Skip to Main Content
Artificial intelligence (AI)
- Refers to the ability of computers and/or algorithms to complete tasks usually associated with human intelligence or behaviour (Copeland, 2024).
AI Algorithm
- Merriam-Webster (2024) define algorithms as “a step-by-step procedure for solving a problem or accomplishing some end”. In other words, it’s a set of instructions designed to carry out a specific task, usually involving a computer.
Algorithmic Bias
- Refers to when algorithms or AI systems make decisions or outputs that reflect, reproduce, or reinforce unequitable social conditions for equity-deserving groups.
- More generally, bias is defined as "a predisposition, prejudice or generalization about a group of persons based on personal characteristics or stereotypes" (Humber College, 2023).
ChatGPT
- One of the more prominent and accessible Generative AI platforms, which offers a free tier and a premium version.
Generative Artificial Intelligence
- A subset of deep learning models that are able to generate outputs in the form of text, images, music, and more content based on the prompts inputted into the platform (Martineau, 2023).
Hallucinations
- Text generation models can generate output that is nonsensical or does not respond/correspond to the input provided by a user. Such instances are referred to as hallucinations (Ji et al, 2023).
Heuristics
- Refer to the ability of computers to make educated guesses with available data to recommend solutions to a problem (Rouse, 2016).
Large Language Models
- "Large language models (LLMs) are a category of foundation models trained on immense amounts of data making them capable of understanding and generating natural language and other types of content to perform a wide range of tasks" (IBM, "What are large language models?," 2024).
Limited Memory Machines
- Limited memory AI, unlike reactive machines, can look into the past and monitor specific objects or situations over time. Then, these observations are programmed into the AI so that its actions can be performed based on both past and present moment data. But in limited memory, this data isn’t saved into the AI’s memory as experience to learn from, the way humans might derive meaning from their successes and failures. The AI improves over time as it’s trained on more data. (IBM, "Understanding the different types of artificial intelligence," 2024)
- This type of machine powers on Deep Learning, which is a branch of Machine Learning that is made up of three neural networks in attempt to model the way humans learn.
Machine Learning
- A subset of Artificial Intelligence that refers to the ability of computers (machines) to learn/adapt without being programmed to do so (Brown, 2021).
Neural Networks
- "Neural Networks are computing systems inspired by the human brain. They simulate the human brain and its basic components and processes, which include cell body (body of the neuron), dendrites (transmitters), synaptic connections (communication), axons (transmission lines), excitation and inhibition of neurons (state of the neuron), neuron activity, and massive parallelism (multiple activities at once)" (Casas, 2020).
Outputs
- Are what you receive from a generative AI system in response to a prompt, and can consist of text, images, video, or music. Generative AI systems use data it has been previously trained on by its programmer to create an output (Harvard University, 2023).
Prompts
- Typically refer to a text-based question or command that is inputted into a generative AI tool in order for it to generate an ouput (Harvard University, 2023).
Prompt Engineering
- "Prompt engineering is the practice of building and refining prompts to ensure quality output is produced by the GenAI tool. Crafting clear and effective instructions or questions for GenAI tools helps them to produce content that matches your expectations" (Deakin University Library, "GenAI prompts," 2024).
Reactive Machines
- AI machines that have no memory and are task specific (IBM, "Understanding the different types of artificial intelligence," 2024). Popular types of reactive machines are Machine Learning Models because they take large amounts of data and use it to deliver recommendations to consumers.
- Reactive Machines do not have the ability to predict future outcomes unless they have been fed the appropriate information.
- An example of a Reactive Machine is Netflix's recommendations which is powered by a Machine Learning Model which processes data and creates specific movie and television recommendations for the user.
Reinforcement Learning
- A branch of Machine Learning . It is a self-teaching system that learns from trial and error. It is always trying to achieve the best outcome. It learns the best behavior and/or path with the goal to obtain the most optimal reward. This type of model is always learning and the best decision is based on the maximum reward.
Strong AI
- A theoretical form of AI that would require human level intelligence. The term is used to describe the mindset of AI developers that seek to create machines "indistinguishable from the human mind" (IBM, "What is Strong AI?," 2024).
Supervised Learning
- Is a machine learning approach that’s defined by its use of labeled data sets. These data sets are designed to train or “supervise” algorithms into classifying data or predicting outcomes accurately. Using labeled inputs and outputs, the model can measure its accuracy and learn over time (Delua, 2021).
Unsupervised Learning
- Uses machine learning algorithms to analyze and cluster unlabeled data sets. These algorithms discover hidden patterns in data without the need for human intervention, hence, they are “unsupervised” (Delua, 2021).