Understanding AI: A Foundation for Responsible Use

Artificial Intelligence (AI) is rapidly reshaping how individuals, businesses and organizations work, communicate and make decisions. As these technologies become more integrated into everyday life, understanding their capabilities, limitations and ethical implications is essential to maintaining public trust and promoting responsible leadership. These principles are at the heart of the NASBA Center for the Public Trust’s (CPT) mission.

Over the next few weeks, the CPT’s Lead With Integrity blog will explore the fundamentals of AI. This first installment provides an overview of AI, introduces common terminology, explains how it works to help individuals become more informed and responsible users of this rapidly evolving technology.

What is AI?

Coined in the 1950s, AI, short for artificial intelligence, is a huge field that spans from simple computing to the generative and agentic models that are more common today.

As this series progresses, it is fundamental to lay the groundwork for common terms and ideas in AI that often arise in the discussion of AI.

Glossary

Machine Learning: Machine learning allows computers to learn without being expressly programmed. Using incredibly large data sets, computers learn to find patterns and make predictions. The better and larger the data set- the better the computer can learn.

Neural Networks: Neural Networks are a mathematical system that mimics the human brain.  These large multi-layered systems use artificial neurons to take in huge data sets and process them to identify patterns that lead to results, much like the brain does.

Large Language Models: More commonly known as LLMs, Large Language Models are one type of neural network that focuses on speech and word sequences. Many familiar AI bots are built off of LLMs.

Agentic AI:  When a company is using “AI Agents”, they are using Agentic AI. Agentic AI can work autonomously to perform tasks, automate workflows, and in many cases, answer basic customer service questions.

Generative AI: The hallmark feature of generative AI is that it generates. Generative AI generates images, text and video based on prompting from a user. Generative AI relies on huge data sets to be able to create what is being asked by comparing the prompt to patterns it has already learned.

Hallucination: Hallucinations refer to completely made up data that AI will attempt to pass of as fact.  When AI “hallucinates” it provides information that is not rooted in fact or research.

Understanding these foundational AI concepts is the first step toward using the technology responsibly. In the coming weeks, the CPT will build on this foundation by exploring how AI works and examining the ethical implications and moral responsibilities of using AI in both our personal and professional lives. 

– Alyssa Gallas, Operations Manager