I’m naturally fascinated by new technology. No one needs to tell me to keep up on tech trends. I do so naturally. I’ve done plenty of dabbling with Generative AI. I’ve written about it a few times on this blog already. If you missed it, you can check it out here, here, and here.

Today’s article is all about Generative AI technology adoption. Some never use it. Some play with it from time to time. Some use it daily. I’m in the latter category, but I recognize the barriers to entry. Simply put, Generative AI isn’t necessary. You can easily function without it, for now. But that very well may change.

30 years ago, people looked up phone numbers in the Yellow Pages. Everyone had a phone book sitting next to their phone in the kitchen. Fun fact: I had a job as a teenager delivering phone books.

For a while, phone books and internet searches co-existed. But slowly, over time, the phone books went away. Now, if you don’t know how to look up a phone number on the internet, you’re out of luck.

I recognize that Generative AI is the tool of the future, and over time, it will displace other ways of doing things. To demonstrate, I’ll share a non-work example that I experienced this week.

The Air Conditioner

This week, it got nice and toasty in Minnesota. For our family, it meant it was time to fire up the central A/C. However, when I flipped the switch to “cool” on the thermostat, nothing happened.

Many of you would dial up your HVAC repair company (after looking it up on the internet, not the phone book). I decided to see how far I could get with the DIY approach. For background, I have absolutely no HVAC repair experience or knowledge. Also, my A/C unit is from 1986, and the air handler is in the attic.

Given my starting point of zero knowledge, I opened up ChatGPT and started the conversation: “My home ac won’t turn on. Please help me troubleshoot.”

Some people make a big deal about “prompt engineering.” I suppose that can be important sometimes, but I can usually get where I need to go simply by iterating and clarifying.

ChatGPT asked me a bunch of clarifying questions, which helped it narrow down the problem. We went down a few rabbit trails together, but didn’t waste too much time.

Eventually, it asked me to physically inspect the air handler in the attic, which I did. Here was my next prompt:

“I can hear main AC unit turn on, but the blower doesn’t blow. If I go to the air handler and give it a whack, the blower turns on.”

I love that I can describe what I’m seeing and doing in non-technical terms, and it figures it out. Here’s the response:

“Ah — this is a classic symptom and your physical “whack” test just gave us the diagnosis. Your blower motor is likely: Failing or has a bad start capacitor, or there’s a sticking relay or worn brushes/bearings.”

It elaborated, but went on to recommend replacing the start capacitor as it was the cheapest and most likely. I ordered a replacement on Amazon for $11, and it arrived the next day.

ChatGPT taught me how to locate, discharge, and safely install the replacement capacitor. Unfortunately for me, that didn’t fix it.

Days are going by at this point. Fortunately, the history of this thread is saved in my ChatGPT history, so even though I switched subjects and interacted with it on different subjects, I could always go back to this thread and pick up where we left off with no loss of context or progress.

Interestingly, at this point, it told me that I needed to replace the blower motor. Since I had seen it spin up just fine (after the whack) and run continuously for hours, I doubted ChatGPT’s advice, so I told it that I thought it was wrong. I find it odd that it diverted from its original diagnosis, but after my nudge, it self-corrected. Then it asked me to find the fan relay. It described it in detail and I found it. It then gave me some steps to take to diagnose whether it was failing or not. Based on our tests, ChatGPT concluded that the fan relay was the culprit. Another overnight Amazon order. This time $22.

ChatGPT gave me the instructions to safely replace the relay, then I crawled into the attic one last time, and got it done. Success! I now have working central air that turns on and off as it should, without me needing to crawl into the attic to give it a whack.

So, what’s the big deal?

Now that I’ve shared this tale, what’s the point? Why is this impressive? Why is this preferable to the old-fashioned way?

Before I used ChatGPT for DIY repair help, I used to visit YouTube and online forums. YouTube can be great if you happen to find a video that someone made with your exact issue. Often, it’s not directly applicable. Even if it is, sometimes they are lengthy and hard to follow. I’ve wasted a lot of time previewing irrelevant videos in the past. Of course, the person who made the video isn’t sitting at his computer ready and waiting to answer your every question. So, you feel like you’re on your own. Similarly, online forums can be great if you find an exact match, but if you have a general question or don’t even know how to exactly phrase what’s going on, you’re going to be out of luck. Best case, you’ll be dealing with very slow, human volunteers for help.

This all started with “My home A/C won’t turn on.” I didn’t know about start capacitors or fan relays. I still don’t know much about HVAC. I only know what ChatGPT taught me in the specific context of our troubleshooting session.

That’s my point. In this case, ChatGPT was far superior to YouTube or online forums. It’s like it has already watched everything on YouTube. It already read all of the online forums. It has already read every HVAC troubleshooting manual, and it’s sitting by my side, ready to answer my every question and describe every step in vivid detail.

It wasn’t perfect, but it’s going to keep getting better.

I think people struggle to adopt new technology like Generative AI because they are still satisfied with the old way of doing things. Get less satisfied. Hopefully, you aren’t still flipping through phone books. Hopefully, you aren’t looking up how to spell words in the dictionary. Hopefully, you aren’t printing off MapQuest maps for your road trips.

Next time something breaks at home, give it a whirl.

How is this applicable to technology adoption at work?

This is extremely applicable. Specifically, it’s normal for emerging technologies to penetrate the consumer space before the business space. Adoption at home leads to awareness of what’s possible, which, in turn, creates demand to push the boundaries at work.

Perhaps you’re not the bleeding-edge AI savant at work. That’s okay. Experiment at home. Be curious. Be like me and fix your own central air unit without any prior HVAC knowledge or experience. It’s wild. Once you experience the power, you realize that the possibilities are endless. We’re just getting started.

Podcast: Play in new window | Download