For the last couple of years, generative AI has been slowly becoming more prominent in everyday life. Their usage ranges from writing entire essays to answering simple questions. Some people even believe that generative AI can even be used instead of a human in a conversation. A question many people have been asking is “Does the AI have the ability to take over human writing/conversation?”
I decided to attempt to answer this question and ask one of the more popular generative AI’s ‘ChatGPT’ about a topic I would like to think I know and could talk a lot about- Matthew Grey Gubler’s character in Criminal Minds, Dr. Spencer Reid. As someone who has seen all 12 seasons of the series (and rewatched it more times than I’d care to admit) I was feeling very confident that ChatGPT would not know anything I didn’t/be able to engage in a more interesting conversation with me about the character than another human. To gauge its level of knowledge I began by asking ChatGPT what it knew of the character- a generally broad question- to which it provided a lengthy and detailed response. At first glance, ChatGPT’s answer to my asking what it knew of Spencer Reid as a character seemed very complex, but after reading it I found that it didn’t say anything about Reid that I (unsurprisingly) didn’t already know. The AI touched on Dr. Reid’s “intellectual prowess, deep passion for reading and learning, and somewhat socially awkward demeanor.” as well as his development as a character over the course of the series, his relationships with other characters, and even Reid’s background and education-all of which you’d know if you simply watch the series-or read an article about the character. I then asked it a (usually conversation-engaging) question-” What do you think of the character Spencer Reid” to which it responded with “I don’t have personal opinions or feelings, but I can provide information based on the general perception of the character Spencer Reid in “Criminal Minds.” It then repeated its original analysis of the character, providing further evidence that it is, in fact, not at all engaging to ‘conversate’ with. In speaking with ChatGPT, I noticed that its responses are very analytical (obviously), as well as easy to read and understand. Interestingly, the bot did not sound human to me in its responses. While most people claim that a lot of generative AI sounds human-like in conversation, in this particular situation, I knew I was speaking to a robot.
Although some claim that AI’s are becoming more and more like humans in this particular conversation, the lack of enthusiasm or favoritism for any character in particular was very un-human-like. If I were asking a fellow Criminal Minds watcher about the series, they would almost certainly respond with either dislike or enthusiasm for Reid, or any other character, instead of providing an academic analysis. This point, I think, is pretty important when discussing the future of generative AI. Some people are worried about AI taking over writing-that they’ll replace human authors entirely. This point is addressed by Erin Kelly in her article “Warming Up to The Power of ChatGPT” She refers to the writing of ChatGPT as “The equivalent of a microwaved chicken breast. It’s cooked, but not what anyone would point to as ideal.” This point, when thinking about ChatGPT’s responses to my questions about Spencer Reid, really rings true. The bot lacked enthusiasm, favoritism, or any emotion at all when discussing a (typically) emotive topic, making it very different from engaging in a conversation about Spencer Reid with another human being- Criminal Minds fan or not.
Kelly, E. E. (2023, July 13). Warming up to the power of CHATGPT (opinion). Inside Higher Ed | Higher Education News, Events and Jobs. https://www.insidehighered.com/opinion/views/2023/07/13/warming-power-chatgpt-opinion#:~:text=If%20much%20of%20the%20boilerplate,see—and%20more%20widely%20appreciated.