Embedded Files
AI Defined
AI Defined
Artificial Intelligence (AI) refers to the broad field of computer science focused on creating systems capable of performing tasks that typically require human intelligence. These tasks include learning, reasoning, problem-solving, perception, and language understanding. AI can be categorized into two main types:
Narrow AI: Also known as weak AI, is designed for specific tasks such as voice assistants, recommendation systems, and image recognition. It performs tasks within a limited scope and cannot generalize to other areas.
General AI: Also known as strong AI or AGI (Artificial General Intelligence), aims to perform any intellectual task that a human can do. It is still largely theoretical and has not yet been achieved.
What about Generative AI (GenAI) and LLMs?
Generative AI is a subset of AI that focuses on creating new content rather than just analyzing or interpreting existing data. This type of AI can generate text, images, music, and other forms of media. See the image to the right for a list of the types of content GenAI can produce. (Image Link)
Large Language Models (LLM) are GenAI models trained on vast amounts of text data to generate human-like responses, recognize patterns, and produce coherent language based on input prompts.
How does GenAI work?
GenAI systems are designed to create new content. This can include text, images, music, and even code. These systems are built on complex algorithms and vast datasets. For instance, ChatGPT, is based on a type of neural network architecture known as transformers, which allows it to understand and generate human-like text based on the patterns it has learned from extensive training on diverse datasets (Resource).
How Generative AI Creates Content
Generative AI models like ChatGPT operate by predicting the next word in a sequence based on the context provided by the user. This involves several steps:
Training: The AI is trained on a massive set of text data, including books, articles, and websites. During this phase, it learns the statistical relationships between words and phrases.
Fine-tuning: After the initial training, the AI is fine-tuned with more specific datasets to improve its performance in particular tasks or domains. This can be done through supervised fine tuning (SFT) by humans or through reinforcement learning (RL) where the system explores its own chain of thought (CoT) for response validation and fine-tuning.
Generating Responses: When a user inputs a prompt, the AI uses its learned knowledge to generate a coherent response. It does this by selecting the most probable next word or sequence of words based on its training data (Resource).
What are AI tokens?
AI tokens are the basic units of text that large language models (LLMs), process and generate. Tokens can be entire words, parts of words, punctuation, or symbols. Typically, a token corresponds to about four characters in English.
Input: Text submitted to the model is converted into tokens. More words equal more tokens.
Output: The model generates responses token-by-token until it reaches a defined limit or stopping point.
What is the environmental impact of token use?
Processing tokens consumes computational resources (electricity, data center cooling), contributing to carbon emissions. More tokens require more computing power, which increases energy use and environmental impact. Efficient token usage reduces computing demands, lowering the carbon footprint (Resource).
AI History Timeline
AI History Timeline
The field of artificial intelligence (AI) has experienced many advancements over the past few decades, transforming from theoretical concepts to practical applications that are integrated into our everyday lives. This timeline outlines the key milestones in the development of AI, from the pioneering work of Alan Turing in the mid-20th century to the sophisticated AI systems of the 21st century. By highlighting these significant events, we can appreciate the progress and increasing impact of AI technologies.
AI History Timeline - Sheet1.pdfCurrent GenAI Programs
Current GenAI Programs
By leveraging AI technologies, educators can enhance students' learning experience, provide personalized support, and foster a more engaging and interactive educational environment. Below are 4 of the most commonly used Generative AI programs and services that are available to educators and students. Use the website links below to find more information about each program.
OpenAI GPT Models
OpenAI GPT Models
History:
Creator: OpenAI
Launch Year: 2023
Website: OpenAI GPT-4
Brief History: GPT-4 is the fourth generation of OpenAI's Generative Pre-trained Transformer. It builds on the success of GPT-3 with improved language understanding, generation capabilities, and broader applications.
Pros:
Generates highly coherent and contextually relevant text.
Supports a wide range of applications.
Improved performance over previous versions.
Cons:
High computational requirements.
Can be expensive to use at scale.
Potential for generating biased or inappropriate content.
Strengths in Education:
Content creation: Develops educational materials and exercises.
Tutoring: Provides detailed explanations and answers to student questions.
Language learning: Assists with language practice and translation.
Anthropic Claude Models
Anthropic Claude Models
History:
Creator: Anthropic
Launch Year: 2023
Website: Claude AI
Brief History: Claude Sonnet 3.5 is a language model developed by Anthropic, designed to offer safe and reliable AI interactions. It focuses on ethical considerations and alignment with human values.
Pros:
Prioritizes safety and ethical considerations.
High-quality language generation.
Designed to avoid harmful content.
Cons:
Newer model with evolving features.
Limited availability compared to other models.
Requires careful tuning for specific applications.
Strengths in Education:
Safe content generation: Produces educational content with ethical considerations.
Student support: Provides reliable and safe responses for academic help.
Research assistance: Helps in generating summaries and analyzing texts.
Google Gemini (formerly Bard)
Google Gemini (formerly Bard)
History:
Creator: Google
Launch Year: 2023
Website: Google Gemini
Brief History: Google Gemini is an advanced generative AI model developed by Google, focusing on improving natural language understanding and generation. It builds on Google's extensive research in AI and aims to provide more nuanced and context-aware interactions.
Pros:
High accuracy in language understanding.
Strong integration with Google's ecosystem.
Continuously updated with the latest AI research.
Cons:
Privacy concerns regarding data use.
Limited availability outside Google's platforms.
High computational requirements.
Strengths in Education:
Content Creation: Generates tailored educational materials and exercises.
Tutoring: Provides detailed explanations and academic support.
Research Tool: Assists in gathering and summarizing information for research projects.
Interactive Learning: Engages students with conversational exercises and simulations.
Apple Intelligence
Apple Intelligence
History:
Creator: Apple
Launch Year: 2024
Website: Apple Intelligence
Brief History: Apple Intelligence is Apple's suite of AI tools and services, designed to enhance user experience across its ecosystem. It focuses on privacy-preserving AI and integrates seamlessly with Apple devices and services.
Pros:
Strong emphasis on user privacy.
Seamless integration with Apple devices.
High-quality natural language understanding.
Cons:
Limited to Apple's ecosystem.
Less accessible to non-Apple users.
Can be expensive due to hardware requirements.
Strengths in Education:
Privacy-Preserving Learning: Ensures student data is kept secure while providing personalized educational experiences.
Content Curation: Curates high-quality educational content tailored to individual learning needs.
Accessibility: Provides tools for students with disabilities, enhancing learning inclusivity.
Interactive Learning: Uses AI to create engaging, interactive learning activities on Apple devices.
Educational Resources Incorporating AI
Educational Resources Incorporating AI
Canva uses AI to assist in design creation, offering smart suggestions for layouts, color schemes, and font pairings. For education, this can help students and teachers create visually appealing presentations, infographics, and other visual learning materials more efficiently.
Google Gemini is an AI model that can be used for various educational purposes, including personalized tutoring, content generation, and language translation. It can assist in creating tailored learning materials and providing instant answers to student queries.
Microsoft Copilot integrates AI across various Microsoft applications, including those used in education. It can help with writing assistance, data analysis in Excel, and creating PowerPoint presentations. For students and educators, it can streamline task completion and enhance productivity.
Grammarly uses AI to provide advanced writing assistance, including grammar checks, style suggestions, and plagiarism detection. In education, it helps students improve their writing skills and assists educators in providing more efficient feedback on written assignments.
Canvas, the learning management system, incorporates AI to enhance its analytics capabilities, providing insights into student performance and engagement. It can also offer personalized learning paths and automated grading for certain types of assessments.
Elicit is an AI-powered research assistant designed to help students, researchers, and academics streamline their research process. Elicit uses AI to assist with Literature search, Paper summarization, Key findings extraction, Research question exploration, Citation Assistance, and more.
Note: The terms GenAI and AI will be used interchangeably throughout, however, it is important to note that in most cases where the term AI is used, it is generative AI that is being referred to.
Disclaimer: The content within this compendium was co-created using AI programs ChatGTP and Claude Sonnet. For more information on the co-construction of knowledge using AI, please see this resource by Robertson et al. 2024 and the AI uses in Education Page.
Page updated
Report abuse