👋 Hey, this is Ben with a 🔒 subscriber-only issue🔒 of Ben’s Bites Pro. A weekly newsletter covering AI trends, ideas, business breakdowns and how companies are using it internally.

I spent many years of my life wanting to build stuff on the internet. I tried to learn to code several times and just kept hitting blockers.

So I built stuff with what we now call ‘no-code tools’. I pushed the limits by building things like Airbnb without code, Instagram and other things. People asked how I was doing it, so that became my startup, Makerpad—a no-code tutorial platform.

Those who can’t do, teach…

I sold it to Zapier in March 2021 and now, AI has entered the chat.

Anyone can use AI to generate code to build applications but there are still blockers. So I asked several folks how AI has enabled them to use code for real-life projects and use cases. Thank you to everyone for their contributions!

Here is how I broke this down:

• The journey from not being able to code, starting coding with AI, and building more complex projects

• A mini-interview with Charlie, who built a Chrome extension in 45 mins with AI

• 3 steps on how to learn to code with AI

• Examples of others who’ve built projects with AI-generated code

• A potential curriculum to help you learn to code with AI

• Resources to help you

• How you can build an app with AI-generated code

• Tips for others looking to use AI to write code

Trying to code before AI

One of the issues of starting to code is learning and maintaining those skills in a timeframe where distractions won’t take over.

Starting 100 days of Python to drop off on day 4 because I wasn’t moving quickly enough, I hadn’t learnt how to build a full project or other work got in the way, is just too common. It's happened a lot.

Prioritisation of learning this new skill vs doing work I already know is tough.

Leaning towards tools or workflows I already figured out takes away from time spent learning to code. It's like being a guitar player and deciding to learn the piano. You have a sense of music, rhythm, and melody, but you have to start with simple tunes and scales to master the new instrument.

Sometimes the specific course, learning format or not even knowing which course is best is enough to put people off.

And sometimes life just gets in the way. Having kids, going to college, a new job, a new house, a new relationship etc.

It feels like there needed to be a bigger NEED to learn.

Perhaps these feel familiar.

Starting to code with AI

That need still has to be there but it’s easier to pick up writing code, thanks to AI.

Previously writing a script felt like a mammoth task which can now be done in seconds using ChatGPT.

The problem is no longer writing code, but understanding what it does and what to do with it.

And needs change when the way with which to get there becomes easier.

Using a no-code tool to build something still requires you to build the thing and takes time to learn the platform, test it, tweak it, and so on.

But there comes a time when off-the-shelf tools don’t cut it or don’t make sense.

It’s not simple to build Chrome extensions or Google App Scripts with no-code tools. And sometimes you just need a script to do a job for you, or you want to work with an API.

Personal projects often get so far that you want to take them further, adding features. It sparks your curiosity about what else you could build.

Note: This is why I always pushed for learning no-code tools first, you figure out what it takes to build something, test and tweak it then launch it. You learn a lot about the process and oftentimes you want to go the extra mile, which often means, learning to code.

Projects at work can be a great motivator to pick up coding. You come across a thousand problems that could ‘just be solved if I could code’. And collaborating with developers on your team becomes much easier.

It could improve an internal process or something that benefits end users.

Personal projects are another great motivator. Everyone thinks they have a $Bn app idea but starting small is much easier to stomach and you learn the process of building and shipping something.

You could build something for your friends, partner, roommates or kids to generate bedtime stories, come up with meal plans or organise trips.

And you need to think of bringing in AI at the right moment, instead of expecting it to spit out the perfect app from one prompt.

Building more complex projects

Once the lightbulb moment has happened, further iterating on what started as a simple project or embarking on new ones becomes less daunting.

A mini-interview: How Charlie Ward built a functioning Chrome Extension in 10 steps and 45 minutes, with no coding experience. Using ChatGPT + Replit

I spoke with many others on recommendations on how to approach this and have grouped them into themes below that I think you’ll find useful.

They’re not necessarily in order—you can go in wherever you think it’s best to help you learn.

The below breaks down as:

• 3 steps on how to learn to code with AI

• Examples of others who’ve built projects with AI-generated code

• A potential curriculum to help you learn to code with AI

• Resources to help you

• How you can build an app with AI-generated code

• Tips for others looking to use AI to write code

Subscribe to Premium to read the rest.

Become a paying subscriber of Premium to get access to this post and other subscriber-only content.

Upgrade⁴⁴⁶ Translation missing: en.app.shared.conjuction.or Sign In³³

A subscription gets you

• All premium posts (1 per week)
• Access to the full premium archive
• Request new topics to cover