#153 (Non) Profiting From AI

Welcoming you back to the new-and-improved On The Fly, with a more personal and focused approach for 2026 and beyond.

Our mission is more streamlined now: we want to bridge the gap between humanity and artificial intelligence to foster both personal and professional growth.

There’ll still be money talk, and advice from experts in various fields, and *most* of the aspects of OTF you’ve come to expect, but the main goal moving forward is to analyze and assess - two crucial skills developed in my career as an underwriter - how our human abilities and access to these incredible tools can complement each other (rather than one replacing the other.)

So in every edition, that’ll be the primary focus, delivered in a more accessible, personal way: humanity 🤝 AI.

Hope you enjoy!

AN AI STORY

How a Retired Volunteer Uses AI to Solve Problems

This isn’t a Silicon Valley AI story.

It’s a story shared by one of our readers, a retiree in North Carolina who works at a non-profit bike shop and uses AI as a much-needed assistant.

Tim spent over 30 years in corporate America working in the food and beverage industry. Upon retiring, he began volunteering at Special Pedals, a non-profit bike shop “working towards an inclusive community where adults with disabilities are employed at jobs that offer equitable hours, pay, and quality of life.”

He wanted to give back while participating in something he’s passionate about (cycling), and this was the perfect opportunity.

But it’s not without its challenges. Special Pedals receives, assesses, rebuilds, and resells donated bikes (not to mention offering tune-up and repair services). This can mean a lot of work, and like many other non-profits, they need to find ways to get the work done with limited staff, resources, and budget.

That’s where AI - or, more specifically, Perplexity AI, a similar chatbot to ChatGPT - entered the picture for Tim.

Example: a customer wanted his old bike rebuilt exactly as it had been when new, with every component accurate to the year and manufacturer. This would be a tough job for such an old bike.

Obstacle #1: Lack of Information - no one could find a complete schematic for the bike, a list of parts, where those parts could be ordered, etc.

That’s light work for Perplexity! Tim opened the app on his phone, typed in what he needed, and boom! He received the:

• Model

• Years of Production

• Correct Components Needed

• Correct Wheel Set

Obstacle #2: Wrong Manufacturer - The customer insisted that listed parts were not right. Some parts were supposed to come from a different company.

Even though Tim disagreed, he decided to follow-up with Perplexity. Turns out that this one year, this bike got its crankset from a different manufacturer! What are the odds? Now, instead of arguing with a customer and insisting on the wrong parts, Tim was back on the right path thanks to Perplexity.

Obstacle #3: Part No Longer in Production - So Perplexity manufactured the part!

Just kidding. But it did give background on when the original manufacturer stopped production and offer alternatives for acquiring the new part (the customer did not want old / recycled parts, so this was a necessity). Problem solved. Another W for Perplexity.

Obstacle #4: Something’s Wrong - Special Pedals thought they were done with this bike, but suddenly, there was a problem: the bike was not functioning properly. (“The Front Derailleur was having difficulty consistently changing chainrings” - I have no idea what this means, but it sounds serious enough.)

So Tim asked Perplexity, “What size chainrings were standard for this model?” It listed different sizes that could work, and Tim quickly realized that the chainrings provided by the crankset manufacturer (see Obstacle 2) were too big.

The chainrings were downsized, the bike was finished, the last hurdle was cleared.

I want to thank Tim for sharing this story.

Tim didn’t start as some “AI power user” nor does he have experience in tech / software. He, like many of us, started small, by beginning with emails and simple searches to help around the house. But that’s all it took to get comfortable enough to solve real problems - trying, failing, adjusting, all helping him to build confidence. That confidence inspired him to turn to Perplexity which helped Special Pedals serve that customer to the best of their ability.

Also, I think there are a few different lessons to be learned from it:

1. I’ve said it before and I’ll say it again: Anyone Can Use These Tools. You don’t need a degree in computer science. You don’t need years working in tech. You don’t need to have gotten an iPad in pre-school or been immersed in tech since birth. You just need to try it. You need to get more reps in and keep messing around with it until you find a way to make it work for you.
   

2. Were there other ways to obtain the information Tim needed to help that customer? Sure. He could’ve Googled it and sifted through hundreds of results and spent hours watching YouTube videos. However, Perplexity saved him a ton of time researching and made him more efficient. This is important for all of us, but for a non-profit with limited resources? It’s essential.

3. While taking action is the most crucial step we can take forward, staying informed is a close second. There are plenty of deep-dive resources out there - Superhuman, The Rundown AI, etc. — but if you’re looking for a more accessible perspective from someone just like you, OTF’s got you covered.
   
   (Tim told me that reading about Perplexity in a past edition of OTF motivated him to start using it, and honestly, I was delighted to hear it. That’s exactly what we’re here for.)

Every time I go on YouTube, I get that ad that tries to make me feel bad for using ChatGPT like it’s Google while 14-year-olds are creating and automating their own businesses.

Maybe I’m out-of-touch, but I like LLMs for improved search. In fact, the one thing I can confidently say about AI - from both a casual user’s and a professor’s POV - is that LLMs are best used as research assistants. They’re not great at replacing original thought (because they’re incapable of original thought), and they’re not great at replacing / developing an original voice (because they talk like robots), but they are excellent at gathering information and finding answers when prompted properly.

You still need to have ideas. You still need to come up with the right questions. You still need to analyze the situation and know where to start and how to follow up. Never say never, but I doubt that part of it - the human part - will ever be replaced. And these are skills that can deteriorate if you don’t keep flexing that muscle in your skull (the brain’s not actually a muscle, but that’s besides the point).

But if you can maintain (or improve) as a critical and analytical thinker, as an inquirer, and as a prompter, tools like Perplexity are godsends.

(This is the one AI app on my phone. After brief trial periods with all the popular ones - ChatGPT, Gemini, Claude, etc. - Perplexity is my favorite for search / research.)

DID YOU KNOW…

• Perplexity AI was created in 2022.

• Perplexity describes itself as an AI “answer engine” that emphasizes cited answers: it shows sources alongside its responses so users can quickly verify information.

• Perplexity has raised funding from major Venture Capital firms. High‑profile backers include Jeff Bezos, NVIDIA, and Shopify founder Tobi Lütke, among others, helping push the company past a billion‑dollar valuation.

• Competitors include OpenAI’s ChatGPT, Google’s Gemini, Microsoft’s Copilot.

Before You Go!

If you have a story you believe could help someone learn more about AI, email us at [email protected] and tell us about it - we’d love to feature more stories like Tim’s to keep learning together.

Next week, we’ll be back with the next evolution of Hustle Hub, featuring content curated by both me and Prof Mike, plus much more!

As always, see you on Tuesday.

Find Dan on LinkedIn

You are now On The Fly & In The Know.