Typical Project Advice Is Now Outdated

For years, aspiring developers were told to build projects. Not because anyone needed another chat app or Netflix clone, but because building those projects demonstrated skill.

If you built a messaging app from scratch in 2018, people were impressed. Today? A motivated developer with AI can build a working version over a weekend.

That doesn't mean development is dead. It means the value of projects has changed.

The question is no longer: Can you build it?

The question is: Was it worth building in the first place?

AI Changed the Economics of Software

Not long ago, creating software required an enormous amount of effort.

You needed to:

  • Learn a framework/language
  • Read documentation
  • Debug constantly
  • Build every feature yourself
  • Design, reasearch, and maintain a tech stack

Today, AI can generate:

  • Authentication systems
  • Best tech stacks to implement
  • CRUD applications
  • Database schemas
  • Landing pages
  • API integrations
  • Entire frontend components

Many projects that once took months can now be prototyped in days. That changes the game. When something becomes easier to create, its value decreases. The same thing has happened to code.

The Era of Portfolio Projects Is Ending

Let's compare two students.

Student A

Built:

  • Netflix Clone
  • Spotify Clone
  • Discord Clone
  • Twitter Clone

Student B

Built:

  • A scheduling tool used by 50 students
  • A platform that helped classmates find tutors
  • A dashboard used by a local business

Ten years ago, Student A might have stood out. Today, Student B is far more impressive. Why?

Reminder:
Because the hard part is no longer writing the code. The hard part is finding a real problem and solving it.

The New Scarce Resources

Code is becoming abundant. Other things are becoming scarce.

Deep understanding of specific topics

In a world where generalists once thrived. The specialist now rules. Deep knowledge about more niche topics or industries

Distribution

Getting users that genuinely enjoy to use the product.

Trust

Convincing people to rely on your product.

Data

Having access to information others do not.

Taste

Knowing what should exist before everyone else does.

These are the new skills companies are now seeking. Hiring managers will begin to ask themselves:

  • What valuable knowledge/context that isn't easily known by an LLM does this candidate possess?
  • Does this candidate understand what a user would like to experience when using a product?
  • Can this candidate intelligently make big picture architecture decisions?
  • Will this candidate understand why certain technologies are used in an application rather than just knowing how to use them?
  • Does this candidate possess a natural knack for good taste and

Build Things AI Can't Invent

AI is incredibly good at recreating things that already exist. It is terrible at living your life. It doesn't know:

  • What frustrates students at your university
  • What problems your fraternity deals with
  • What your tutoring clients struggle with
  • What annoys people in your hometown

Only you do. That knowledge is now the most valuable thing you know. Instead of building another task manager, ask:

What problem do I understand better than most people? For example, a University student might build:

  • A class planning assistant
  • A tutoring marketplace
  • A rush management platform
  • A roommate matching system

Those ideas may sound smaller than a Netflix clone. They are actually much harder to copy because they are rooted in real experience. Additionaly, this type of project can become the most valuable in your entire portfolio.

AI Raises the Ceiling

Many people assume AI lowers the bar. That's true. But it also raises the ceiling.

Before AI, a student might spend three months building a basic application. Now that same student can spend three months building something much more ambitious.

Reminder:
The opportunity isn't to build easier things. The opportunity is to build bigger things.

What I Think New Developers Should Build

I am avoiding ❌:

  • Netflix clones
  • Spotify clones
  • Weather apps
  • Generic task managers
  • Another chat application

Instead, I would build ✅:

  • Tools for communities I belong to
  • Software that solves my own problems
  • Products that can get real users
  • Applications with unique datasets
  • Projects that teach me something difficult

Reminder: The goal isn't to impress people with code. The goal is to create something useful.

A Simple Test

Before starting your next project, ask:

If AI could write every line of code for me tomorrow, would this still be a good idea?

If the answer is no, the project probably wasn't very interesting to begin with. If the answer is yes, you're probably onto something worth building.

Final Thoughts

The best project in 2026 isn't the hardest thing to code. It's the hardest thing to understand.

AI is rapidly reducing the cost of implementation. That means developers are being pushed higher up the stack.

The future belongs to builders who understand people, problems, and communities. Not just programming languages.