Uncovering ChatGPT: A Deep Dive into AI Capabilities

January 3, 2024

Kathlyn Phan

11th Grade

Fountain Valley High School



From helping you answer your homework to effortlessly generating multiple complex lines of code, ChatGPT is an all-in-one generative AI chatbot that has the solution to any problem powered by artificial intelligence (AI) and machine learning. Although this technology is relatively new as it only came out a year ago, ChatGPT has been rapidly growing in usage to roughly 100 million monthly users in only two months. The AI is continuously learning and evolving itself with every question it’s asked with the help of machine learning. With other generative AI being produced by industry giants like Facebook and Twitter, what makes ChatGPT so special? How does it manage to make itself stand out and accumulate such a vast user base? 


ChatGPT outputs answers to any question or prompt that is inputted in mere seconds, but how could this be possible? By using deep learning technology, ChatGPT “trains” its algorithm to come up with predictions of what the user is trying to say and produce a human-like response. By making educated predictions, the algorithm can increase the rate at which answers are given. With an abundance of data and extensive training on human language at its disposal, ChatGPT can extract pieces of information to formulate coherent and precise answers. ChatGPT sources its information from Wikipedia, books, articles, etc. Since its AI is forever changing by self-enhancement, ChatGPT will not output the same reply for every user even when asked the same question. Transformer neural networks contain mechanisms that tend to each segment of the user’s text input by processing and basing its response on the contents. These neural networks provide a broader context to what the user is trying to convey and lead it to perfect its performance.


After the computer system analyzes your prompt to understand what is being asked, ChatGPT will then utilize the data it has stored from numerous past trainings to provide the most appropriate and quality response as a natural language processing tool. The AI is constantly learning more and more as it is being given data and taught what responses are most effective each time that it is used. To give an example, the thumbs up or thumbs down icon located beneath every response ChatGPT provides allows the user to rate how useful or accurate the answer was. Using this feedback, the AI will add to its training to better improve itself. During the development of ChatGPT, developers would run similar tests by ranking the responses they received to aid in the advancement of its machine-learning capability. 


ChatGPT is known for how seamless and convenient the user experience is and how naturally human the responses sound. How could a machine algorithm come so close to mimicking the dialect and conversational skills of a real human being? The AI’s linguistic skills are trained with the help of online text to understand our natural language. Transcripts are also fed to the AI so that it can learn human conversational phrases. These examples provide ChatGPT with learning material of what our basic rules of speech are, so it will generate replies that give users the impression that they are talking to a real human rather than a robot. 


As we further emerge into the AI-generated realm, we are constantly pushing the boundaries of how advanced technology can become. By learning more about new and unfamiliar forms of machine learning, we can utilize these computer systems to our advantage by improving the efficiency, accuracy, and convenience of solving everyday problems. ChatGPT’s user-friendly interface and easily adaptable machine learning capabilities explore AI’s potential by providing its users with fast and quality answers using familiar human-like speech.

Reference Sources

Amazon WorkSpace. “What Is GPT AI? - Generative Pre-Trained Transformers Explained - AWS.” Amazon Web Services, Inc.


https://aws.amazon.com/what-is/gpt/#:~:text=The%20GPT%20models%20are%20transformer. Accessed 21 Dec. 2023.


Guinness, Harry. “How Does ChatGPT Work? | Zapier.” Zapier.com, 6 Sept. 2023, 


https://zapier.com/blog/how-does-chatgpt-work/


Hetler, Amanda. “What Is ChatGPT? Everything You Need to Know.” WhatIs.com


www.techtarget.com/whatis/definition/ChatGPT#:~:text=ChatGPT%20uses%20deep%20learning%2C%20a.


IBM. “What Is Natural Language Processing? | IBM.” IBM, 2023, 


www.ibm.com/topics/natural-language-processing


Porter, Jon. “ChatGPT Continues to Be One of the Fastest-Growing Services Ever.” The Verge, 6 Nov. 2023, 


www.theverge.com/2023/11/6/23948386/chatgpt-active-user-count-openai-developer-conference.


Scribbr. “Where Does ChatGPT Get Its Information From?” Scribbr


www.scribbr.com/frequently-asked-questions/chatgpt-information/#:~:text=ChatGPT%20is%20an%20AI%20language.


Semrush Team. “How Does ChatGPT Work? (Simple & Technical Explanations).” Semrush Blog, 4 Oct. 2023, 


www.semrush.com/blog/how-does-chatgpt-work/.


Somoye, Funmi Looi. “Does ChatGPT Give the Same Answers to Everyone?” PC Guide, 9 June 2023, 


https://www.pcguide.com/apps/chatgpt-same-answers/#:~:text=Well%2C%20the%20truth%20is%20that. Accessed 21 Dec. 2023.