The opinions expressed by Entrepreneur authors are their very own.
Generative AI has been around for over a yr, disrupting the public relations industry and leaving communicators wondering about the way forward for their work. People are uncertain, especially given all the unknowns that technology brings.
However, this fear prevents people from understanding the possibilities of artificial intelligence, leaving them feeling unable to prepare for the future. Unfortunately, many communicators lack the knowledge to accurately describe what this technology is, how it really works and what it is able to, each in terms of the organizations they represent and in terms of their very own background knowledge.
That’s why I wrote a short glossary of commonly used AI terms in plain English so that any communicator can understand what these buzzwords mean and explain what is going on on.
artificial intelligence
artificial intelligence is a technology that enables computers and machines to simulate human pondering and intelligence, in addition to solve problems at a human level.
It covers every little thing from autonomous cars to weather forecasting models, machine learning, robotics and much more. Each of those examples is a “subset” of AI, and entire articles could be written on each of them. However, given that this text is about generative AI, let’s delve into the lexicon surrounding such a AI.
To do this, we’d like to look at the “machine “learning” subset of artificial intelligence.
Machine learning
The goal of machine learning, or “ML,” is to use algorithms that can learn and generalize information. Basically, the machine learning algorithm receives information. A question is then asked and the algorithm comes up with an answer based on the information received.
Within machine learning, there are dozens of subsets. These include “decision trees” that are used in chatbots. There is “linear regression” which is useful for predicting future events based on past data reminiscent of weather models. There is also “clustering,” which is how the adtech algorithm knows when and how to sell you a product or service.
All of those subsets take the information that has been fed into them to make predictions about the future based on past events. They are all useful and influence our each day lives. However, there is one other subset of machine learning called “deep learning.” This is the subset where we discover generative artificial intelligence.
Deep learning
Deep learning means that there are greater than three layers of neural networks. The (*5*) are the brain of the algorithm, while the “layers” are the depth of pondering the algorithm can do.
In standard machine learning, there is an input layer (i.e. What will the weather be like today?); a “thinking” layer, for example taking all the wind, rain and temperature data from past events and applying it to the current situation; and then the output layer (i.e. the weather forecast will probably be sunny). All these layers form a neural network.
In deep learning, the neural network consists of greater than three layers. This allows the algorithm to think deeper and with more nuance. In fact, this deep and shallow way of pondering is where the phrases “deep AI” and “shallow AI” come from.
In addition to the difference in the variety of algorithm layers, the way information is entered into these algorithms is also different. This is because the deep learning algorithm relies on fundamental models.
Basic models
“Core models” are giant stores of knowledge, and each data point is called a “parameter.” Deep learning models are trained on these base models full of knowledge and then “tuned” to perform in a specific way. Some base models have over 1 trillion parameters.
There are several sorts of basic models, including “large language models” or “LLM”. They are called so because they are large – they will have over a trillion parameters – and are designed to process and generate normal human language. Other basic models include vision models for generating video, audio models for generating various kinds of sounds, and even biological models for predicting how proteins will interact with each other.
Core models are vital because they are huge repositories of knowledge that could be accessed by any paying subscriber. Instead of spending thousands and thousands of dollars and 1000’s of hours compiling all this data, a company can subscribe to an already existing model (reminiscent of an OpenAI model or a Google model) and use this information to train its generative AI.
AI application
These core models form the basis of “AI applications.” The application itself could be anything from a piece of the platform to a full application that tunes the underlying model in a specific way. A great analogy for AI applications is to look at how applications are developed in general.
If you look at an app in the Apple Store or Google Play, you will see that it was built to run on that particular app store’s underlying technology infrastructure. AI applications work on the same principle – they are built to work with the underlying technology infrastructure of the AI model.
So where does generative AI fit in?
“Generative AI” includes models specifically created to generate latest content. It is something that is created using a knowledge base of underlying models combined with tuning from AI applications to achieve the desired result. This is how video generators like Sora or language generators like Perplexity or ChatGPT work.
In short, generative AI is used in artificial intelligence applications that use deep learning neural networks trained on foundational models to generate a specific, never-before-seen piece of content.
It is vital for us as communicators to fully understand the terms associated with artificial intelligence so that we will enable society to understand how this world-changing technology works. We hope that PR professionals will have the opportunity to use this glossary to higher communicate what artificial intelligence is, in addition to higher understand how it may possibly be implemented in their on a regular basis lives.