattacks at planetary scale. An estimated 10 million systems infected. Roughly $10 billion in damage. Twenty-four hours, wall to wall.

At the office, the same partner got caught by the same trick twice in the same week. The same email. The same attachment. The same crash.

Thank God for those cassette tape backups. Those weren't just paranoid IT overhead anymore. They were the lifeline. That week changed how everyone thought about security — and for those of us who already lived inside these machines, it meant something had shifted permanently.

10M+ Systems Infected
$10B Estimated Damage
24hrs Global Spread Time
~2 YEARS AGO
05

THE CODING ITCH
RETURNS

By 2011 I joined the State Board of Administration of Florida, managing a real estate portfolio that would grow from roughly $4 billion to $20 billion over the next decade. And somewhere in that decade, a familiar frustration set in.

We were using Excel as a database. Off-the-shelf software that bloated, compromised, settled — and never actually did what you needed it to do. The friction was becoming untenable.

Then Harvard dropped something extraordinary — free, on YouTube — called CS50. Introduction to Computer Science. And this time, the curriculum was different. Python. SQL. Flask. Algorithms. This time, it clicked.

CS50x — Harvard's Free Computer Science Course

Taught by David Malan, CS50 covers Scratch, C, Python, SQL, HTML/CSS/JavaScript, and Flask, plus foundational concepts in algorithms, data structures, and cybersecurity — available free on YouTube for anyone with the curiosity and the time.

2022 – PRESENT
06

AND THEN
AI HAPPENED

While I was working through CS50, OpenAI dropped their first GPT. It was slow. It did some curious, impressive little tricks. Nobody fully understood what it was going to become — including the people who built it.

That version looks almost quaint now. Because today, we're living in something else entirely.

THE CURRENT STACK

Four subscriptions. Four distinct toolsets. One singular mission.

OpenAI / ChatGPT

The pioneer. Broad capability, deep integrations, the platform that changed the public conversation about what AI could be.

Google Gemini

Deep research capability and native integration across Google's data ecosystem. A different lens on the same problems.

xAI Grok

Real-time data, X platform access, and an irreverent edge that makes it useful for research no one else is willing to touch.

Anthropic / Claude

Long context, careful reasoning, and an extraordinary ability to write production-grade code — the architect in the stack.

Used individually, each is remarkable. Used together, communicating across agents and subagents, coordinated with intention — they provide direction and utility that would have been science fiction five years ago. Not to me personally. To everyone. To humanity.

"It's the most profound thing I have experienced in my 54 years."

NOW
07

BUILDING
REPI

The friction is finally gone. Not reduced — gone. After decades of working around bloated, off-the-shelf software that always compromised, always made you settle, we now have something extraordinary: the ability to describe your problem in plain language and have it built for you in pure, beautiful code.

That's what REPI — Real Estate Portfolio Intelligence — is. A late-stage hail Mary turned living system. Built from scratch. Built right. No vendor. No compromise. No settling.

# REPI — Real Estate Portfolio Intelligence
# State Board of Administration of Florida

REPI_ONE        → Flask API · portfolio data engine
REPI_Dashboard   → Plotly Dash · live analytics interface
OpenClaw        → AI agent orchestrator · Mac Mini M4
@MarinoFSUBot    → Telegram bot · mobile command layer

# Portfolio: $4B → $20B+ under management
# Status: RUNNING

If you have a little imagination and just a little bit of time, you can literally build what your organization needs — exactly the way it should work. No compromises. No more suffering through the worst of what enterprise software offers. That era is over.

THE PEOPLE BEHIND IT
08

THE WORK
NOBODY SEES

Here's what the story leaves out if you're not careful: none of it was possible alone.

While REPI was being built, our analyst and Senior Investment Officer — two of the best this team has ever had — spent six months and countless hours doing the work that no one photographs and no one puts in a press release. Fixing Excel errors. Chasing down missing data entries. Literally rebuilding the entire back-of-house data infrastructure from the ground up.

"The stuff that nobody sees. The hard work. The stuff our introverts do and never really get credit for."

They were the motivation. Not the technology — the people. They are the reason this project got done. The architecture, the code, the dashboards — those are visible. What they built is invisible in the best possible way: a clean, trustworthy foundation that makes everything else work.

A Note on Credit

The most important work in any data transformation isn't the front-end that people see — it's the unglamorous, meticulous, thankless work of making the underlying data true. That's where real analysts earn their place in the story. And they earned it here.

THE NEXT 5 YEARS
09

FASTEN YOUR
SEATBELTS

There are two roads ahead. And the choice between them is real.

The Old Road

💸 Continue spending millions on bloated SaaS
📋 Consultants who deliver decks, not solutions
⏳ Wait for vendors to build what you actually need
🔒 Talented people blocked by the wrong tools

The New Road

🤖 A real AI-focused task force — internal talent
🛠 Empowered builders with the right tools
🏗 Own the architecture. Own the data. Own the output.
🚀 Describe the problem. Build the solution. Ship it.

The brightest people are already inside the building. They just need to be empowered and given the tools to develop. That's not a technology problem. That's a leadership decision.

Either way — whether organizations embrace this and build a real roadmap, or continue spending millions on consultants and SaaS that almost-does-what-you-need — the next five years are going to be like nothing we have seen.

FASTEN YOUR SEATBELTS.

From a Commodore 64 in 1984 to AI agents building production software in 2024 — if the last 40 years taught us anything, it's that the people who stay curious, stay in the room, and keep building are the ones who get to shape what comes next.

THE STORY
ISN'T
OVER.

From a Commodore 64 in 1984 to AI-powered portfolio analytics running on a Mac Mini in Tallahassee — the itch to understand and build has never gone away.

It just keeps getting more interesting.

THE ARC

1984

Commodore 64

First computer. Telnet, early forums, the community center computer club. Age 10. Couldn't put it down.

1980s

Building the First PCs

Intel 8086, 4MB hard drives, MSDOS terminals. Building machines from parts before build guides existed.

1990s

The Great Digital Leap

Real estate appraisal by day, network engineer by night. Digital cameras, comp databases, Windows 2000. First-in-industry implementations.

2000

ILOVEYOU Worm

May 4th. The day security became everyone's problem. Tape backups saved the day. Twice.

2011

SBA of Florida

Director of Acquisitions. Managing a real estate portfolio that would grow from $4B to $20B+.

CS50

Harvard Free · Python · The Itch Returns

CS50 on YouTube finally made code click. Python. SQL. Flask. The foundation REPI would be built on.

2022+

The AI Era

GPT drops during CS50. The world shifts. ChatGPT, Gemini, Grok, Claude — four subscriptions, one vision. REPI goes live.

NOW

REPI — Real Estate Portfolio Intelligence

Flask · Plotly Dash · OpenClaw AI Orchestrator · Telegram Bot. No compromises. No settling. Built exactly right.