Last year, between Christmas and New Year’s, I vibe coded my first web application. It took me one weekend of intense struggle and confusion, followed by six months of curious bewilderment about what to do next.

This holiday season, I vibe coded 4 web apps, deployed a mobile app to two new app stores, ideated on an iMessages app concept with my husband (which is now in TestFlight), and even helped a local small business owner remake his website with AI.

I did all of this in 10 days.

I’m not alone.

If you work in tech, chances are, your social graph is looking a little like this right now:

• “I spent all vacation building apps with Claude Code. Now I’m obsessed.”

• “I built a legal dataroom in NotebookLM that replaced human lawyers on a recent deal.”

• “I built an entire software suite with coding agents, without being ‘technical.’”

• “I built an app while running a marathon.”

I’m sorry. What now?

Suffice it to say that building apps is easier than ever. Indoctrinated or not, the barrier to entry is so low that you can literally build and deploy software with your voice alone while running 26.2 miles with a Macbook strapped to your back.

The only question is: What are you going to do about it?

Before we get into how to start to build apps for yourself, I’d like to set one thing straight right off the bat: You do not need to be an engineer to get building.

Here are three things in common with the people in my network I know who have gotten the most out building with AI:

1. They are willing to play with unfinished tools in public.

2. They refuse to outsource permission to job titles or social scripts.

3. They build tools in problem spaces they already understand deeply.

That’s it. You don’t need to be “technical.” You don’t need to have millions of dollars. You do need time. And you also need to know what problems you (yes, you!) are uniquely and personally equipped to solve.

A huge amount of what people call “AI slop” exists because we’re building tools without anchoring them in lived expertise. When everyone has access to the same models, the differentiator isn’t output. It’s human judgment.

Think about what makes someone compelling on Instagram. It’s not just technical skill; it’s having a point of view. The way they see sourdough bread, press-on nail art, or drone footage that no one else "gets” in quite the same way.

If you don’t use technology to amplify the judgment you already have, you’ll just produce more noise even faster.

Here’s one example.

Even though I have access to the same tools as someone like Fred Wilson (who built a legal data room with NotebookLM over the holidays), I couldn’t have built that app.

Why not?

• I don’t have deep intuition for what makes a financing document good or bad.

• I don’t have years of past deals sitting around as raw material to work from.

• I don’t have a steady stream of founders pulling me back into that problem over and over again.

In other words, tools aren’t enough. What makes something stick is the combination of judgment from lived experience, raw material to learn from, and enough external pull to force iteration.

It’s no surprise to me that so many people showed up with brilliant new ideas after the holiday season, when we are typically “off cycle” of working. Why? Because creative sparks show up when your mind has time to wander.

In his latest post, Anil Dash talks about the importance of the idea of “ambient efficiency” or “reserve capacity” that exists at larger tech companies: It’s only when people are a little bit bored that they can find the time to tap into their cognitive surplus and come up with something new and original.

This is what happened during our holiday break. The building didn’t happen because people suddenly had better tools: It happened because they briefly had a little bit of down time.

I’ve spent most of my career in small, operationally intense companies, where there’s rarely enough slack to tap into my own cognitive surplus.

It’s only through self-imposed constraints (via fractional work and founder-mode cycles) that I’ve made time to build at the edges that matter to me.

That’s how I learned to make AI work for me last year. And that’s how you can make it work for you, too.

If you’re a solo-preneur wondering how technology might actually change your life this year, start by protecting time to build, tinker, and play. (This post from Khe Hy is a great place to start if you’re serious about getting started in command line building.)

And if you’re a manager wondering why your team isn’t adopting AI faster, stop staring at the tools and starting looking at the calendar. Without time to experiment, nothing gets built.

To see what problems people are thinking about in the Build First community, check out this public library of real problems (and real AI prompts) to get you started.

Last year, between Christmas and New Year’s, I vibe coded my first web application. It took me one weekend of intense struggle and confusion, followed by six months of curious bewilderment about what to do next.

This holiday season, I vibe coded 4 web apps, deployed a mobile app to two new app stores, ideated on an iMessages app concept with my husband (which is now in TestFlight), and even helped a local small business owner remake his website with AI.

I did all of this in 10 days.

I’m not alone.

If you work in tech, chances are, your social graph is looking a little like this right now:

• “I spent all vacation building apps with Claude Code. Now I’m obsessed.”

• “I built a legal dataroom in NotebookLM that replaced human lawyers on a recent deal.”

• “I built an entire software suite with coding agents, without being ‘technical.’”

• “I built an app while running a marathon.”

I’m sorry. What now?

Suffice it to say that building apps is easier than ever. Indoctrinated or not, the barrier to entry is so low that you can literally build and deploy software with your voice alone while running 26.2 miles with a Macbook strapped to your back.

The only question is: What are you going to do about it?

Before we get into how to start to build apps for yourself, I’d like to set one thing straight right off the bat: You do not need to be an engineer to get building.

Here are three things in common with the people in my network I know who have gotten the most out building with AI:

1. They are willing to play with unfinished tools in public.

2. They refuse to outsource permission to job titles or social scripts.

3. They build tools in problem spaces they already understand deeply.

That’s it. You don’t need to be “technical.” You don’t need to have millions of dollars. You do need time. And you also need to know what problems you (yes, you!) are uniquely and personally equipped to solve.

A huge amount of what people call “AI slop” exists because we’re building tools without anchoring them in lived expertise. When everyone has access to the same models, the differentiator isn’t output. It’s human judgment.

Think about what makes someone compelling on Instagram. It’s not just technical skill; it’s having a point of view. The way they see sourdough bread, press-on nail art, or drone footage that no one else "gets” in quite the same way.

If you don’t use technology to amplify the judgment you already have, you’ll just produce more noise even faster.

Here’s one example.

Even though I have access to the same tools as someone like Fred Wilson (who built a legal data room with NotebookLM over the holidays), I couldn’t have built that app.

Why not?

• I don’t have deep intuition for what makes a financing document good or bad.

• I don’t have years of past deals sitting around as raw material to work from.

• I don’t have a steady stream of founders pulling me back into that problem over and over again.

In other words, tools aren’t enough. What makes something stick is the combination of judgment from lived experience, raw material to learn from, and enough external pull to force iteration.

It’s no surprise to me that so many people showed up with brilliant new ideas after the holiday season, when we are typically “off cycle” of working. Why? Because creative sparks show up when your mind has time to wander.

In his latest post, Anil Dash talks about the importance of the idea of “ambient efficiency” or “reserve capacity” that exists at larger tech companies: It’s only when people are a little bit bored that they can find the time to tap into their cognitive surplus and come up with something new and original.

This is what happened during our holiday break. The building didn’t happen because people suddenly had better tools: It happened because they briefly had a little bit of down time.

I’ve spent most of my career in small, operationally intense companies, where there’s rarely enough slack to tap into my own cognitive surplus.

It’s only through self-imposed constraints (via fractional work and founder-mode cycles) that I’ve made time to build at the edges that matter to me.

That’s how I learned to make AI work for me last year. And that’s how you can make it work for you, too.

If you’re a solo-preneur wondering how technology might actually change your life this year, start by protecting time to build, tinker, and play. (This post from Khe Hy is a great place to start if you’re serious about getting started in command line building.)

And if you’re a manager wondering why your team isn’t adopting AI faster, stop staring at the tools and starting looking at the calendar. Without time to experiment, nothing gets built.

To see what problems people are thinking about in the Build First community, check out this public library of real problems (and real AI prompts) to get you started.