Allan Nascimento

The best investment right now is the best AI model. That’s the conclusion I came to, and today I’m doubling down on what I pay for it.

Why I Upgraded

I’ve been working a lot with AI, mainly for coding. My business partner Danilo handles the commercial side and manages the services we provide, while I focus on development and automation. While building a system we plan to sell to clients, I was almost hitting the session limit, by 11 a.m. I was already close to the weekly limit, which would have left me without AI for an entire day until the limits reset tomorrow. That would have drastically reduced my speed, even though I can still code on my own. To avoid being stuck in the middle of work, I switched from the 5x plan I was using to the 20x plan.

Paying With Real Revenue

I paid for the subscription with money that already exists, it came from the revenue of a recent client deal. I don’t subscribe to something hoping I’ll be able to pay for it later. If I didn’t have the cash, I would have started with the cheap plans, taken on more freelance projects, and moved up the staircase gradually. That’s exactly what I did until this morning: I was on the $100 plan, and now I’m on the $200 plan. In Brazil that amount is more than half the minimum wage, so it’s not a trivial expense.

Gaining Time, Not Just Power

The upgrade didn’t just give me more AI usage; it bought me more of my own time. My time is the only bottleneck right now. For example, this afternoon I worked with Fable, delegating tasks to multiple sub‑agents that ran all afternoon while I practiced jiu‑jitsu. If I had to handle everything sequentially, the workflow would have stalled and I would have lost a lot of time.

Balancing Projects

I’m ambitious, and I know the current AI wave won’t last forever, so I need to work as fast and as well as possible. I’m working on big projects that aren’t generating revenue yet but have strong potential, while also handling urgent client requests that do generate income. Right now I can keep the big projects moving without pausing for urgencies because I can run several projects in parallel: typically one large project plus two or three smaller ones. I stay in contact with clients, solve emergencies quickly, especially when a website goes down, and let the AI handle the heavy lifting. I still review everything, write tests, and manually verify results; I don’t rely 100 % on AI.

Study Routine Gap

I’m still the bottleneck, so I need to keep studying and stay focused. The better plan gives me more time, but it also raises the expectation that I’ll be more productive. I realized I’ve missed daily study for a few days, and I need to get that back. Now that workflows can run for hours, I can start a workflow, know it will take 30 minutes to give feedback, and use that half hour to study instead of scrolling on Instagram or Twitter.

Knowledge as the Arm

If I had paid this amount a year ago, I wouldn’t have gotten the same results because I had less experience. I’ve spent four years solving real problems, building WordPress sites, integrating cloud code, creating freelance products, and that knowledge is what multiplies when I use AI. AI is the lever; my years of experience are the arm that pulls. Someone without my background wouldn’t be as productive, even with the best models.

Investing in the Arm

What I need to keep doing is investing in the arm, my knowledge. The real problems I’ve faced over the years are the heavy end of the lever, and I must keep my study routine strong. AI makes my time worth more, allowing me to deliver more value to clients, not just more dollars.

If you’re using AI, let me know in the comments what plan you use, what you’re working on, and how you plan to invest in your own arm. That’s it for today.