What a time to be building.

I’ve been heads down working in and on the businesses, so I haven’t spent much time sharing here lately. Building lots of things. Spending late nights working with agents. Designing new software, improving old software, and getting through an amount of work that would have otherwise been impossible given my time constraints.

I’m coming up for air to share my latest thoughts on AI, what I’ve been learning, and why I think everyone should be putting these tools to work.

A year ago, even six months ago, I didn’t think agents were quite there yet. The latest frontier models, like Claude Fable 5 and 5.1, Astra, and Grok 4.6, have changed what I’m willing to take on. They are getting so good, so fast.

As a creator, I’m excited to hit the keyboard every chance I get. Old code. New code. Small ideas. Big ideas. Things I would have put off because I simply didn’t have the time are now things I can actually work through.

But the biggest point I want to make is this: agents are for everyone.

Finance, marketing, sales, operations, engineering. Whatever your role, my advice is to spend time learning how to use them. There is too much opportunity here to assume this is just something for the programmers.

One example I see firsthand is our director of finance. He is programming every day, building full applications for forecasting, cash flow planning, profitability analysis, budgeting, and financial triage. He had never touched code before joining us.

Now he’s building financial tools that used to be available only to companies 10 times our size, with far bigger budgets and teams.

He knows finance. He knows what the business needs. And now he can build it.

Small teams can accomplish mighty things now.

That example has also changed how I think about specialists and generalists.

I believe the opportunity for generalists is expanding dramatically. If you’re curious, understand how different parts of a business fit together, and are willing to learn, you can take an idea much further on your own than you could before.

You can understand the problem, research possible solutions, design something, build it, and figure out how to get it into people’s hands. The gaps in your experience are becoming things you can work through.

We all have specialties. But I think the path gets narrower if we insist on staying entirely within them.

A finance professional can build an application. A designer can take a concept all the way to working software. A salesperson can analyze their own data and build a process around what they find. An operations person can fix the workflow they’ve been frustrated with for years.

That doesn’t make their original specialty less valuable. It gives them more ways to use it.

Our director of finance’s knowledge is exactly why he knows what to build. The agents help him do something with that knowledge that he couldn’t do before.

I feel the same thing in my own work. I can take on more design, cloud architecture, data engineering, and product management. I can build on what I already know and learn the parts I don’t.

That’s what I mean by expanding your areas of influence. Keep developing your specialty, but give yourself permission to work beyond it. You may find that you’re capable of a lot more than your job description suggests.

And you don’t have to build software to get value from agents. You might need to investigate why margins changed, research a group of sales prospects, put together a campaign, or get a weekly reporting process off your plate.

You probably already know which parts of your work need attention. You’ve just been too busy doing the work to improve how it gets done.

That’s where I would start.

Pick something that eats up your week, or an idea you keep putting off. Give the agent the background, examples of what you work with, and a clear explanation of what you’re trying to accomplish. Let it work through the problem with you. Ask questions. Check what it produces. Keep going when the first attempt falls short.

You can learn while doing. In fact, I think working on a real problem is one of the best ways to get comfortable with these tools.

Then start thinking about how that work could keep moving without you having to kick off every step.

That’s what excites me about the agentic side of this technology. Agents running throughout the day on our behalf. Picking up work when something happens. Investigating problems. Connecting information across systems. Building solutions. Testing them. Moving projects forward while we’re working on something else.

A customer request comes in and an agent starts gathering the information needed to resolve it. Financial results change and an agent investigates what drove the change. An error appears and an agent starts working through the cause.

Getting those processes working takes effort. But imagine how much work you could move forward once you do.

In software, we used to be the factory. Now, each of us has the opportunity to build and run the factory. I think about that a lot.

I’d also encourage you to spend time watching the people pushing these tools the furthest. Seeing how someone else works can open your eyes to something you didn’t know was possible.

DHH calls this “Endless Execution.” That describes a lot of what I’m feeling. I have no shortage of things I want to build, and more of them feel within reach every day.

These two interviews are worth making time for:

And if you aren’t watching what is happening with Omarchy, take a look. An operating system embracing agents as part of the everyday computing experience has me excited about where this is heading.

DHH is one of the people I most enjoy following. His enthusiasm for building with agents is contagious. A couple of his posts worth checking out are here and here.

Tobi Lütke is another. It’s awesome to watch people who have already built so much get this excited about what they can create next.

Look through Omarchy’s founding patrons, too. Tobi Lütke, Patrick Collison, Michael Dell, Jack Dorsey, Jason Fried, and others. There are some remarkable builders paying attention to this.

We’re also using Laravel Cloud for more of what we build. It takes a lot of the work out of deploying and running applications, which matters when agents are helping you create more software than ever.

You still have to get that software into people’s hands. Making that part easier gives small teams more time to keep improving what they’ve built.

Shoutout to Taylor Otwell and the Laravel team. Between the tools helping us build and the platforms helping us get things live, it’s a pretty remarkable time to be creating software.

You do have to put in the time. I’m still learning, and there is a lot I haven’t figured out. But every difficult problem I work through gives me a better understanding of what I can do next.

If you tried these tools a while ago and weren’t impressed, I’d encourage you to give them another shot. Bring a real problem and spend enough time with it to get past the first attempt.

You don’t need to have everything figured out before you start. Choose something worth working on and begin. Learn the tools. Learn from other people. Keep testing what you think you’re capable of.

Specialist or generalist, whatever your role, there is room for you in this.

Agents are for everyone. Get curious. Get going.

Now, back to building!