The First Answer Is Just the Beginning — Getting More Out of ChatGPT

by Cary and Patty ~ My Simply Sage

So, you’ve tried ChatGPT. You asked it a question, and within seconds, you had an answer. Pretty impressive, isn’t it? But here’s something we’ve discovered while learning to use AI ourselves.
That first answer might be just the beginning of something much more useful. And you don’t need special training, complicated instructions, or a collection of fancy questions to find out. You just need a little curiosity.

Let’s Start With Something Around the House

We have used ChatGPT to help investigate a problem with a GFCI electrical outlet on our back patio.

You know, those outlets with the little TEST and RESET buttons, often found in kitchens, bathrooms, garages, and outdoor areas.

Suppose you ask ChatGPT:

“Why isn’t my electrical outlet working?”

It might suggest checking the circuit breaker, looking for a tripped GFCI outlet, or determining whether other outlets have lost power. All reasonable suggestions. But which one applies to your situation? That’s where the next question becomes important.

You might explain that the circuit breaker hasn’t tripped, but several outlets aren’t working.

Now ChatGPT has more information.

It (ChatGPT) can explain how one GFCI outlet may protect other outlets farther along the same circuit. Suddenly, you’re learning something about how your home’s electrical system works.

In our particular case, the problem involved this outside GFCI outlet. What I hadn’t realized was that this outlet also protected other outlets on the same circuit.

To complicate matters, I couldn’t positively identify which circuit breaker controlled the outlet.

Through our questions and answers, ChatGPT helped me better understand first, how GFCI outlets work and second, why several outlets might lose power at once.

Replacing an outlet isn’t necessarily a complicated job, but working on one without being certain the electricity is turned off is another matter entirely.

ChatGPT suggested it might be time to call an electrician.  And that’s exactly what I did.

He quickly discovered that the old GFCI had simply reached the end of its useful life. He replaced it, made the appropriate wiring connections, and everything worked again.

So, how was AI helpful?

It didn’t repair anything. But it helped me understand the situation, recognize what I didn’t know, and decide when to call someone with the proper knowledge and equipment.  Sometimes knowing when not to do something yourself is just as valuable as knowing how to do it.

The important point is that one question can lead to a better understanding of the problem and give you more information about further questions to ask and how to proceed.

Our Air-Conditioning Questions — Something Isn’t Right!

We’ve also spent time discussing our home’s air-conditioning system, filters, and the fine dust that seems to find its way into everything.  But this time, we had a problem. Our air-conditioning filters were actually being pulled out of shape by the airflow through the system.  Why was this happening?

We started asking ChatGPT questions.

Could the filters be too restrictive? What do those MERV ratings on the packaging mean? Does a higher MERV rating always mean a better filter? And could our air-conditioning system be working harder than it should?

As we continued asking questions and examining the answers, another possibility emerged.

We were also using a humidifier. Could that have something to do with the problem?

ChatGPT helped us explore how certain humidifiers can release tiny mineral particles into the air, which can accumulate in air filters. Combined with filters that were too restrictive for our system, this could reduce airflow and cause problems.  Now we were getting somewhere!  But rather than simply accepting AI’s explanation, we hired an air-conditioning specialist to investigate.  He confirmed the problem. Our filters were too restrictive, and particles from the humidifier were contributing to the buildup.  We switched to less restrictive filters, and the problem was solved.

So, what did AI accomplish?

It didn’t repair our air-conditioning system. But through a series of questions and answers, it helped us identify a possible cause, understand what was happening, and know what to discuss with the specialist.  And along the way, we learned something important: The filter with the highest MERV rating isn’t necessarily the best filter for your particular air-conditioning system.

Sometimes a better answer comes from asking a better second question, or third, or fourth!

That’s the difference between getting a quick answer…and making an informed decision.

What About This Guitar? Something’s Wrong. It Just Doesn’t Sound Right!

What’s causing it? Could it be the strings? Something in the setup? Or could the humidity in our house have something to do with it?  We live in a very dry climate, southern Nevada.

That’s where another conversation with ChatGPT begins.

We can explore how changes in humidity affect a wooden instrument, what symptoms to look for, and what Taylor (the guitar mfg.) recommends for proper care.

Then come more questions.

How dry is too dry? Would a case humidifier help? How do I measure the humidity accurately? And when should I have a guitar technician inspect the instrument?

The important thing is not to assume that humidity caused the problem. It’s to investigate the possibilities before making changes. Each question helps narrow things down. And along the way, I’m learning more about caring for an instrument I enjoy playing.

You Can Actually Have a Conversation With ChatGPT

This is something we think many people overlook.  You don’t have to start over every time you ask another question.  ChatGPT can follow the conversation and use what you’ve already discussed.

You can say things like:

  • “I don’t understand that. Can you explain it more simply?”

  • “What if I tried something different?”

  • “Are there other possible explanations?”

  • “How do you know that’s correct?”

  • “Can you walk me through it one step at a time?”

You don’t need special wording. Just talk to it as you would someone helping you figure something out. And don’t be afraid to disagree!

But What If ChatGPT Is Wrong?

It happens.

ChatGPT can misunderstand your question, overlook important information, or give you an answer that sounds convincing but isn’t correct. That’s why we believe one of the most important things you can learn about using AI is this:

Don’t automatically believe everything it tells you.

If something doesn’t sound right, question it. Ask where the information came from. Ask ChatGPT to check its answer against reliable sources. For important information, open those sources yourself and make sure they actually support the answer. And remember, asking ChatGPT to check its own work doesn’t guarantee that it will catch every mistake.

For medical, financial, legal, or potentially dangerous household problems, be especially careful. AI can help you understand information and prepare better questions, but it shouldn’t replace qualified professional advice when you need it.

You Don’t Need 180 Special Prompts

We’ve seen advertisements offering hundreds of ready-made questions, sometimes called prompts, to help people get more out of AI.

A prompt is simply what you type or say to ChatGPT. Some prepared prompts can be useful, especially when you’re learning something unfamiliar. But there’s no secret language you have to learn. You can ask a question in your own words, read the answer, and decide what you want to know next. In fact, that’s how we use ChatGPT when we’re researching articles for My Simply Sage. We ask questions. We examine the answers. Sometimes we disagree. We look for evidence. We make corrections. And quite often, one question leads to another we hadn’t even thought about. That’s when learning really begins.

A Little Curiosity Goes a Long Way

Think about how we learned things before AI came along. We asked someone a question. Maybe we read a book, looked something up, or talked to somebody who knew more about the subject. And when we didn’t understand something, we asked another question. ChatGPT gives us another way to do much the same thing, often much faster. But speed isn’t the same as understanding. The real value comes from taking a little time to think about what we’ve learned. Does the answer make sense? Is something missing? What else would be helpful to know? Could there be another explanation?

Those are the questions that move us forward.

So, Where Do We Go From Here?

The next time you use ChatGPT, try something different.

Ask your question as you normally would. Then, before closing the conversation, take another look at the answer. Maybe there’s something you don’t understand. Maybe you’d like an example. Maybe you wonder whether the information is accurate.

Ask another question.

And another, if you need to. You might be surprised where that simple conversation takes you. We certainly have been. And perhaps that’s the most useful thing we’ve learned about AI so far:

The goal isn’t simply to get an answer. It’s to understand the answer and know what to do with the information.

Simple is not where you start, simple is where you arrive!