If you are looking for validation that you have made the main stream, one sure way is being featured on South Park. Congratulations ChatGPT, you’ve made it.
This week, the perpetually in 4th grade group of misfits took on the challenges of school and dating in the modern day. From TikTok to tweets and messaging, keeping up on social media is tough enough. Add in the burden of a girlfriend and Stan is looking for help. His friend Clyde provides a one-word answer, “ChatGPT”. A few clicks of copying and pasting and suddenly Stan is the perfect boyfriend. Anyone familiar with South Park knows that these is a plot twist coming. And it usually involves Cartman. I won’t spoil the episode for you, but suffice it to say, ChatGPT features prominently.
Knowing there is a perspective of truth in every story, the gang lays bare some realities of our AI powered tool. For example, who should be responsible for policing ChatGPT’s usage in a educational setting? Students may begin relying too heavily on its responses instead of thinking for themselves. These context-appropriate answers to simple questions will improve over time and students might become further dependent on it.
Unfortunately, ChatGPT’s reliance on natural language processing causes it to be unable to accurately capture the nuances of complex topics. In the initial scene, the text messages seem scripted. When asked “Do you think about me all the time?” the response was “It certainly seems like all the time.” Though accurate, the messages sounded like they were written by someone for whom English was a second language. I haven’t processed the text questions myself, but that’s the basis for another post.
In addition to complex topics, AI tools can struggle with tasks that require a higher a higher emotional IQ. For example, ethical dilemmas or creative problem solving are currently more complicated than the well defined questions that ChatGPT can answer.