Ever wonder what a bot has to say about proton therapy for prostate cancer? I sure did! As a retired IT guy and technology fanatic, you can imagine my excitement about the recent easy access to artificial intelligence software—AI. The ChatGPT program from OpenAI has become ubiquitous seemingly overnight. Did I sign up to try it? You bet!
What exactly is it, and where does it get its information? Here is ChatGPT’s answer, in its own words:
“ChatGPT is an artificial intelligence program that is designed to respond to user input in a conversational manner. It uses advanced language processing to generate human-like responses to questions and prompts. ChatGPT is trained on a massive dataset of text data collected from the internet, including websites, books, and other written works.”
It neglected to tell me what the “GPT” stands for, so I asked a follow-up question, and it replied:
“A Generative Pre-trained Transformer (GPT) is a type of deep learning model that uses the transformer architecture and is pre-trained on a large corpus of text data. It is generative because it can generate new text based on the input it receives, rather than simply selecting from pre-determined responses. The pre-training allows the model to have a strong understanding of language, which can then be fine-tuned for specific tasks, such as language translation, text generation, and conversation.”
Okay, I’ll stop there and resist my tendency to tumble headfirst into every rabbit hole I encounter. Instead, let’s dive in and put ChatGPT to the test. It should be at least amusing and possibly useful to ask it a few questions about the topics near and dear to you and me—proton therapy and prostate cancer. I will share the results of this fun exercise in the form of a dialogue between ChatGPT and me. It supposedly communicates in a human-like conversational style, just like me. We’ll soon find out.
I tried to get ChatGPT to keep its answers brief, but it can’t seem to resist giving some long-winded responses. To be fair, I am admittedly often guilty of this, too, and I am not a bot (as far as you know). Also, in nearly every response, ChatGPT repeated the standard mandatory liability caveats—patients should consult their doctor, results depend on various factors, and so on. To make the answers more natural, I omitted those caveats and quoted only the most relevant part of each answer. Other than that, what you see here is 100% verbatim—no editing by me.
Ready? Let’s chat:
Your fight is not over, your strength is unbroken,
You’ll rise above cancer, your spirit awoken.
With every step forward, your courage shines bright,
Illuminating the path, and guide you to light.
You’ll face every challenge, with courage in your heart,
And come out on the other side, stronger from the start.
With hope in your heart and love in your soul,
You’ll beat this disease, and reclaim your goal.
So keep your head held high, and never lose faith,
You’ve got this, my friend, and you’re on the right path.
With every step forward, you’ll leave cancer behind,
And reclaim your life, with joy in your mind.
Well, that was certainly fun. Now I have three questions for you: Does ChatGPT sound human-like? Which of us sounds more like a bot, ChatGPT or me? Are there more questions I should ask ChatGPT and share in a follow-up post? Send your answers HERE. Thanks!