Skip to content

Ideas & Observations

66 essays across 8 topics on building products, systems thinking, AI, and the internet.

Essays

Published writing

66 essays · All topics

Building AI Products While Everyone Else Is Still Debating AI

Most people are asking whether AI will replace jobs. Meanwhile, a smaller group is already building autonomous systems and AI workflows. The real divide is execution, not access to information.

June 2026·8 min
Read →

Building in Public Feels Weird Until It Doesn't

Posting unfinished work is uncomfortable. But people connect more with the messy middle than polished endings. Building in public creates momentum and unexpected opportunities.

June 2026·7 min
Read →

The Future Internet Will Be Run by Agents

Software today assumes humans manually do everything. The next phase is AI agents operating across workflows autonomously. One person running systems that previously required entire teams.

June 2026·9 min
Read →

Why Most Side Projects Never Get Finished

Side projects usually die from scope, unclear feedback, and treating them like hobbies instead of problems worth solving. Finishing is mostly a systems problem.

July 2026·6 min
Read →

I Stopped Waiting for Perfect Ideas

Good ideas are rarely perfect when they first show up. They become obvious only after you start building, testing, and paying attention to what people actually need.

July 2026·5 min
Read →

What Building 5 Products Taught Me

Each product teaches a different lesson about timing, users, distribution, and what matters once the novelty wears off. The pattern is rarely what you expect.

July 2026·8 min
Read →

The MVP Is Usually Smaller Than You Think

Most people overbuild the first version because they are trying to prove the product before they have evidence. The real MVP is usually a single sharp loop.

July 2026·6 min
Read →

Why Every Developer Should Build for Themselves

Building for yourself is the fastest way to learn taste, urgency, and product judgment. You stop guessing what matters because you feel it directly.

July 2026·6 min
Read →

Shipping Fast Doesn’t Mean Shipping Bad

Speed and quality are not opposites. The real issue is whether speed comes from focus or from skipping the parts that matter.

July 2026·5 min
Read →

How I Decide Which Ideas Are Worth Building

I look for pain, urgency, leverage, and whether the thing would still matter after the novelty disappears. Good ideas survive contact with reality.

July 2026·7 min
Read →

Why Features Don’t Build Companies

A feature can impress a user once. A system that solves a recurring problem can actually create a business. Companies are built on outcomes, not checklists.

July 2026·6 min
Read →

AI Won’t Replace Developers, It Will Replace Boring Work

The real shift is not full automation. It is removing the repetitive parts that keep good developers from doing actual product work.

July 2026·6 min
Read →

Why AI Is the Biggest Productivity Tool Since the Internet

AI compresses research, drafting, coding, and iteration into a shorter loop if you know how to work with it. That changes what one person can do.

July 2026·7 min
Read →

The Difference Between AI Wrappers and AI Products

A wrapper borrows intelligence. A product uses intelligence to solve a repeated, high-friction job end to end.

July 2026·6 min
Read →

The Era of One-Person Companies Has Started

With modern tools, a single builder can now do work that used to require a small team and a budget. The constraint is no longer headcount.

July 2026·7 min
Read →

What Every Student Should Learn About AI

Students need practical AI literacy: how to question outputs, use tools well, and avoid mistaking fluency for understanding.

July 2026·5 min
Read →

Building AI Products Without Millions in Funding

The advantage is not capital anymore. It is clarity, distribution, and building narrowly enough to get real usage fast.

July 2026·7 min
Read →

My AI Stack for Building Products in 2026

The stack should be boring, reliable, and replaceable. The goal is leverage, not collecting tools for screenshots.

July 2026·6 min
Read →

AI Is Becoming the New Operating System

AI is slowly becoming the layer that coordinates intent, tools, memory, and action across the rest of the software stack.

July 2026·8 min
Read →

Why Small Teams Beat Big Companies

Small teams move faster because coordination costs are lower and the feedback loop between decision and result is tighter.

July 2026·6 min
Read →

Distribution Matters More Than Features

A great feature without reach is invisible. Distribution turns useful work into something the market can actually notice.

July 2026·6 min
Read →

Every Startup Is a Marketing Company

Even technical companies need a way to make their value legible, memorable, and easy to repeat. If nobody understands it, nobody buys it.

July 2026·6 min
Read →

Why Nobody Cares About Your Idea

People care when the problem is sharp, the demo is real, and the value is obvious within seconds. Ideas only matter once they become evidence.

July 2026·5 min
Read →

Build Distribution Before Your Product

If nobody can hear you, then even a good product has no path to learn, improve, or compound. Reach is a product requirement, not an afterthought.

July 2026·6 min
Read →

Why the Best Startups Solve Their Own Problems

Founders who feel the pain can judge urgency, quality, and fit much faster than outsiders can. They know what broken feels like.

July 2026·5 min
Read →

Lessons From Launching Multiple Products

Each launch teaches what carries, what stalls, and which parts of the process are actually repeatable. Shipping becomes pattern recognition.

July 2026·8 min
Read →

How I Think About Product-Market Fit

Fit is not a slogan. It is when the product starts pulling demand instead of you constantly pushing it. The market should feel some amount of hunger.

July 2026·7 min
Read →

Why I Build Instead of Doomscroll

Building gives the day a shape. Doomscrolling gives you noise, anxiety, and the illusion of being informed.

July 2026·5 min
Read →

Being 18 Doesn’t Stop You From Building Real Software

Age matters less than consistency, attention to users, and a willingness to ship imperfect work in public. Real software is built by people who keep going.

July 2026·6 min
Read →

The Internet Rewards Consistency More Than Talent

Talent matters, but the web mostly remembers people who keep showing up long enough for the work to stack.

July 2026·5 min
Read →

You Don’t Need Permission to Start

Waiting for validation is just a slower way to avoid uncertainty. Starting is how you get useful feedback.

July 2026·5 min
Read →

Why Curiosity Is the Most Valuable Skill

Curiosity keeps you moving through unfamiliar territory long enough to actually learn something valuable. It is the engine behind most useful builders.

July 2026·6 min
Read →

Learning in Public Changed Everything

Once your process is visible, feedback arrives faster and the work becomes easier to improve honestly. Visibility creates acceleration.

July 2026·6 min
Read →

Why I Chose Building Over Waiting

Waiting feels safe until you realize that motion, even messy motion, creates more opportunity than hesitation. Building is how you get unstuck.

July 2026·5 min
Read →

My Favorite Mistakes So Far

The best mistakes are expensive enough to teach you something and small enough to keep you in the game. They make the next version better.

July 2026·7 min
Read →

How I Built LeadScrapper

LeadScrapper started as a way to reduce the boring parts of finding prospects and turned into a workflow system for discovery, qualification, and outreach.

July 2026·10 min
Read →

Inside the Architecture of Itiner

Itiner is built around a calm workflow for travel designers, where AI drafts the structure and humans keep control of the final output.

July 2026·9 min
Read →

Designing Nucleus CSS

Nucleus CSS was built around simplicity, responsiveness, and speed. The goal was a small system that stays useful instead of becoming a burden.

July 2026·7 min
Read →

Why I Chose Next.js for My Portfolio

The portfolio needs performance, simple routing, and content-friendly structure. Next.js fits that job well without becoming too heavy.

July 2026·6 min
Read →

Lessons From Building an AI Travel Platform

Workflow-heavy software only works when AI helps without taking over. The product has to respect human judgment.

July 2026·8 min
Read →

What I Learned While Building an AI Flashcard App

Students do not want novelty. They want something fast, usable, and accurate enough to help them study under pressure.

July 2026·7 min
Read →

Building a Browser Game From Scratch

A browser game forces you to think about feel, feedback, and pacing. The hard part is making the experience expressive without making it heavy.

July 2026·8 min
Read →

How I Structure Large Next.js Projects

The folder structure, boundaries, and habits matter more than a clever architecture diagram. Keeping the project understandable is the real win.

July 2026·7 min
Read →

Degrees Are Becoming Proof of Persistence, Not Knowledge

The credential still signals effort, but the internet increasingly measures what you can actually build.

July 2026·5 min
Read →

The Resume Is Slowly Dying

Static summaries cannot compete with live products, public work, and proof that people use what you make.

July 2026·5 min
Read →

The Best Portfolio Is Something People Use

Usage beats presentation because real users are harder to fake than polished screenshots.

July 2026·5 min
Read →

Everyone Wants to Be a Founder Until They Have to Sell

Selling is where assumptions collide with reality, and most founders discover that too late.

July 2026·6 min
Read →

AI Has Made Execution the Only Moat

Ideas are cheap, tools are abundant, and the differentiator is how fast you turn intent into outcome.

July 2026·6 min
Read →

Software Is Becoming a Commodity. Distribution Isn’t.

If building gets easier, the scarce thing becomes attention, trust, and access to the right audience.

July 2026·6 min
Read →

The Next Billion-Dollar Companies Will Be Tiny Teams

Leverage now belongs to small groups that can move fast, stay focused, and use software as force multiplication.

July 2026·6 min
Read →

The Browser Is the New Operating System

For a lot of work, the browser has already become the default workspace where everything important happens.

July 2026·6 min
Read →

How to Build a SaaS From Scratch

Start with a narrow problem, a clear user, and a launch path that does not depend on luck.

July 2026·8 min
Read →

Best AI Tools for Students

Focus on tools that save time, improve understanding, and fit real study habits instead of hype cycles.

July 2026·6 min
Read →

Best AI Tools for Developers

The best tools help with scaffolding, debugging, documentation, and the repetitive parts of building software.

July 2026·6 min
Read →

How to Build Your First Startup

The first goal is not scale. It is learning how to identify a problem and make something people want.

July 2026·7 min
Read →

How to Learn Web Development in 2026

The path should combine fundamentals, projects, shipping, and enough judgment to avoid tutorial dependency.

July 2026·7 min
Read →

How to Validate a Startup Idea

Validation is about testing demand with small, real actions before you overbuild a product.

July 2026·6 min
Read →

What Is Building in Public?

A practical explanation of showing the process, the tradeoffs, and the work as it happens.

July 2026·5 min
Read →

How to Create a Developer Portfolio

A portfolio should be a proof surface: clear work, clear thinking, and obvious evidence that you can ship.

July 2026·6 min
Read →

Why Every Developer Needs a Personal Website

Your site is where your work, ideas, and identity can live without platform constraints.

July 2026·5 min
Read →

How to Go From Student to Founder

The transition is mostly about compounding skills, building signal, and treating small shipped work seriously.

July 2026·7 min
Read →

Build Log #1: Improving LeadScrapper Search

A weekly note on what changed, what got cleaner, and what still makes prospecting slower than it should be.

July 2026·4 min
Read →

Build Log #2: Rebuilding Itiner’s UI

A look at interface decisions, hierarchy, and why calm design matters in workflow software.

July 2026·4 min
Read →

Build Log #3: My SEO Experiments

A practical log of what I tried, what indexed, and what turned out to matter more than expected.

July 2026·4 min
Read →

Build Log #4: Growing a SaaS With $0 in Ads

What happens when growth has to come from product quality, distribution, and persistence instead of spend.

July 2026·5 min
Read →

Build Log #5: What Broke This Week

The most useful post format is often the one that tells the truth about bugs, bottlenecks, and recovery.

July 2026·4 min
Read →

These essays are works in progress. I update them as the work and my thinking change.

Got thoughts? Email me →