For those of us raised on sci-fi movies featuring robot world domination, there is something especially chilling about the free artificial intelligence tool ChatGPT that has quickly won the heart of the nation with its conversational interaction.
What is ChatGPT?
ChatGPT, “Chat Generative Pre-trained Transformer,” is an AI (Artificial Intelligence) created as part of a group of language models known as GPT-3. These models were designed to understand and generate human-sounding answers to questions and prompts by training them on vast amounts of data, including books, articles, websites, and even social media, making humans feel like they are communicating with “a friendly and intelligent robot” to quote ChatGPT itself. ChatGPT can do this because it was trained to predict what word or string of words generally follows the last one in a given context.
OpenAI, an AI and research company, released a trial version of ChatGPT to the public on November 20, 2022, explaining that it was a simple text box that would allow people to carry on a conversation with the AI bot. Chat GPT had more than one million users in the first five days after being launched and 100 million active users in two months. For a point of reference, TikTok reached 100 million active users after nine months.
ChatGPT can be used to write essays, create apps, write codes, build resumes, write Excel formulas, summarize large amounts of information, write cover letters, write poetry, create fiction, draft letters, and more. What makes ChatGPT unique from traditional internet search engines is that ChatGPT cannot search the internet for information. Additionally, its knowledge base is limited to 2021, which wouldn’t allow ChatGPT to respond if you asked for top news stories for April 20, 2022.
What Are the Concerns about ChatGPT?
With cheating and plagiarism already rampant in schools and higher education, professors have been playing catch up, trying to figure out how to respond to the wave of AI-produced work being turned in. Traditionally, plagiarism has been defined as using someone else’s work or ideas without giving proper credit to the author. However, with ChatGPT being something and not someone, the umbrella of plagiarism leaves ChatGPT in an awkward grey area.
“Calling the use of ChatGPT to pull reliable sources from the internet ‘cheating’ is absurd,” said Brown University sophomore Jacob Gelman. “It’s like saying using the internet to conduct research is unethical. To me, ChatGPT is the research equivalent of Grammarly. I use it out of practicality, and that’s really all.” Other students think of ChatGPT as an “online encyclopedia.”
Some public school systems like New York City have already banned the use of ChatGPT and other AI-assisted tools to combat plagiarism and cheating. However, many other universities hesitate to take such a drastic step.
While some students may ask ChatGPT to write them a 1000-word book review on Pride and Prejudice, print it out, and turn it in as their own work, surprisingly, the majority of university students would be hesitant to attempt such an act. According to a survey by Best Colleges, 51% of university students reported that using ChatGPT to complete assignments and exams is cheating. 20% disagreed, and the remainder were neutral.
However, despite these sentiments, in an article by Forbes, 89% of students reported using the AI to complete homework assignments, 48% reported using the technology to take an at-home test or quiz, 53% said they had it write an essay, 22% said they had it create an outline for them, 72% of college students say ChatGPT should be banned, 82% of professors are aware of the new AI technology, and 72% of those professors are concerned about the impact it will have on cheating. Interestingly, only 34% of educators say that ChatGPT should be banned, while 66% think students should have access to the technology.
One teacher wrote anonymously on the survey: “I love that students would have access to another resource to help answer questions. Do I worry some kids would abuse it? Yes. But they use Google and get answers without explanation. It’s my understanding that ChatGPT explains answers. That [explanation] would be more beneficial.”
To further complicate the issue, Insider reported that researchers had ChatGPT take the United States Medical Licensing Exam (a three-part exam qualifying students for residency), and ChatGPT “performed at or near the passing threshold for all three exams without any training or reinforcement.”
Self-publishing authors have flooded the Amazon Kindle store with ChatGPT-authored books, releasing Youtube videos and Reddit forums explaining how to use the technology to write novels.
Some argue that the use of these technologies will result in mental atrophy, while others believe this is the natural process society goes through with any new technology. Take the calculator, for example. Many believed the introduction of the calculator would result in the decline of basic math skills. However, research done in 2012 proved this inaccurate, showing that calculators had neither a positive nor negative impact on math skills. While this may be true for mathematical skills, there is still insufficient research on this new technology to know whether it will impair the human brain.
The Future of AI
ChatGPT is only the beginning of easily accessible Artificial Intelligence, and simply ignoring the controversy will not allow leaders, educators, and citizens to handle its developments over the months and years to come. Technology advances faster and faster each year, requiring people to respond and act quickly lest they be trampled underfoot.
OpenAI announced the release of GPT-4 on March 14, a next-generation language model. This new advancement means that it is multimodal, which is that it can process both text and images. Additionally, OpenAI has released paid versions of the application, such as ChatGPT Plus, costing $20 a month, allowing full access to ChatGPT, ensuring the busiest times of the day, faster response times, and priority access to new features and improvements. It is unclear whether OpenAI will keep the free version.
While any new technology brings excitement or apprehension, the important thing to remember is that these ethical questions need definitive answers. Technology is like a train speeding down a track with no breaks, and the last thing we can do is stand aware in its way.
Edited by Sean Mulryan
