This section of our project takes our findings a step further, analyzing how ChatGPT may be influencing our lives in a broader context with its various capabilities, potentially good and bad. To do so, we integrate secondary information from different credible sources to provide deeper insights.
What does research say about ChatGPT?
In “‘What Can ChatGPT Do?’ Analyzing Early Reactions to the Innovative AI Chatbot on Twitter”, Viriya Taecharungroj, an Associate Professor of Marketing at the Mahidol University International College (MUIC), presents the various functionalities of ChatGPT in order to shed light on the general impressions and uses of ChatGPT during the initial month of its launch. In this article, researchers found that during the first month of ChatGPT’s launch, users mostly used ChatGPT for the following five domains (which were overly saturated fields based on the topic analysis) – creative writing, essay writing, prompt writing, code writing, and answering questions. Considering the tool’s frequent usage for writing and coding, we can understand that the tool is very useful in the technology and education sectors, as we examine in this project.
Researchers also mention in this article key issues with the advent of ChatGPT, including future job displacement, a rise of a new technological landscape, debates about AI ethics, and AI’s growth in terms of approaching artificial general intelligence (AGI). As Taecharungroj himself puts it, “one concern is ChatGPT’s ability to write high school essays, homework assignments, and take-home exams effectively” (Taecharungro). This defines the dramatic shift in the education system as students now have easy access to answers generated by AI chatbots. Some might argue that this is not an issue, on the grounds that ChatGPT helps users achieve goals productively and it could improve the quality of users’ work. Yet we realize that heavy and total reliance on AI technologies could bring negative consequences that impact our ability to think as humans.
In “Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning”, David Baidoo-Anu, a member of the Faculty of Education at Queen’s University in Canada, and Leticia Owusu Ansah, from the University of Cape Coast, present an overview on ChatGPT and its different benefits and drawbacks in the education space in particular. Baidoo-Anu and Ansah also argue that the tool is beneficial in many ways; but they establish that it should not completely replace human-led teaching. As the authors themselves put it, “Given the increase in AI even in workspaces, integrating generative AI tools in the classroom and teaching students how to use it constructively and safely will also prepare them to thrive in an AI-dominated work environment after school. Therefore, educators could harness generative AI models tools like the ChatGPT to support students’ learning.” Although some people believe ChatGPT might destroy education in the future, the authors show how it can used to promote education in various ways. We agree with them in the sense that it would be in our best interest to learn how to maximize the potential of this tool as we integrate it into our education system, and learn how to minimize the harms along the way. Some might completely object to the tool on the grounds that humans will lose the ability to think on their own if they resort to using the tool for almost any need they have. However, in a society where AI is making its way into every space, we might actually put ourselves at a disadvantage by refusing to learn how to utilize the tool. Therefore, we argue that it would be in our best interest to learn how to embrace this tool in an ethical manner.
In sum, then, current literature shows both positive and negative aspects regarding ChatGPT. Our team similarly found from our visualizations and sentiment analysis a lot of positive sentiment as users found the tool “productive”, “inspirational”, and “useful” in helping achieve their goals more efficiently. However, hindering these initial positive views are concerns mentioned in the articles about decline in future job prospects for white-collar and creative professionals, as some AI technologies are already replacing jobs in the creative industry. There are also numerous debates about AI justice and ethics, and considerations about the extent to which we can allow AI’s influence in our current society, as there is the looming possibility of artificial intelligence dominating our lives. Existing literature presents mixed perspectives about the launch of ChatGPT, expressing that though ChatGPT can achieve successful goals at the moment, there are broader consequences that can impact humanity in the long run.
The need for effective regulations on the use of AI technologies
The development of new AI technologies has unlocked a level of potential that has never been seen before, and thus we do not yet know what the future beholds. Therefore, it is imperative that we develop new regulations that guide the use of these technologies in a humane way and prevent decline in our own human race – an important point to make given the continuous development and advancement of technologies in the current society.
In the article “A College Student Created an App That Can Tell Whether Ai Wrote an Essay”, NPR journalist Emma Bowman covers a story about a computer science student who built GPTZero, an application that can easily detect whether a given prompt or text was written by an AI chat bot like ChatGPT or a human. Edward Tian, the creator of GPTZero, was driven to prevent the increase of AI plagiarism simultaneously occurring with the rapid development of AI technologies like ChatGPT. GPTZero uses two indicators: perplexity, which measures how complex the text is, and burstiness, which measures the variations in sentences in terms of the sentence structures or lengths. It reasons that AI generated texts tend to be less complex and more uniform with less variance, in terms of these indicators. At last, Tian argued that he was not against the use of AI technologies like ChatGPT but with such development of new technologies, we should have “safeguards” in order to adopt the technology responsibly.
While ChatGPT has numerous positive influences, unregulated use of such technologies can bring severe consequences and risks. One big problem that the development of ChatGPT poses is AI plagiarism, with students submitting AI generated responses, for instance, raising concerns about the education system. People like Tian, however, give a sense of hope and relief to these concerns by providing solutions like GPTZero. Although, like with any technology, it would not be able to guarantee complete accuracy, the development of such tools make a technology that would otherwise be risky and detrimental into something more regulated and controllable. Technologies like GPTZero provide a good example of how AI technologies like ChatGPT could be regulated.
Why should we care?
As ChatGPT was just released a few months ago and is therefore still fresh to society, individuals—as well as various corporate sectors—seem to be exploring novel functions of the product and finding them impressive. Though many users seem to be benefitting from the new technology at the moment because of its increasing efficiency in the workplace and helping make decisions faster, concerns about ChatGPT stealing existing jobs, generation of bias within the program, and inaccuracy of the results have been rising simultaneously. It is quite evident that the product is being received with mixed feelings across society, with some people excited to embrace the tool and others fighting against it. From an individual’s standpoint, one might be confused as to whether artificial intelligence technologies like ChatGPT will be doing more harm or good to society in the long run.
This project dug deeper into the sentiment surrounding ChatGPT on Twitter – as it is an accessible platform where many individuals and companies are vocal about their thoughts – and broadly analyzes the content and sentiment distribution of these tweets in order to inform readers about how the public at large is reacting to ChatGPT, as well as how it is being perceived in specific sectors such as business, education, and technology. Through our extensive sentiment analysis on ChatGPT across different industries and society as a whole, we hoped that we could inform readers of common pros and cons of the tool and potential concerns and benefits people are identifying and experiencing with the product so that readers can make a more informed decision as to whether there is a place for the tool in their own personal circumstances and if so, what potential concerns they should be aware of.
Conclusion
ChatGPT: How is it influencing our society? is a project analyzing a dataset consisting of 219,294 tweets about ChatGPT, one of the most powerful natural language processing AI tools, collected from the first month of its launch. This project explores the sentiment toward ChatGPT amongst the general population as well as the sentiment coming from specific industries in order to identify common pros, cons, and concerns people are experiencing with the tool so that based on this, readers can make an informed decision about how the tool may be utilized in one’s own circumstances.
To summarize our findings, we found a good amount of overall positive sentiment toward ChatGPT, which encourages us to take advantage of the tool in an appropriate manner as students, educators, and professionals. However, the tool is not treated equally in all sectors of industry, as we found that there is a slightly more positive sentiment in the business and technology sectors, whereas the sentiment is almost split in half between negative and neutral sentiment versus positive sentiment for the education sector.
Contrary to our initial expectation that a tool as powerful as this would certainly disrupt employment and job prospects, we were pleasantly surprised to find that ChatGPT actually seems beneficial for employment purposes in the workspace and is not as concerning as we expected. However, various other concerns such as fear about where society is headed and the lack of ethical regulations do exist. Because of this, thoughtful regulation over the uses of ChatGPT is necessary. For example, the development of new technologies such as GPTZero can provide a solution to making AI technologies like ChatGPT more controllable, not necessarily completely restricting the use of technology but regulating it to some extent so that it does not cross harsh boundaries and instead promotes ethical uses of such technologies.
Overall, there are a lot of benefits we can attain from the tool, but it must be used with caution. While we might need to modify current systems to become more compatible with the tool, we should not become paralyzed by fear of what the future holds. As explained by one of the tweets right from our dataset:
“If you think ChatGPT is going to kill off your job, just remember that: Calculators were supposed to kill off math. Fitness machines were supposed to kill off outdoor recreation. TVs were supposed to kill off the theater. No one can predict the future.”
References
Click here to access our full annotated bibliography.
