What is Artificial Intelligence (AI)?

Artificial Intelligence is a “branch of computer science that studies how to design systems to process information and perform tasks that would otherwise require human intelligence" (World Book,  2024.) This is a popular buzzword in conversations with friends, family, and coworkers, as well as something you might see/hear discussed on the news or your social media feed. Frequently the implications of this technology on our society are discussed, but usually a basic explanation of Artificial Intelligence (AI) and how it works is skipped over. 

AI has two layers that provide the framework for all its functions; “Machine learning” and “Deep learning”. Both involve the development of algorithms (problem-solving operations used by computers) that are designed and based on how the human brain makes decisions. This enables computers to  “learn” from a data set and make inferences that become increasingly accurate over time. The difference is in the scale of complexity. "Machine Learning" in the classic sense uses networks with one input layer, a hidden layer, and an output layer. Typically, these depend on data that is structured or labeled by humans for the algorithm to process it. This is considered  "supervised learning.” 

"Deep Learning" uses an input layer, many hidden layers(usually hundreds), and an output layer. These multiple layers enable "unsupervised learning", namely the automatic extraction of information from large unstructured sets of data. Both machine learning and deep learning algorithms use artificial neural networks modeled on the decision making processes of the human brain to  process huge amounts of data, then make predictions about what that data means.

Scientists have divided the field into two basic categories all types of AI can be classified under. "Weak AI", also known as "narrow AI" or "artificial narrow intelligence" (ANI) is AI that is trained on performing a specific task(s) within certain predefined parameters. This is the most common type of AI that we have today. Examples of this include Siri, Alexa, and self-driving vehicles like Tesla. Weak AI does not imply that the technology is weak, just that it is limited in scope."Strong AI" is still in its theoretical stages. Also referred to as " artificial general intelligence" (AGI), this type of AI would constitute a machine having an intelligence equal to humans and be self-aware and learn/plan for the future.The most advanced form of this type of AI would be "artificial super intelligence" (ASI) , or a computer that would surpass the intelligence and capability of humans. This of course, does not exist yet, but there are many great examples from beloved sci-fi media, like HAL from "2001: A Space Odyssey or R2-D2 and C3PO in "Star Wars".

There are many ways humans make use of  AI in our daily lives we may not even think of as using "Artificial Intelligence". The most common is on your mobile phones, many of you use speech to text on your mobile device to call someone through carplay while you are driving or using Siri on Apple Devices to convert measurements while cooking in the kitchen. If you have made use of customer service chatbots on company websites to answer frequently asked questions about products and services, that is also AI. Even meteorologists and weather broadcasters have used AI in weather forecasting to make more accurate predictions from algorithms run on supercomputers for many years!

Artificial Intelligence has come a long way since its inception. It was first conceptualized by Alan Turing in the 1950's in his paper entitled "Computing Machinery and Intelligence", where he posed the question "can machines think?". However, the term "AI" wasn't coined until 1956 at Dartmouth College by John McCarthy. He defined it as computers capable of learning and the ability to reason logically. Later that same year the "Logic Theorist", the first ever running AI software program was created by Allen Newell, J.C. Shaw, and Herbert Simon.  You may have heard of IBM's computer "Deep Blue" defeating world chess champion Garry Kasparov in 1997 or "Watson," another IBM computer who managed to beat both Ken Jennings and Brad Rutter, former Jeopardy champions, at their own game in 2011. 

More recently, in 2023, we have seen computers with "natural language processing", that gives them the ability to understand, analyze, and manipulate language as humans do. "Chat GPT" is an generative AI model, meaning that it can create new content based on a pool of pre-existing data. The rise of it's popularity caused President Biden to sign an executive order last year, directing a number of federal agencies to develop standards and best practices involving AI,  to "minimize its harms and maximize its benefits." As we continue to understand this fast growing technology and field of study, scientists continue to push the boundaries of what is possible!  

Want to learn more? Be sure to check out our online resources using the library's A to Z databases page and make sure to check our catalog for books and magazines about technology! 

Resources:

LSU LibGuide on AI

Purdue LibGuide on AI