In the current day, artificial intelligence (AI) has been dominating so many branches of modern society. This new transformative force has been reshaping education, social interactions, art and the economic landscape.
So what is artificial intelligence, also known as AI?
Artificial intelligence is a modern technology that utilizes machine learning, a technique that allows machines to learn from data from a number of sources and improve their performance based on that data.
A model is trained using a “training set” of data, a massive set of information usually gathered from across the internet, which allows it to make classifications and recognize patterns. Using this process, computer systems have become incredibly efficient at problem-solving and performing countless tasks.
First, an important distinction should be made. There are three general capability types of artificial intelligence: Narrow AI, Artificial General Intelligence, and Artificial Superintelligence. Even within these three categories, the latter two are still heavily theoretical as technology hasn’t evolved that far yet.
The most commonly used technology of the three is Narrow AI. Narrow AI is artificial narrow intelligence tool that is specifically designed to complete certain tasks or commands. Examples of narrow AI include self-driving cars or digital voice assistants, like Alexa. This type of AI is the one that most people might interact with on a day-to-day basis, such as when they use a search engine.
Artificial General Intelligence describes a form of AI that can act as highly adaptable assistants. This form of AI would be able to perform similarly to a human being and be able to complete a wide range of tasks.
Artificial Superintelligence (Super AI) is what we think of when we imagine what AI will be like in the future in science-fiction entertainment. It is a type of AI that is able to think, reason, learn, feel and want, such as the AI chatbot in the 2013 film, “Her”.
The most recent major leap in narrow artificial intelligence has been the creation of Large Language Models (LLMs). AI Chatbots like ChatGPT, Claude and Gemini are among the most widely utilized LLMs.
Large Language Models are machines that use statistics to generate the most correct response to a prompt from a user. Unlike early versions of artificial intelligence that needed a large amount of labeled data which worked like examples and answers to be taught about what the expected result should be, LLMs use a less guided approach to learning. This means the machines are left to guess what the correct answer is to the unorganized data it’s given.
For example, if a model was told to predict the next word in the sentence “I like my tea with”, it would be able to figure out that correct answers include words like “sugar” or “honey” after some trial and error. After several years of progress in the training of LLMs and other AI, it has gotten more attention as it’s gotten much more impressive, leading it to be connected to more facets of modern life.
Despite how much more AI is being talked about now, it has been a thing for decades, just in different forms. In the current day, the topic of AI has turned into more of a point of contention for many because of how many finite resources it requires to function and perform tasks.
There is also debate over how the data it requires to be able to learn should be regulated and how copyright laws can be respected in the growing use of LLMs and generative artificial intelligence.
