THIS WEEK: WE'RE ALL SOFTWARE ENGINEERS NOW

🌟 Editor's Note

Thanks for reading H4CKER, my newsletter on AI, marketing, and tech-powered growth hacking. Share your feedback and thoughts with me at [email protected] or on X at @nathanabinford.

🫵 Wouldn't You Like To Be An Engineer Too?

Something incredible happened this year. We all became software engineers overnight. If you have any interest in building things or learning technology, you did anyway.

There are a lot of naysayers still denying this reality but they’re increasingly being revealed as digital Karens and disappearing into the shadows.

But if you’re in the “bubble” of the AI revolution, the reality is setting in. There are no more excuses for not building the tools and systems you’ve always wanted. And there is less and less reason to wait for someone else to create your perfect solution.

What do I mean by that?

I mean that if you have the slightest technical proficiency and are willing to dedicate the time to it, you can vibe-code custom software to do the things you’ve always wished you had software to do for you.

It’s not a question of cost (not really), and not a question of effort exactly (AI does the work). It’s simply a matter of time and persistence and willingness to learn.

You don’t need to learn how to code, necessarily. You do need to learn how to read code and how the different layers of software work together though: operating system, data layer, application layer, client/server architecture, REST / MCP (aka the API layer), and so on.

The AI can be immensely helpful when learning this stuff but you can’t simply take it for granted the AI generated the best possible solution. You don’t have to write the code, but you will need to generally understand code and fully understand the architecture.

We’re not all coders, we are all software engineers.

⚙️ Adopting An Automation-First Mentality

Speed and scale are the most important benefits that AI provides today. To take full advantage everything needs to be done at scale.

Admittedly, this is challenging for me because I like to do things the right way first. To figure them out manually, refine them iteratively, and then automate what’s working well.

But this produces ROI at a fraction of the volume and pace that simply running recklessly at a half-baked idea does —when you take that idea to scale.

And this is the reason that we’re all engineers now. It’s not that you want to be, necessarily. It’s that you’re almost being forced to engineer your own solutions because it’s the only way to do what you uniquely do, at scale.

I’ve been dealing with this situation since I started my business back up last year.

I tried hiring. I invested into lots of different software platforms.

In the end, I’ve struggled to grow because marketing strategy is changing very quickly and the noise in the marketplace requires serious volume to compete.

Moving fast AND producing output at scale is very challenging if you’re starting every task as a manual process and converting it to an automated one.

And that recent realization has me retooling my thinking on automation, scaling, and how to go from idea to execution; faster, bigger, sooner.

Instead of starting from a manual prototype, is it more efficient to start automating from the very beginning with AI to streamline this process?

A “fail forward faster” approach to AI-powered marketing…

🤩 Not Just A Theory: I’m Going To Do This & You Should Too

I’m not just saying this, I’m beginning to do it.

It’s not exactly the same thing, but you can sort of chart the trajectory I’m already following…

Let’s review what I’ve managed to create so far since August 7th, 2025 when GPT-5 was released and this golden age of software began in earnest.

Not all of these are automated workflows, but they all eliminate redundant work for me, making it easier for me to scale my efforts. I expect this trend to accelerate.

Automated Content Briefs

Using Make I built an automation that takes targeted search terms from a Google Sheet and uses multiple calls to Perplexity and ChatGPT to develop article ideas, create outlines for the articles, and package up research data for RAG, outputting everything I need to generate each article automatically.

This has streamlined content production dramatically; and I put out about 2 dozen pieces of content per client each month…and growing (up from 4-5).

“Speed To Lead” / Auto-SMS Lead Outreach

One of my clients was losing 10+ leads a month at $200/ea to bad habits (calling once during work hours and then never trying again). I solved it by vibe coding a solution that automatically texts the new leads within 1 minute of filling out a web form.

Now the client is thrilled with their lead volume (and up several hundred thousand in revenue) and the only change was this automation.

Wordpress Plugins: Gallery, Shortcodes, Gutenberg Blocks

In my business, one of the most redundant tasks still eluding automation has long been Wordpress site updates: adding local service pages, updating content, tweaking SEO. I’ve also been complaining about the bloat of Wordpress page builders since they became popular. After building a custom gallery plugin in a couple hours one afternoon, I decided to abandon page builders entirely and just vibe code my own custom themes, plugins, etc.

Now I’m building better looking sites that run (much) faster and tie into my processes exactly how I want. And, I’m building in the same time it was taking me to do it in Elementor previously, and all the elements are modular and reusable (saving even more time in the future).

My point is, this is all very accessible.

It’s been years since I was coding on the regular and I was able to pick it up and be producing viable code in only been a few months! And I spent maybe $100 producing all that code, and another $50/m hosting the my own production and staging environments.

You could easily spend more than that each month paying for different SaaS platforms to accomplish the same thing, but with less customization that’s specific to you.

💸 Just Because You Can Doesn’t YOU Have To

It takes time for a change this significant to settle in but anyone can build now. And if you don’t want to learn, or don’t have time, that’s fine.

You know someone who does.

What’s important, in my mind, is that you know that the possibility exists: that you can solve any problem, eliminate any inefficiency, automate any redundancy, and someone around you has the technology and nerdiness to figure it out.

That…is incredible. Take advantage!

⚙️ The Laboratory: Closing The Feedback Loop With Cursor 2.0

Cursor, one of my favorite companies right now, released their v 2.0 major update with a slew of new agentic features, an in-app browser (HUGE, more on this later), and most notably, their own coding model Composer (previously tested in the wild under the name Cheetah).

It’s a beast of a model: Super-duper fast, as smart as Sonnet 4.5 (my estimation), and tiny bit reckless (much like Sonnet). It is, apparently, based on Chinese open-source models however.

And while that’s cool (and good for Cursor’s bottom line), the REAL big win is that Cursor’s effectively closed the feedback loop on web development, between the AI and the environment, so it can test its work semi-autonomously.

Because THAT is super exciting…

Here’s what I show in the video:

✅ Switching Between Agent & Editor Dashboards : Agent view to run multiple Cursor agents. Editor view to manage Codex and Composer at the same time.

✅  Codex For Projects, Composer For Admin: New workflow: 1. Project guide in markdown. 2. Build features with Codex. 3. Simultaneously create tests with Composer.

✅ Autonomous Testing: Composer runs tests autonomously, using CLI and Cursor’s in-app browser. 👀

Nathan Binford
AI & Marketing Strategist

I hope you enjoyed this newsletter. Please, tell me what you like and what you don’t, and how to make this newsletter more valuable to you.

And if you need help with AI, marketing, or automation, grab time on my calendar for a quick chat and I’ll do my best to help!

Thanks For Reading H4CKER!