From Saying to Doing: My First Impressions of Experimenting with AI Tools
A few years from now, I will surely smile at this post and realize how little I knew and how naïve I was, but I’m going to share it anyway.
There’s a long way from saying to doing. Or is there? That’s the first thing I realized when spending some hours playing around with AI tools. My thesis so far is that the effectiveness of AI technology is dependent on the resourcefulness of the human operator, as it will only be able to perform to the level of their abilities.
In my first interactions with AI, I learned that the way from saying to doing was very short (as opposed to the popular saying), but the way from thinking to expressing was not that straightforward. After spending a fair share of time on YouTube watching videos like “GPT for beginners”, “how to use chat GPT more productively”, and other tools that “will blow your mind”, I began to test some of the things on my own. I quickly realized that without a deep understanding of the subject matter, it is easy to fall into the trap of “garbage in, garbage out.”
I first tried to use mid-journey to generate an image that I had in my head. Total failure…
I quickly realized that the way from imagining to describing was not as straightforward as I had hoped. I struggled to describe the image to the machine in a way that it could depict what I wanted.
The lack of privacy in my setup for dummies made me uncomfortable, and with so many people in the chat, I had to scroll up and down constantly to check if my image had been generated. I also found myself getting distracted by images of rubber ducks in firefighter gear, mini dinos with glasses, a large open living room, and other creations being made by people in the community in real time.
Then I tested a few “AI chat” solutions and, although this experience did blow my mind, I acknowledged that again it would be my own capacity for imagining what I could ask AND my ability to putting it into words that would determine the power of the answers I got.
Clearly, the garbage in, garbage out principle is especially relevant when it comes to utilizing this type of AI technology. While these tools can be incredibly powerful, I feel that they can only be as effective as the human operator using them. Without a thorough understanding of the problem and the subject’s current state of the art, AI-generated output may be limited or even flawed. As a result, it’s essential to ask well-informed and thoughtful questions to ensure that the outputs generated by AI are relevant and useful.
While there are a growing number of tools available to help people with limited technical expertise use, I believe that nowadays, having a strong foundation in the subject matter is still critical for the effective use of AI technology. Without this foundation, individuals may not be able to ask the right and interesting questions or trust the output. In conclusion, in February 2023, the garbage in, garbage out principle remains a critical challenge for individuals and businesses that want to leverage the power of AI technology.