4 Lessons from Building My Own AI Software (and Breaking It)

Long Story Short:

I built a YouTube video summarizer in under an hour to avoid another subscription cost, then promptly broke it by trying to add too many features. This experience taught me four critical lessons that apply to both AI development and entrepreneurship:

(1) thorough planning is essential,

(2) celebrate small wins before moving forward,

(3) retrace your steps when things break, and

(4) sometimes starting over is the best path forward.

Business Want and Need to Cut Cost

Those that know me know with the size of my operation, I often complain about how much I spend on tools and software. When you're running a 7-figure service business, launching a startup like VendorCall, and diving deep into technology and writing all at once, both time and money become precious resources.

My goal this year? Cut my software costs by 50% using AI and custom tools I build myself.

So far in 2025, I've already reduced my monthly software expenses by $318 (that's over $3,800 annually). You can follow my journey at Alternative Programming.

“Cut The Cord Becomes Cut The Software”

This journey reminds me of the cord-cutting revolution that began a decade ago. Remember when cable and satellite became insanely expensive, and streaming services like Netflix ushered in an era of consumer empowerment? The same disruption is happening now with AI.

Expensive software platforms that once cost thousands can be built in days by individuals with the right knowledge, tools, and determination. It's democratizing technology in ways we've never seen before.

“YouTube is still the best university for learning”

YouTube University has been my go-to college for continuous education since I started coding. It's incredible for learning practically anything — from changing spark plugs to building software empires. But there's one major drawback that drives me crazy: excessively long videos.

I get it. YouTube is the largest video platform, and creators need to monetize their content. But sometimes, I just want to Get To The Point (GTTP). When I'm short on time or attention, I need a summary to decide if a video is worth watching fully or just to extract the essential steps.

I don't want to watch a 30-minute video for a 3-minute solution.Breaking Down the Barriers of Access

“AI can level the playing field, if you let it”

Commercial Solutions: Another Subscription?

Looking online, I found dozens of YouTube summarizer tools, some better than others. The drawback? I'd need yet another subscription on my credit card. These tools ranged from $8 to a whopping $249 monthly.

At that higher price point, I might as well watch the hour-long video with ads! Since I'm actively reducing subscriptions, not adding them, I set myself a challenge:

Build my own YouTube summarizer in 1 hour (with a 30-minute buffer).

For the project, I used Claude for coding, Google for the YouTube API, Gemini API Studio for documentation, and Lovable as my building platform. I completed the project in 58 minutes.

But the journey taught me far more than just coding. Here are four powerful lessons that apply equally to AI development and entrepreneurship: AI Your Ally

“AI isn't about replacing humans; it's about amplifying our capabilities”

Lesson 1: Planning Is Everything

As with any venture, preparation determines success. I used AI to help create project requirements, flow diagrams, designs, and resource lists. The more thoroughly you plan, the less your AI builder (or your team) has to figure out on the fly, and the more predictable your outcome.

Business Application: There's an old saying: "If you build it, they will come." But that's rarely true in business. Success comes from focusing on strategy and desired outcomes, not flashy features.

You have to look at the outcome you want — not the praise, the shiny cars, or the social validation, but the actual solution you're creating. In business as in AI development, starting with a clear vision of the end result makes all the difference.

Lesson 2: Start Small and Celebrate Wins

I built exactly what I set out to create in about 75 minutes. But then ambition kicked in — "I can do more, add more features, go further!"

And guess what happened? I broke it. Completely. I tried to add fancy features and an interactive chat component that the simple architecture couldn't support.

Business Application: As entrepreneurs, we love pushing boundaries. We reach a goal and immediately stretch toward the next one without pausing. This constant reaching leads to unnecessary problems, failures, and wasted time.

My grandmother wisely said, "Stop and relish in the greatness of you." This means you have to celebrate your wins. When you reach a milestone or objective, take the time to acknowledge that achievement. If you don't, you'll always feel like what you've accomplished isn't enough.

In "The Gap and The Gain," Dan Sullivan and Dr. Benjamin Hardy discuss measuring backward from where you started instead of fixating on an infinitely moving target in the future. I had a working application that met my needs, but because I saw more potential (and AI, like life, always suggests you need more), I broke something beautiful.

Lesson 3: Retrace Your Steps When Things Break

After breaking my creation, frustration set in. I felt defeated yet my pride wanted to push forward. I kept trying but was getting nowhere.

After taking a break, clarity returned. I decided to retrace my steps.

In AI building, you can instruct models to think "Step-By-Step" using "Chain-of-Thought" reasoning. This methodical approach helps the AI find its way when lost.

Business Application: In life and business, we have a tendency to move too fast when things are going well, not paying attention to the environment around us. While driving gets you there faster, walking teaches you more about the journey.

Speed is valuable, but excessive speed leads to crashes if you're not careful. Sometimes you need to slow down, assess the situation, and retrace your steps. If all else fails, you can begin again, which brings me to the final lesson.

"If you don't know AI, you're going to fail. Period." Mark Cuban

After multiple failed attempts to fix my broken project, I had to make the tough call. "CPR, mouth-to-mouth, chest compressions, defibrillator padded — all failed. Like a doctor in the ER, I had to call it."

Time of death: 6:18 PM, March 4th, 2025. My Scrappy YouTube Summarizer was dead.

Was I frustrated? Absolutely. Did I want to quit? Yes. But then I heard my mother's voice in my head: "Try again." She passed away last year from pancreatic cancer, but her teachings remain with me forever.

I would practice for competitions as a kid, and when I got frustrated, she refused to let me quit — "Try again," she would say. "Learn from your mistakes. If you did it once, you can do it again."

So I set my timer for another hour and began rebuilding. This time was different. I was armed with knowledge, experience, and the insights from failure. Yes, failure. You learn far more from failure than from success.

Thirty-two minutes later, I had a working tool — no fancy name or bells and whistles, just the functional solution I needed.

Business Application: This mirrors my earlier business journey. When I started, I knew nothing about business fundamentals. I was a great hustler but a terrible businessperson.

I grew to $20,000 monthly revenue and thought I had it made — the BMW 750, the Atlanta penthouse — but my foundation was built on speed and ambition rather than sound principles.

I nearly lost everything. I had to lay off my entire team and find employment elsewhere. In that moment of humiliation, my mom told me, "Don't worry about it, Johnny. You can get it back. You did it once; you can do it again."

And I did. I stepped back, educated myself on business fundamentals, took an internship, and eventually learned to code through the IVMF Onward to Opportunity program. I became one of the first certified Cassandra Database developers in the world! I rebuilt my business, which has now generated over $8 million in revenue.

The Power of Starting Fresh

Entrepreneurs and high-achievers often obsess over their failures and flaws. They get too deep in the weeds and too hard on themselves about mistakes. It's not healthy, and it doesn't get you any further — in fact, it holds you back.

It's okay to step back. It's okay to start over. Most times, it means you'll come back stronger, faster, and with a greater sense of purpose.

Possibility Begins With Action

Building my scrappy YouTube summarizer taught me that in both AI development and entrepreneurship, the principles that lead to success are surprisingly similar:

1. Plan thoroughly — Know exactly what you're building and why

2. Celebrate small wins — Don't rush past achievements without acknowledging them

3. Know when to step back — Sometimes you need to retrace your steps to move forward

4. Be willing to start fresh — Beginning again isn't failure; it's applying hard-won wisdom

It's okay to start over. It's okay to fail. But don't give up!

What about you? Are you trying to cut costs with AI or custom tools? Have you experienced the frustration of breaking something you built? I'd love to hear your stories — hit reply and let me know!

Want more insights on using mindset, AI, and automation to achieve your goals? Subscribe to Alternative Programming for weekly strategies and inspiration. If you'd like to try my YouTube summarizer tool for yourself, send me a message!

If you're ready to harness the power of AI for your business or personal brand, visit Inspired By Brands—where we help businesses implement AI solutions to increase efficiency, improve competitiveness, and unlock new opportunities. Reach out today through LinkedIn or our website to see how we can help you stay ahead in the AI age.