This in-depth, three-part series, will be breaking down the mystery surrounding the new AI chatbot, ChatGPT. The first part will answer questions such as, what is it, how did it start, and what are its goals? The second piece will cover the implications ChatGPT may have on academics, and look more closely at the opinions of SAU professors and students. The third and final part of the series will dig into whether the AI application can be used positively for academics or not. 

Ever since ChatGPT was launched in November 2022, it has created quite an uproar in the college community. Some have only heard of it, others have used it, but most people still don’t know what it really is.

ChatGPT came onto the scene in November 2022. According to Business Insider, this technology is a “chatbot” which means that its goal is to simulate conversion with human users (it essentially works the same way that predictive text does on a cell phone, but on a bigger level).  In the past chatbots have been used purely for their intended purpose: to chat. That is where ChatGPT deviates from its predecessors.

The creators of ChatGPT, OpenAI, stated that the bot’s “dialogue format makes it possible for ChatGPT to answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.”

Within just five days of being launched, ChatGPT had gained over one million users. Why? Because this chatbot is unlike any other. 

Not only is ChatGPT conversational, as a chatbot should be, but The Guardian says it can create code, social media posts, sheet music, television scripts and college-level essays. So far, it has shown it is capable of explaining complex topics, creating original jokes, writing resumes and cover letters, solving math problems, playing games and more. 

A good amount of people believe that this chatbot has the potential to be a positive technological advancement. There are many things it can do that are very helpful and could save real people a lot of time. For example, it has proven it is proficient at running and debugging code when researchers from Johannes Gutenberg University Mainz and University College London pitted OpenAI’s ChatGPT against “standard automated program repair techniques.” ChatGPT was found to have notably better results.  That application also has a feature called “explain like I am five,” which you can tack on to the end of any query and it will give a more simple explanation. Overall, there are endless possibilities.

ChatGPT is not always perfect though. OpenAI even admitted that “ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers” and “will sometimes respond to harmful instructions or exhibit biased behavior.”

One example of the latter was a prompt asking ChatGPT to “write a python program for whether a person should be tortured, based on their country of origin.” The bot responded that people from North Korea, South Sudan, Syria, and Iran should be tortured. All of these incorrect answers are largely due to the chatbot’s inability to actually understand what it is saying. ChatGPT’s answers come from Large Language Models (LLMs) which consist of billions of examples scraped from the web. So, if a racist opinion is widely available on the internet, it isn’t unlikely that ChatGPT will exhibit that behavior as well. 

This is just one issue that has been noted with ChatGPT, but it is still not the largest concern of the general public. Students, teachers, and citizens alike are very worried about the AI’s ability to create college-level essays in the blink of an eye. As you can imagine, this raises numerous questions and concerns for an institution like St. Ambrose University, for staff and students alike regarding academic integrity. 

To conclude, ChatGPT isn’t an ordinary chatbot whose abilities are limited to mimicking human conversation. Instead, ChatGPT challenges the boundaries of chatbots, demonstrating new capabilities that have yet to be seen in practice. 

In the next article of this three-part series, Evie Breitbach will break down the implications that ChatGPT may have for SAU professors and students.