How I’m fighting bias in algorithms

By Joy Buolamwini

MIT grad student Joy Buolamwini was working with facial analysis software when she noticed a problem: the software didn’t detect her face — because the people who coded the algorithm hadn’t taught it to identify a broad range of skin tones and facial structures. Now she’s on a mission to fight bias in machine learning, a phenomenon she calls the “coded gaze.” It’s an eye-opening talk about the need for accountability in coding … as algorithms take over more and more aspects of our lives.

AI Modules for K-12 Education

Welcome to Artificial Intelligence in K-12 Education

Content Developed by: Todd Edwards, Melissa Goodall, Jim Kiper, Ann Haley MacKenzie, Brady Nash, Sherrill L. Sellers, Barry Wittman Project Manager: Sherrill L. Sellers Instructional Designer (Miami Online): Jennifer Culbertson

This AI module stemmed from the Computer Science and Computational Thinking modules. During our annual review, the development team recognized that a significant update was needed to include Generative AI. As its impact has become too substantial to incorporate into the existing modules, we created a new AI-focused module designed to engage both preservice and in-service teachers. There’s a wealth of content to explore!

Each unit offers practical, hands-on learning experiences, starting and ending with “Apply” sections, while the “Learn More” sections include references and supplemental resources. To support comprehension, we’ve included knowledge checks at the end of each section. You’re encouraged to explore the modules at your own pace and according to your interest—whether to dive in deeply or skim key points. Upon completion, you’ll also receive a certificate of participation.

AI Beginner’s Glossary

Artificial Intelligence (AI) is undergoing rapid advancements and is significantly impacting a wide range of industries. Thus, it is crucial to understand the key concepts and terminology associated with this transformative technology. This glossary provides concise definitions and examples of fundamental AI terms to help enthusiasts, students, and professionals navigate the field.

1. AI Ethics – The study of moral implications and societal impacts of AI technologies, addressing issues like bias, privacy, and job displacement. Example: Debates on the ethical use of AI in surveillance and data collection.
2. Artificial Intelligence (AI) – The simulation of human intelligence processes by machines, particularly computer systems. It includes learning, reasoning, and self-correction. Example: AI-powered assistants like Siri and Alexa use AI to respond to voice commands.
3. An Algorithm – A sequence of rules given to an AI machine to perform a task or solve a problem. Common algorithms include classification, regression, and clustering.
4. Bias in AI – Systematic errors in AI models that lead to unfair or discriminatory outcomes due to biased training data. Example: AI hiring tools that favor male candidates over female candidates due to biased training data.
5. Chatbot – An AI-driven software application that interacts with users through text or speech, simulating human conversation. Example: Customer service bots on e-commerce websites that answer frequently asked questions.
6. Computer Vision – The field of AI that enables machines to interpret and analyze visual information from the world, such as images and videos. Example: Security cameras with AI-powered facial recognition use computer vision to identify individuals.
7. Data Mining – The process of discovering patterns and insights from large sets of data using AI and statistical techniques. Example: Retailers use data mining to predict shopping trends and stock inventory accordingly.
8. Deep Learning – A specialized form of machine learning using neural networks with multiple layers to analyze complex patterns in data. Example: Self-driving cars use deep learning to detect and recognize objects on the road.
9. Edge AI – AI models and computations performed locally on devices rather than relying on cloud computing, improving efficiency and privacy. Example: AI-powered voice assistants on smartphones that work without needing an internet connection.
10. Ethical AI – The development and use of AI systems that align with ethical principles, ensuring fairness, transparency, and accountability. Example: AI regulations that prevent facial recognition misuse in surveillance.
11. Explainable AI (XAI) – AI systems designed to provide human-understandable explanations of their decisions and behaviors. Example: Healthcare AI provides reasons for diagnosing a disease based on patient data.
12. Generative AI – AI models that can generate new content, including text, images, and music, based on training data (e.g., ChatGPT, DALL-E). Example: DALL-E can create realistic images from text descriptions.
13. Large Language Model (LLM) – An advanced AI model trained on vast amounts of text data to generate human-like language, such as GPT models. Example: OpenAI’s GPT-4, which can write essays, code, and answer complex questions.
14. Machine Learning (ML) – A subset of AI that enables systems to learn from data and improve their performance without explicit programming. Example: Netflix’s recommendation system uses ML to suggest shows based on viewing history.
15. Natural Language Processing (NLP) – A branch of AI focused on enabling computers to understand, interpret, and respond to human language. Example: Google Translate uses NLP to convert text from one language to another.
16. Neural Network – A computing system inspired by the structure of the human brain that processes information through interconnected nodes (neurons). Example: Facebook uses neural networks for facial recognition in photos.
17. Prompt – It’s an input that a user feeds to an AI system in order to get a desired result or output.
18. Reinforcement Learning (RL) – A learning paradigm where an AI agent interacts with an environment and learns by receiving rewards or penalties. Example: AlphaGo, the AI that beat human champions in the game Go, was trained using reinforcement learning.
19. Singularity – A hypothetical future point where AI surpasses human intelligence, potentially leading to transformative changes in society. Example: The idea of AI becoming self-aware and improving itself beyond human control.
20. Supervised Learning – A type of machine learning where a model is trained on labeled data, meaning the correct output is provided for each input. Example: A spam filter trained on emails labeled as spam or not spam.
21. Turing Test – A test proposed by Alan Turing to determine whether a machine exhibits human-like intelligence. Example: If an AI chatbot converses so naturally that humans can’t distinguish it from a real person, it may pass the Turing Test.
22. Unsupervised Learning – A type of machine learning where a model finds patterns and structures in data without predefined labels. Example: Customer segmentation in marketing, where AI groups customers based on behavior patterns.
23. Voice Recognition – Also called speech recognition, is a method of human-computer interaction in which computers listen and interpret human dictation (speech) and produce written or spoken outputs. Examples include Apple’s Siri and Amazon’s Alexa, devices that enable hands-free requests and tasks.

Source: ChatGPT generated this glossary.