By Clare Aman

OpenAI Logo. Photo courtesy of Flickr user Ishmael Daro.

Two years ago, in November 2022, the company OpenAI released ChatGPT, a free artificial intelligence chatbot. Chatbots are specifically designed to simulate conversation with human users. Though ChatGPT is one of the most commonly used AI chatbots today, there are also many others, such as Grok, which is affiliated with X, Google’s Gemini, and MyAI, developed by Snapchat. AI platforms like these have a plethora of different abilities, but their main goal is to eliminate the need for humans to complete time-consuming tasks, such as making lists, inputting data, and creating content that technology is capable of doing more quickly than humans. 

OpenAI was originally founded in 2015 by a group of investors and technology industry leaders, including Tesla and SpaceX CEO Elon Musk. Musk is no longer affiliated with OpenAI, and the current CEO, Sam Altman, has held the position since 2019. There were three versions of the platform before the current model, GPT-4: GPT-1, GPT-2, GPT-3. “GPT” stands for “Generative Pre-training Transformer.”

When the latest version was released on March 14, 2023, ChatGPT quickly became a viral sensation on social media. TikTok, X, and Instagram reels were filled with thousands of videos of people showing how ChatGPT can be used to do things like write a resumé, answer questions, solve math problems, and generate pieces of writing based on a prompt. 

As people began to experiment with ChatGPT, it started to gain users rapidly. Within the first week, it had over one million users. This amount of attention caused mixed reactions and plenty of feedback. Educators were anxious that because it was so simple to use, students would take advantage of its abilities and use it to get around doing school work. TikToks circulated the internet of students raving about how ChatGPT is “plagiarism-free” and cannot be traced back to a source, which supposedly makes it easier to get away with using on writing assignments. A YouTuber under the name Drew Gooden posted a video entitled “AI is ruining the internet,” where he went into an in-depth rant about how he believes that AI has affected almost every mainstream social media platform in some way. After almost two years, ChatGPT is still raising questions and concerns as humans learn to navigate the complex world of artificial intelligence. 

AI systems are often described as learning from their users. The more users use ChatGPT, the better it gets at understanding their requests and responding accurately. In the spirit of writing this article and learning about AI, I turned to ChatGPT to help speed up my pre-writing process. 

When preparing for my interview with Middle and Upper School Instructional Technology Coach Rachael Rachau, I struggled to come up with creative and in-depth questions. I logged on to ChatGPT and typed: “Give me 15 questions to ask someone who knows a lot about AI.”  The questions that it came up with were far more meaningful and specific than anything I would have thought of. A task that could have taken 30 minutes was completed in a matter of seconds. This experience opened my eyes to how convenient and helpful artificial intelligence can be. 

AI usage in schools is a controversial topic across the nation. While there are many benefits to it, there are also worries about students using it to cheat or do their homework for them. I sat down with Rachau to get some perspective from someone who handles AI in the classroom daily. She described her job by saying, “I work with teachers to use technology in intentional ways in the classroom.” She made a point that artificial intelligence is not something that we can avoid. It is evolving at a fast pace and isn’t slowing down, so we have to work alongside it. 

Many parents and educators fear that AI can do so much that students aren’t actually learning anything for themselves and are becoming too reliant on technology. Rachau touched on this, talking about how a common question that teachers wrestle with is, “What skills are we not developing when we use this tool instead?” There is a fine line when it comes to teaching alongside technology, and Rachau strives to help teachers and students use AI responsibly and efficiently in the classroom. 

ChatGPT can individualize learning for a child in a way that a teacher cannot. Teachers might explain things a certain way in class. However, students’ brains process information differently, so some kids will inevitably struggle. AI and technology help level the playing field for students by giving them a resource that can explain concepts in multiple ways and create endless practice problems for them. ChatGPT can create study guides and practice tests, summarize essays, and come up with topics for writing assignments. There is no limit to how many times ChatGPT can execute these actions. This allows students to have access to material that can enhance their academic performance. 

The most common worry among parents and educators for students using AI is access to a tool that can be easily used to cheat. While this is a valid concern, Rachau made the point that “It’s a problem that hurts the cheater most.” If students get comfortable using AI to write their papers for them or answer pledge set questions, then they aren’t ever going to advance those skills, which will affect them most in the long run. Rachau also believes that some of the responsibility falls onto the teachers, who should adapt to changing technology and create work for students that is difficult to complete using AI. She said, “If I give you an assignment, and it can be fully and exceptionally well accomplished by AI, I also need to rethink what my assignment is.” Students using AI to get around doing work honorably is a problem, one of a few moral issues that come along with AI. 

Schools across the country are beginning to take measures to resist the impact that AI can have on learning. Since the start of the 2023-2024 school year, Collegiate has a general policy concerning the usage of AI which states that, “The use of ChatGPT and other Artificial Intelligence tools to complete class assignments and assessments is strictly prohibited unless specifically designated by the instructor.” The policy also states that this rule is in place to “ensure an academic community built upon personal responsibility, fairness, and trust.” By banning the use of AI on assignments, Collegiate aims to continue to sharpen student’s writing skills so that their abilities don’t deteriorate without practice. In his course policies documents for his English classes, Upper School English teacher and Match advisor Vlastik Svab writes, “Even with all of the technology available to us, expressing yourself and your ideas clearly and effectively is still a valuable skill.” 

There are several ethical concerns to consider when it comes to AI usage. Platforms like ChatGPT have consumed almost everything that humans have ever created and put on the internet. Rachau said, “All the AI models were technically trained on stolen goods,” and “the amount ChatGPT has consumed is an impossible figure for us to understand.” Everything on the internet that humans have written, spoken, and created is programmed into these models. It has even memorized things that aren’t published publicly, like personal information that the user shares with it, which poses another layer of moral concerns. Because of this, it has also taken in everything problematic, dark, and horrible that humans have written and said. This means that it can be biased by producing content that may be racist, sexist, homophobic, and more. AI can potentially produce awful things because it only knows what humans have given it. 

Artificial intelligence has already impacted the job market significantly. Jobs that are repetitive and structured, like data entry, customer service, and manufacturing, can be done entirely without humans through AI. We often interact with AI more than we realize. There is the facial recognition feature on many cell phones, and many online shopping websites utilize chatbots to communicate with customers, rather than a real person. This is a far more affordable way for companies to increase productivity without hiring more people. AI can perform this work more efficiently and with fewer errors than humans can. The shift from manual labor to technology is already impacting the amount of job opportunities there are for employees in those fields. 

There is a widespread fear among Americans that AI will soon have the ability to take over many other types of jobs, eliminating the need for humans in those positions at all. Many companies, such as Google, DuoLingo, and Klarna have laid off hundreds of employees due to AI’s ability to fill those roles. Journalism jobs have also been lost, since AI can write stories faster and for less money. There is still hope, though, because while AI is great at completing straightforward tasks, it lacks qualities like creativity, emotional intelligence, and decision-making abilities, which can be crucial in many fields of work. As time passes, this may result in certain jobs being redefined rather than removed completely. 

Though AI does not have a soul, and cannot form an original thought, it still has the power to create original things. By combining all of the different information and media that it has consumed, it can create art. For example, if you ask ChatGPT to create a painting in the style of Vincent Van Gogh using the image creation AI tool DALL-E, it can create something “original” that mimics his Post-Impressionist style. The concern is that ChatGPT is stealing the creativity humans have spent centuries practicing and perfecting. ChatGPT is hiding behind a facade of originality when it’s not creative at all, but just coded to respond by replicating humanity. In the same way, AI is not intelligent, which is ironic, considering its name. It is only as intelligent as the humans who have created it, meaning it is limited in its abilities by what humans have achieved.

AI-generated fake photo of Donald Trump in prison. Photo courtesy of Flickr user Thomas Quine.

Another ethical issue is that AI allows anyone to produce fake photos and videos using AI. AI makes it possible to create fake media of events that never took place, called deepfakes. This is a problem because social media can be used to spread these deepfakes, which have already started to make people question what is real and what is fake.

People have used the photo-generating features of ChatGPT and other apps, such as Wonder, Flux AI, and Vimo, to make fake political images and videos. During the presidential debate on September 10, President-Elect Donald Trump made comments about Haitian immigrants in Springfield, Ohio eating pets. Following the debate, Republicans created images of Trump saving pets from floods and of him being chased by Haitian immigrants while carrying kittens. These images were taken by some people as evidence for the remarks he made during the debate, only to find out that these images were fake. In another election-related case, someone used AI to replicate Joe Biden’s voice and made phone calls spreading misinformation to Americans across the country. Misinformation online is already rampant, and more often than not, posts on social media aren’t entirely true. News sources are less reliable than they used to be, which is problematic and causes people across the globe to have the constant fear of being misled. How AI will be used to alleviate or expand these concerns remains to be seen. 

Featured image courtesy of Ai Busted.