What comes to many people's minds when they think of Artificial Intelligence (AI) is actually Large Language Model Artificial Intelligence (LLM's or LLMAI's). These types of AI are part of Generative AI, meaning they generate human-like content in the form of text, images, and, increasingly, video. LLM's are trained on huge sets of data, mostly pulled from the internet, though some LLMAI models can be trained on more specific data for specialized tasks.
The following definition is from ChatGPT 4o:
LLM AIs (Large Language Model AIs) are powerful computer programs that can understand and generate human language. Imagine it like a super-smart robot that can read, write, and answer questions by learning from tons of text, like books, websites, and articles. It doesn’t think like a human, but it can give useful responses based on the patterns and information it has learned. (Response generated October 7th, 2024).
One thing to keep in mind about LLMAI's is that they do not actually know anything. In simple terms, responses from a LLM is them sifting through their data and go through other responses found in their data. Then they string together words that appear most common to create a seemingly coherent, factual-sounding response to a user's input. LLMAI's are simply guessing at the correct response, which can often bring about an AI phenomenon called "hallucination".
An AI hallucination is where the LLM makes up facts in an effort to create a correct response to a user's input. LLMAI's can really be thought of as over-eager assistants-- they want to please the user as quickly as possible, but sometimes will get in their own way. This is why it is important to have at least a base-line understanding of the topic that you are questioning the LLMAI on or to fact-check an LLM frequently.
But how we question an LLMAI can be just as important. There are many techniques to create inputs for LLM's to get an appropriate response. The process for getting these inputs is called Prompt Engineering.
In January, 2025, Ethan Mollick assembled a good introduction to the major general-use AI chatbots.
We can only offer some generalizations that might help put your own experiments in perspective. The below list assumes a conversation consisting of 1-3 human prompts and AI replies.
Interested in Learning More?
Greater depth: What are LLM AIs?