What Laypeople Should Know About AI

Programming Club

The terms “Artificial Intelligence” (AI) and “machine learning” are often used interchangeably, but they describe two distinct ideas. AI is a term that describes technologies enabling computers to imitate human behavior, including machine learning, large language models (LLMs), neural networks, and natural language processing. Machine learning is a subset of AI and is the mechanism through which computers can learn without being explicitly prompted. To do this, computers are trained on lots of data and learn to recognize patterns in that data. Then, it applies this pattern recognition to novel situations. In contrast, ChatGPT and other chatbots are large-language models (LLMs): AIs trained on huge amounts of text so they can imitate human speech and writing. In other words, AI is simply a predictive algorithm. While it gets more complicated for things like image and video generation, AI is, essentially, predicting each next “word” based on its training data. Part of what makes AI so powerful is its ability to find its own connections within the data; this can, however, lead to unforeseen consequences. ChatGPT and other AI tools are often presented as sources of truth, but this is far from accurate. ChatGPT, image generation models, and even simpler models can be plagued by bias. AI is at the mercy of the data that is fed to it, and unfortunately, not all of this data is of very high quality. For example, in 2018, Amazon created an AI bot to screen résumés and trained it on the current team members’ résumés. However, all of the team members were men, and when the model was applied to applicants, it rejected every single candidate who wasn’t a man. 

It is also important to note that AI has existed in different forms for many years before ChatGPT. Tools like Google Translate are technically LLMs just like ChatGPT, Gemini, and DeepSeek. Many large store websites use AI to match your search to products. You can try it out yourself by typing something like “gifts for my 9-year-old niece” into the Macy’s search bar. Even though none of the items are explicitly labeled as “gifts for nieces,” they use AI to figure out what is most similar to your prompt. 

The public misconception of AI is, in part, intentionally crafted. The sparkles icon that has become closely associated with AI is meant to invoke a sense of magic and novelty, whereas the reality of AI is purely algorithmic. Trusting AI too much can be a bad idea; using it as a friend or therapist can be dangerous because it is just an algorithm that tells you what it thinks you want to hear.