Category: Uncategorized

The build

I needed a personal assistant that could do things Claude, Claude CoWork, and Claude Code simply couldn’t. They kept hitting walls — permissions, file access, raw capability. Every time I wanted something real, I ran into the edge of what they’re allowed and able to do.

So I built my own, Rouh, to cover everything those tools left on the table.

Then I wired in a feedback loop. A single sentence of like or dislike becomes a feature request, gets built, deployed, and goes live in a couple of minutes. I say “I wish it did X,” and before I’ve finished my coffee, it does X.

As that loop tightened to its limit, I felt the power of it. The gap between wanting something and having it running had almost vanished. That’s a strange thing to feel in your hands — desire and reality collapsing into the same moment.

Letting it loose

The next move was obvious and unnerving. What if I stopped driving it and let it study its own code, its usage, my integrations, my emails, my calendar, my recent events — and decide for itself what to change to improve my life?

I set it up as a nightly run. It builds overnight while I sleep, not during work hours.

It works. Each morning it has genuinely come up with useful features and deployed them on its own.

Every deployed feature sits behind a feature flag. I can turn it off, trial it, modify it, review it, or just leave it running. That’s the one thing standing between me and chaos.

The assistant started changing in front of my eyes. Evolving continuously, right at the limit of what I can comprehend and comfortably use.

I could push it to evolve even faster, but past a point it becomes too difficult to work with. My ability to keep up is the real cap — not the machine’s.

Even so, I could do work many times faster and more capably than with anything I’d used before. Well beyond CoWork and Claude Code.

Living with it

It becomes a bit like a living thing you cooperate with but don’t fully control, because you can no longer track all of its functionality. There are features in there I didn’t ask for and don’t fully remember approving.

People don’t come with an inbox-list of their thoughts or a settings page of who they are. You talk to them, get to know them little by little, and build trust. That’s exactly what living with Rouh feels like.

You end up in a relationship with an avatar that represents you in a world more capable than you are yourself — living on the edge of your ability to control something this powerful.

Naming the pattern

This isn’t a quirk of my setup. It’s a pattern worth naming. I call it Continuous Co-Evolution.

At its core it means software that evolves on its own.

It’s co-evolution because as the software evolves, its owner — a person or a company — evolves too, becoming more capable and more efficient alongside it.

It’s continuous because once the setup is right, the change in both the software and the person never stops.

The software is set up to examine everything happening on it: its own code, its usage, the logs, the data. From that examination it generates ideas to improve itself and the situation, triages them, and builds the features.

The most impressive version runs without any oversight, though it can also be done with oversight. Running without a human in the loop, the software can be improving by the hour or by the day, adding efficiency with every upgrade it ships itself.

Into the enterprise

My personal version is honestly reckless. No oversight means it can lead to security compromise, and it’s purely experimental. But strip out the recklessness and the same engine becomes a serious asset.

The personal example is a foundation others can translate into enterprise needs — a custom-built CRM, for instance.

At BigBongo, my company, we build custom software so clients own the codebase, the process, the outcomes, and the data. Owning the codebase is what lets a company embed continuous co-evolution — and get CRM upgrades daily or hourly.

A self-evolving CRM would look at your people, your data, and its own codebase, know the goal, and optimize for it on its own. It would analyze your top performers to understand why they’re the best, help you pick better people, and lift everyone else. It would build features to compensate for weaker salespeople, feeding them exactly the information they need.

Enterprise requires heavy governance, privacy, and security — all of which can be built and provided. It needs testing, validation, and SDLC protocols, with every requirement passed before any feature deploys. It’s more complicated, takes more steps, and demands planning in advance. But it is absolutely possible, and it is the future.

What this becomes

This is the future of apps for everyone — enterprise, small clients, personal assistants, all of it.

Software will no longer be fixed, normal software. From now on it’s “live software” — always evolving, co-evolving with its owner.

Having built and lived in a relationship with a self-evolving tool, I’ve had a glimpse of that future. From there, today’s best tools — Claude, CoWork, Claude Code — already look a generation old. Not because they’re bad, but because they’re the infrastructure and building blocks, not the final product or the vision.

The real vision is bigger and stranger: a world where we co-live with AI creatures — living things made of different building blocks. At some point we may even see software with no owner at all. But that’s a different story.

I just realized Elon is the GOAT.

I realized how incredibly coherent, so ahead of us, doable, so brilliant Elon’s concept is of the future.

Here is my breakdown.

AI models today are so good they can do almost all tasks better than humans. We’ve reached this point already. Why does this matter? Speed, efficiency, quality and price will drive replacement of almost all cognitive work. It’s solved – just a matter of integration, which will grow faster than we expected. Elon happens to have one of the best LLMs with one of the best private data sources (Twitter) and massive data centers to power it. Not top notch like anthropic but let’s say frontier.

What’s left is physical labor, and he’s also been solving physical labor – transportation through robotaxis and general-purpose robots for everything else.

Nothing is accidental. To have robotaxis, you need real-world training. For training, you need cars with camera/sensor data. Tesla provided all this. Now autonomous vehicles are basically solved – mostly regulation left. Optimus is next, and Musk says Tesla will primarily be a robot company.

He’s not solving for cars or specific labor. He’s solving for everything.

Quick jump and I will connect the dots.

We can’t build a space civilization with humans – too dangerous, too expensive, too fragile. But we don’t need humans anymore. We’ve solved for both cognitive and physical labor. Robots are perfect for space by design.

Space throws radiation that kills humans – doesn’t affect silicon and metal. No oxygen needed. Extreme temperatures handled. Zero gravity that destroys human bones actually helps robots – less wear on joints, easier movement. They’re modular – swap parts, upgrade components, last decades instead of needing daily food and water.

Data centers in space make perfect sense now. You need massive computational power to coordinate thousands of robots building your space civilization. Space is cold – free cooling. Solar panels work 24/7 with no atmosphere – unlimited clean energy. No real estate costs, no regulations.

Those space data centers serve Earth too. SpaceX’s million satellites aren’t just internet – they’re distributed computing nodes. The same infrastructure coordinating space robots runs Earth’s AI operations. Twitter data feeds the AI managing both Earth factories and space construction.

Everything connects. Terafab iterates chips in days – straight to space infrastructure. Reusable rockets become routine supply runs. Optimus robots trained in Tesla factories learn skills for building in vacuum.

He’s not just going to space. He’s moving our entire industrial base where it can grow without limits.

Once you have this complete stack in space, something else becomes possible. Robots build the habitats, AI designs them. Perfect gravity through rotation, optimal atmosphere mix, controlled temperature and light – everything Earth got wrong, fixed by design.

But that’s just the start. AI improving at exponential rates, and that rate itself accelerating because AI makes the next AI faster. Extrapolate that trajectory and we start unlocking science that’s been pure fiction.

We’re already seeing hints – 3D printing organs, creating materials on demand, understanding biological systems at molecular level. Full control of physics means things like controlled fusion, maybe even teleportation. Complete understanding of biology means designing perfect organisms, controlling entire ecosystems. Not just copying Earth’s flora and fauna – improving them.

Up until now, Earth was our only choice because we lacked both understanding and control. Nature was better than anything we could build. But once AI pushes past a threshold, once we truly understand how matter and energy work at fundamental levels, we can engineer habitats that surpass Earth in ways we can’t imagine today.

The same way building in space with humans seemed impossible until we solved it with robots, creating better-than-Earth habitats seems impossible until AI unlocks the science. He’s not just building the infrastructure – he’s building the platform where these breakthroughs will happen. Away from Earth’s constraints, with unlimited computation and robot labor, discoveries that would take centuries happen in years.

I wonder how many other ideas in his head are already in motion that we don’t understand yet.

Sometimes, life feels like a puzzle where most of the pieces are scattered, and no matter how hard you try, something is just… off. Then suddenly, one key piece falls into place, and everything shifts. That’s exactly what happened with my brother.

For a long time, he was working hard but not feeling fulfilled. Then, he joined my company. It wasn’t just a job change—it was a structural shift in his life. He got to work on something exciting, set his own schedule, go to the gym in the morning, and most importantly, he wasn’t trapped in draining work. The difference was huge. His mood, motivation, and even his social life improved.

From the outside, it looked like everything just magically worked out. But it wasn’t magic. It was one key shift that unlocked everything else.

Finding the One Missing Piece

If you’re stuck, the real question is: What’s the missing piece for you? What’s that one shift that could make everything click?

The answer is surprisingly simple: turn toward what you love.

But there’s a catch. Loving something isn’t enough. You have to:

1. Do it a lot. Not just as a hobby but as something you dedicate time and effort to.
2. Get better at it. Push yourself. If you love drawing, don’t just doodle—learn new techniques. If you love marketing, study the best case studies and experiment. Improvement is key.
3. Do it for yourself. There’s a different energy when you’re working toward your own growth rather than just fulfilling someone else’s goals.

This is what changed things for my brother. It wasn’t just about joining my company. It was about aligning with something he was naturally good at (sales), doing it every day, improving, and doing it in a way that built something meaningful for himself.

You Don’t Always Have to Start the Train—You Can Hop On

One interesting realization is that you don’t always have to come up with your own grand vision. Sometimes, someone else’s train is already moving in the right direction, and if it excites you, you can hop on.

AI was my thing—I had been obsessed with it, pushing my knowledge, finding the best ways to explain it to people. But when my brother saw what I was doing, it resonated with him. He recognized that this was something he could contribute to and grow with.

The key is being in touch with yourself. If you chase things just because other people say they’re valuable—whether it’s a certain career, money, or status—you might find yourself stuck in a life that doesn’t actually fulfill you.

The Formula for Getting Something Real Going

To recap –

1. Find what you love – something you can be genuinely excited about.
2. Go deep – learn, practice, improve.
3. Have your own path – you will tap into your inner energy

It will naturally become valuable to someone. 

This isn’t a shortcut. It takes time. In my case, it took two years of diving deep into AI. In my brother’s case, it was two years of fully committing to sales. But once you hit that point where your effort compounds, things take off.

It’s not a complete formula for success, but it’s necessary and largely sufficient. If you do this, you’ll be in a strong position. If you don’t, things probably won’t click the way you want them to.