Most people are using AI wrong.
They're using it to write LinkedIn posts and summarize emails. That's fine. But there's a different category of person
who's using it to build systems that generate income without their direct involvement — and the gap between those two
groups is widening fast.
In 2026, the question isn't "are you using AI?" Everyone is. The real question is whether you're using it to consume
or to produce. Whether you're saving 20 minutes a day on email, or whether you've built something that runs while
you're asleep.
That's what this newsletter is about. I'm not here to tell you AI will make you rich. I'm here to show you exactly
what I'm building, what's working, what's breaking, and which tools are actually worth paying for.
What I'm Building This Week
The crypto arbitrage bot is live — and it's more complicated than I expected.
The concept is simple: crypto prices aren't identical across exchanges at any given moment. If ETH is trading at
$2,840 on Kraken and $2,847 on another exchange, you buy on one and sell on the other and pocket the spread. The bot
monitors those price deltas in real time and executes trades automatically when the gap crosses a threshold that
covers fees and still leaves profit.
Here's what's actually working: the detection layer is solid. The bot is scanning multiple trading pairs across
exchanges and flagging real opportunities several times per hour during high-volatility windows.
Here's what's not: execution speed. By the time an order hits the second exchange, the spread has often collapsed.
Latency is the enemy of arbitrage. I'm currently testing co-located servers closer to the exchange APIs to cut that
lag. The math checks out on paper. The infrastructure is the problem, not the strategy.
Next week I'll have real numbers on whether the server upgrade changes the execution rate enough to make this
profitable at scale.
One Tool Worth Your Time: n8n
n8n is an open-source workflow automation platform that lets you connect apps, APIs, and AI models into automated
pipelines — without writing much code. Think of it as the backbone that makes your tools run on their own.
I use n8n to wire together my sports EV scanner — it pulls odds data from an API, runs it through a calculation layer,
and pushes alerts to Telegram automatically. You can build the same setup for almost any data-driven workflow: price
monitoring, newsletter automation, lead qualification, client onboarding.
The self-hosted version means you're not paying per execution as you scale — which matters once your automations are
running hundreds of times a day.
Pricing: Cloud plans start at €20/month. Self-hosted on your own server is free.
Next Week
I'm going to break down the YouTube automation pipeline — how I'm using AI to script, voice, and publish videos
without appearing on camera, and what the revenue picture looks like at 30 days in.
If you know someone building a side income with AI, forward this to them.
Find me on Twitter: @theoperatorai2