Execution at the speed of thought

I’ve always had a low tolerance for fads.

Superfoods, avocado toast, NFTs – they show up out of nowhere, carried by a loud minority claiming they’ll change the world. That kind of noise is usually the fastest way to turn me off. So, naturally, when AI started popping up in every conversation, pitch deck, and news article, I rolled my eyes.

To me, “We’re an AI-powered company” sounded no more meaningful than “We’re a Java first company built on the cloud.” Who cares what you’re built with? Tell me what you do.

Yes, I played around with the AI art filters that turned my face into something straight out of Studio Ghibli. Cute gimmick. But it didn’t move me. I remained unconvinced – until about 90 days ago. That’s when something shifted.

The Experiment

I hadn’t written a line of code since 2007 – pre-GitHub, deep in C++ land. But as we scaled Rabbit into Saudi Arabia, I found myself stuck. We had more ideas than engineering bandwidth. Prioritizing became a brutal game of resource triage: one pack of wet food, too many hungry pets.

So I asked for something ludicrous and quite characteristic. I asked the team for a copy of the codebase (yes, a copy – I didn’t know what a commit was) to see how far I would go on a week-long side quest with an AI side kick: Could I, a founder who hadn’t coded in over a decade, build something real?

I treated it like a puzzle – something to mess around with after hours. Once the team helped me set up the dev environment and explained Git like I was five, I jumped in.

The first day was a mess. Terminal commands, foreign syntax, cryptic file trees – I understood nothing. But I was hooked. That night, I traded doomscrolling for YouTube tutorials. By morning, I had a plan.

Armed with coffee and curiosity, I mapped out what I wanted to build, improve, and simplify. I set a modest target: upgrade three modules in our 60+ module system by the end of the week.

By 11:30 PM that same day, I was alone in the office, locked in after security closed down the building with half the list already done.

The code wasn’t elegant. It wasn’t efficient. But it worked. And it was mine. I wasn’t just writing code – I was imagining what the world could look like and then bending reality to match it. 

What AI Really Is

That’s when it clicked: AI isn’t the product. It’s the amplifier.

We’ve seen this before. The computer unlocked new productivity – but only for the few who could afford and understand it. Then came the PC, then the smartphone, each wave democratizing access further.

AI is the next step. It’s not just a tool. It’s a force multiplier. It lowers the cost of creation, of experimentation, of imagination. The friction between idea and execution has never been this low.

The Great Equalizer

As someone who remembers life before the internet – and every major tech leap since – I can say with confidence: this is different. AI doesn’t just promise to change industries. It promises to change who gets to play.

You no longer need a Stanford CS degree or Silicon Valley connections. You just need curiosity and internet access.

That shift is profound.

We’ve always been supply constrained. More problems than problem solvers. More ideas than engineers. AI flips that. Suddenly, the bottleneck isn’t labor or capital. It’s imagination.

Sure, it’ll displace jobs – just like tractors replaced farmhands. But it’ll also free people. No one dreams of being a call center agent. They dream of helping people. AI lets us do the latter without the grind of the former.

Why does it take a decade of study to become a doctor? Because you have to learn and memorize vast amounts. What happens when that knowledge is available on-demand, flawlessly? We’re not far from finding out.

Fewer taxi drivers. More scientists. That’s the world I want.

The Dark Side

Of course, it’s not all utopia.

AI agents don’t sleep. They don’t blink. If directed toward good, they’ll build entire industries. If directed toward malice, they can just as easily destroy a person’s reputation overnight. Think Eagle Eye, but cheaper.

As compute costs drop and context windows expand, the ability to weaponize AI grows. We must build guardrails. Fast.

Still, I remain optimistic. I believe in building. In progress. In bending tools toward better futures. Nuclear energy scared us too – until it lit up entire cities.

Wait not, into the storm step you must

There are two types of people in uncertain times: those who wait for clarity, and those who move toward the fog. I’ve always believed the best way to understand the storm is to walk into it.

At Rabbit, AI isn’t an experiment. It’s infrastructure.

Just like mobile and cloud are no longer departments – they’re defaults – AI is becoming foundational. It’s how we ideate, plan, and build.

We’re already seeing the impact.

The time from idea to delivery used to be gated by bandwidth. How fast can a team scope, prototype, test, and ship? Now, that cycle is compressed. A PM and a designer with access to the right AI stack can go from wireframe to working prototype before lunch.

We use AI to write job descriptions, model unit economics, analyze delivery heatmaps, handle support tickets, and even refactor frontend code. Not because it’s novel – but because it lets our people do more of what matters.

It’s changing how we work – and where we spend our intellectual resources.

Developers spend less time on boilerplate. Marketers spend less time formatting slides. Ops spends less time crunching spreadsheets. Everyone spends more time solving actual problems.

We’re not adding AI to Rabbit. We’re writing the next chapter of Rabbit hand in hand with AI.

It’s not a new tool – it’s a new mindset. And we’re all in.

So, where can you start?

let’s look at the 3 layers of the stack (while this is not technically correct , its how my complexity avoidant brain tends to bucket them)

1. The Model Layer
   Foundation models like GPT or Claude are like AWS. Don’t try to compete. Build on top.

2. The Application Layer
   Interfaces like chatgpt, Notion AI, or Canva. These make AI accessible. Try them.

3. The Platform or Orchestrator layer
   High level aggregators and workflow managers that stitch models and tools into something magical and truly bespoke. This is where you can truly create magic.

We’re entering an era where the only thing standing between you and the future is the courage to build.

So let’s build.

For our communities. For our countries. For the dreamers watching us now, wondering if they can do it too.

They can.