Large language models (LLMs) are a type of artificial intelligence algorithm. They have undergone extensive training on vast amounts of data. This makes LLMs mainly used in natural language processing tasks, allowing them to engage in conversations, write essays and poems, provide translations, and much more. Nowadays, LLMs are employed in a wide range of domains, including education, healthcare, the legal system, finance, and linguistics(Ecosystem,2023).
Evolution of language models: From ELIZA to LLMs
LLMs are trained on a huge amount of data, often referred to as corpus (Kerner, 2023). These models are composed of neural networks, which consist of interconnected nodes resembling the human brain (What Is a Neural Network?, n.d.). They can predict the probability of the next word depending on a sentence’ context (Fisher, 2023). In fact, after completing their training, LLMs are capable of producing predictions. Following the pattern it acquired during the training stage, LLMs can anticipate which token is coming next, upon request. In this context a token is a unit of text used in processing and generating language (Johnmaeda,2023). LLMs keep predicting the next token till achieving a certain length or when coming across a stop signal, such as a full stop. Nowadays, ChatGPT is one of the most known large language models.
One of the first examples of an AI language model is ELIZA, which emerged in 1966 in MIT (Tewari, 2022). Eliza was able to generate natural and conversational responses by miming human interactions. For instance, it has the ability to model human conversation, allowing it to respond to the user in a more natural and conversational way. Eliza marked the beginning of Artificial Intelligence usage and served as a base for future technological advancements. Large language models have been evolving ever since. In 1990, LLMs were able to effectuate more advanced tasks thanks to their use of neural networks for data analysis.
Application of Large Language Models in various fields
LLMs constitute a huge part of our current society and are being used in many fields. In fact, LLMs can be used in healthcare to help formulate diagnosis, identify treatments, and predict new virus variants (Ecosystem, 2023). In addition, LLMs are used in technology businesses, to describe products, write mission statements, and many other functions enhancing their operations (Ecosystem, 2023). Another field using LLMs is retail. LLMs can provide companies with customer insights and their preferences to make better decisions that benefit the company’s growth (Ecosystem, 2023).
Large Language Models and writing/ reading practices
Nowadays, Artificial Intelligence tremendously impacts reading and writing practices. In fact, it is becoming more prominent in our daily lives. A notable example of this is how LLMs are now an integral part of educational systems. Many students use AI to write their essays, draft emails, and complete their assignments (Pantelimon et al., 2021). Nevertheless, this may be detrimental since it encourages relying on external tools rather than individual thinking. Over reliance on LLMs can also hinder people’s creativity and critical thinking skills; thus, knowing how to use these tools ethically is crucial (Bai et al., 2023). In addition, it is fundamental to make sure that the generated information is correct, since LLMs can occasionally produce inaccurate responses. As such, it is crucial to use AI as a tool to support the learning process, rather than a substitute for individual work.
Personally, I generally use LLMs only for brainstorming, or for finding ideas. I input my main idea or topic in the AI tool, then wait for it to generate a response. It usually gives me keywords and sentences that are related to my main question. This is really helpful as it enhances my cognitive processes by giving me new ideas. This approach allows me to infuse an individualized touch into my work, which makes it feel more authentic.
References
Bai, L., Liu, X., & Su, J. (2023). ChatGPT: The cognitive effects on learning and memory. Brain-X, 1(3). https://doi.org/10.1002/brx2.30
Ecosystem, E. (2023, May 16). A quick introduction to the large language model (ChatGPT). Medium.https://becominghuman.ai/a-quick-introduction-to-the-large-language-model-chatgpt-a2f5f54b4d5e
Fisher, P. (2023, September 21). A (relatively) short guide to Large Language Models (LLM). GoDaddy Blog. https://www.godaddy.com/resources/skills/guide-to-large-language-models
Johnmaeda. (2023, October 24). LLM AI tokens. Microsoft Learn. https://learn.microsoft.com/en-us/semantic-kernel/prompt-engineering/tokens
Kerner, S. M. (2023). Large language models (LLMs). TechTarget. https://www.techtarget.com/whatis/definition/large-language-model-LLM
Pantelimon, F., Bologa, R., Toma, A., & Posedaru, B. (2021). The Evolution of AI-Driven Educational Systems during the COVID-19 Pandemic. Sustainability, 13 (23), 13501. https://doi.org/10.3390/su132313501
Tewari, G. (2022). Introduction to large language models. Omega Venture Partners. https://www.omegavp.com/articles/introduction-to-large-language-models/#:~:text=The%20history%20of%20large%20language,Weizenbaum%20at%20MIT%20in%201966
What is a Neural Network? (n.d.). Amazon Web Services, Inc. https://aws.amazon.com/what-is/neuralnetwork/#:~:text=A%20neural%20network%20is%20a,that%20resembles%20the%20human%20brain