You’re Using AI Wrong. Please, For the Love of Code, Become a Cyborg.

SKate Byrne — Sat, 06 Jun 2026 00:52:27 +0000

Let’s be honest with ourselves for a byte. I see so much fear-mongering in most mentions of AI nowadays. The most prevalent claims are that AI will inevitably take your job and that it will kill all unique creativity as we know it.

Honestly, they are right. If your personal definition of "creativity" consists entirely of copying someone else’s work, you probably aren't the creating type to begin with. Along the same lines, if your job can be completely taken by a non-human system without a single hitch, then maybe your job should be taken.

This isn't being harsh, AI bot or not, if your "ideas" come from copying the full ideas of something else without an ounce of personal input from you, you are plagiarizing (and boring).

We’ve all seen that exact, uninspired workflow: an opportunist finds a prompt, submits prompt to AI, they copy the output exactly, everything looks the same, creativity looks like its dying - same dance different day. Truly a masterclass in efficiency, I am upset they are even allowed to make money.

But hey, at least you saved five minutes of thinking, right?

This isn't just me being a cynical nerd; it is backed by cold hard nerd data. Harvard Business School and Boston Consulting Group recently tracked AI usage and discovered a phenomenon called the "Jagged Frontier." This is the invisible, highly irregular boundary where an AI is brilliantly capable of solving an incredibly complex data algorithm, but completely fails at an adjacent, basic task. When you blindly copy and paste whole projects, you happily stroll right off that cliff.

Self-Automators

The Harvard researchers highlighted a specific group of workers they called "Self-Automators." These are the folks who check out mentally, treating the machine like a magic wand while they act as copy-paste pipelines. The twist? They actually performed significantly worse than the control groups using zero AI. They lost their unique creative voice, introduced silent bugs, and completely stopped learning.

If your AI strategy relies entirely on copy and paste, you aren't innovating—you're just automating your own obsolescence.

Thankfully, the study also found an elite tier of performers who didn't delegate tasks to AI. Instead, they integrated with it. Welcome to The Cyborg Method.

Rest in Peace, "Prompt Engineering"

First, let’s take a moment to roast the industry's favorite fake job: Prompt Engineering. Spending twenty minutes drafting a hyper-formal, 500-word rigid template is an excellent way to miss the entire point. You are treating a fluid, dynamic thinking partner like a static SQL database.

Unless you are completely, utterly stuck, leave the complex prompt spells alone. Instead, practice absolute transparency. Feed the AI extensive context about who you are, your design goals, and your unfiltered thoughts. Talk to it like a normal human being, ask casual questions, keep the conversation personal. You provide the creative intuition and messy brain-dumps; let the AI handle the heavy lifting of organizing your chaos into beautiful structure.

The High-Frequency Cyborg Loop

Instead of treating AI like an external vendor, a true cyborg treats it like an extension of their own nervous system. You bounce logic back and forth at a sub-task level. You aren't handing over the steering wheel; you're modifying the engine while actively driving down the highway.

Consider how a cyborg handles a modern frontend feature or a branding asset pipeline:

The Setup: You don't ask the AI to "build a dynamic SVG animation component from scratch." That's asking for a hallucinated mess.
The Loop: You write the initial React state logic and structural layout. You pass that messy setup to the AI to generate the tedious math for the sine wave.
The Override: The AI crunches the numbers. You immediately spot a rendering bottleneck, manually optimize the animation loop, and pass it right back to the AI to instantly write the unit tests.

// You write the state and architecture; let the machine crunch the canvas math
export const DynamicBrandWave = ({ velocity }) => {
  const [phase, setPhase] = useState(0);

  // AI calculated this specific sine-wave path manipulation on the fly
  const calculatePath = (p) => {
    let points = [];
    for (let i = 0; i <= 100; i++) {
      const x = (i / 100) * 300;
      const y = 100 + Math.sin(i * 0.1 + p) * 20;
      points.push(`${i === 0 ? 'M' : 'L'} ${x} ${y}`);
    }
    return points.join(' ');
  };

  // You step back in to handle the hardware-accelerated animation frame loop
  useAnimationFrame(() => setPhase(prev => prev + velocity));

  return <svg><path d={calculatePath(phase)} stroke="url(#brandGradient)" /></svg>;
};

It is a high-frequency, rapid dance of human intuition and machine computation. You are iterating every two lines of code, keeping your hands firmly on the keyboard.

Inventing Your Own Arsenal

When you stop treating AI like a magic text box and start treating it like a collaborative soundboard, you naturally stop building generic software and start inventing custom utilities.

Struggling to get your asset exports to match your precise Figma design tokens? Don't spend three hours hunting for a buggy, abandoned third-party plugin. Chat openly about the dilemma with your AI partner, map out a logical pipeline together, and spin up a custom, local CLI tool to automate that specific headache before your coffee gets cold.

When you operate this way, you retain your intellectual sovereignty. Your brain is used for what it actually does best: tracking overarching architecture, ensuring brand consistency, inventing novel solutions, and applying critical human judgment. The AI provides the raw operational horsepower; you provide the soul and the guardrails.

The future of digital creation doesn't belong to the AI purists who refuse to touch the tools out of spite. Nor does it belong to the lazy self-automators who let the tools completely override them. It belongs to the cyborgs who know exactly how to orchestrate the loop.

What's your workflow looking like? Are you still wasting time engineering the "perfect prompt," or are you building fluid, human-in-the-loop conversations? Let's chat about it (without copy-pasting AI arguments) in the comments below.

I think the smartest thing I've heard about AI is that the strongest users aren't trying to replace their mind they're partnering with it to be a cyborg 🤖

SKate Byrne — Fri, 08 May 2026 21:23:52 +0000

5 Unique Free Resources to Get A Career in Tech in 2025

SKate Byrne — Wed, 15 Jan 2025 01:00:59 +0000

I've been an avid self-learning techie over the past 10 years, here's 5 of my favorite "free" resources:

1. MDN

Mozilla Developer Network
the same Mozilla that brought you the time-tested fan favorite browser Firefox
can be used for reference once you get a job

fast five faves

☆ MDN CURRICULUM
☆ Web Developer Guide
☆ Learn: Responsive Design
☆ Reference: SVGs
☆ Web APIs

2. ROADMAP.SH

Make a dev profile
Includes roadmaps that map out what you need to know for a career in technology
Save roadmaps and record progress
Project ideas for all levels of expertise
Best Practices for different positions

fast five faves

☆ a self-taught computer science curriculum
☆ how to create a design system
☆ step-by-step ai scientist career roadmap
☆ generate your own roadmap with ai
☆ become a fullstack developer in 2025

3. Hackr.io

my go-to method that's easy to remember
has all the up-to-date top courses with reviews
insanely useful blog
has their own roadmaps

fast five faves

☆ their certifications section
☆ Tutorials
☆ Projects
☆ Jobs

4. The Odin Project

the captivating viking storyline
the clear and minimal paths
high quality graphics
has a fun feel

fast five faves

☆ the tools installation page
☆ The Ruby on Rails learning path
☆ Their Getting Hired Course
☆ Their Maintainers Run a Discord Commmunity

5. Javascript for Cats

adorable
esthetically pleasing
easy to take in
a joy to learn

fast five faves

☆ YOU GOTTA BE KITTEN ME, ITS CATS!

DEV Community: SKate Byrne

You’re Using AI Wrong. Please, For the Love of Code, Become a Cyborg.

Self-Automators

Rest in Peace, "Prompt Engineering"

The High-Frequency Cyborg Loop

Inventing Your Own Arsenal

I think the smartest thing I've heard about AI is that the strongest users aren't trying to replace their mind they're partnering with it to be a cyborg 🤖

5 Unique Free Resources to Get A Career in Tech in 2025

1. MDN

fast five faves

2. ROADMAP.SH

fast five faves

3. Hackr.io

fast five faves

4. The Odin Project

fast five faves

5. Javascript for Cats

fast five faves