Exploring the Power of ChatGPT

When I sat down to draft this post, I clicked over to ChatOpen.ai for title suggestions. However, when I arrived, the server was overwhelmed, and ChatGPT kindly told me to come back later.

As I waited, I worked on the post. Then around 9:30 pm, I was granted access. ChatGPT proudly showcased its success when asked for title suggestions: “Exploring the Power of ChatOpenAl: A Look into the Future of Al-Powered Conversations.” I modified the post title.

The recent hype about ChatGPT reminds me of the flood of articles, posts, videos, and books about social media networks and apps from 2010 – 2012. Each month there were dozens of events, happy hours, and Meetups with social media gurus sharing the cool possibilities and applications of Facebook, Twitter, Foursquare, Google+, etc. I remember thinking it was like a digital gold rush. Everyone was rushing to make money.

Overnight, dozens of books and social media experts appeared, speaking at conferences, gaining followers like a rockstar, and writing books while charging high fees to transform your company’s boring marketing into social media magic. I was part of this digital wave, however, my presentations focused on the pros and cons of social media and digital communication.

I’ve been testing ChatGPT for the past week by submitting different requests. First, my inquiries – book title suggestions for my new novel, a title for this blog, and one of my son’s college assignments. Next, I asked questions requiring advanced thinking, such as marketing strategy for using ChatGPT. The results were impressive and reminiscent of marketing 101. However, humans are essential for determining if the strategy generated from ChatGPT is viable and aligns with the company’s overall marketing strategy.

When I searched for my son’s college assignment, he did not use the information. He doesn’t have an account. I can understand why high schools and higher ed would be concerned. ChatGPT offers students condense information – it’s like having all the library’s research capabilities at your fingertips. However, the student must know how to critically think and interpret the information. I tested ChatGPT comparative analysis skills and asked it to complete an analysis of World War I and II. The results were slow to arrive and when they appeared were a fact summary of both wars—no insight or critical thinking.

The information or data ChatGPT provided would need a human to analyze across multiple disciplines – critical thinking, logic, philosophy, history, and incorporate the most important part – our human experience. As much as society would like to delegate work, data still requires our human expertise and ability to connect the dots that do not appear in the data.

Companies and organizations that seek to cut costs will use machine learning tools like ChatGPT and Jasper so employees can generate content faster. But, the more everyone uses the same system, it becomes a feedback loop unless a human modifies the content to maintain an original voice.

After a week of testing the platform, ChatGPT will be another tech tool we use, and it functions like a turbocharged Wikipedia. These intelligent bots offer bits and pieces of generic content faster, repackaged, and washed for reuse. It’s a word factory. As we’ve learned, from our use of digital tools, there are benefits, disadvantages, and serious problems. How we use these tools determine if they are beneficial or not.

I’m a silent techie and as much as my inner techie had fun this past week testing ChatGPT, I know from experience the potential risks of using any tools, digital or physical. I’m concerned that we are sacrificing our human brainstorming and creative abilities for production speed. Another concern of mine is our hyperfocus and a slight obsession with data and information without taking the time to reflect on the data and if the information is necessary.

We’re about to experience similar growing pains that occurred when social media went mainstream, and everyone was actively participating and learning at the same time. It’s about to an interesting ride as machine learning goes mainstream.

Post originally appeared on www.dillon5.com