What is AI?
Artificial Intelligence is a machine's ability to learn, adapt, and change based on data/input and perceptions. AI has the ability to solve problems by finding patterns, making associations, and drawing conclusions without the need for human intervention. The Turing Test is a test of a machine’s ability to exhibit intelligence comparable to a human. Check out this video to learn more about the Turing Test and the AI agents who have come close to passing it.
Expert Systems and Machine Learning
Expert systems employ a knowledge base and inference engine to simulate intelligence, representing an early form of AI. In contrast, machine learning models don't memorize training data and don't recall it later. They adjust parameters (weights) to minimize the error measure and don't keep any record of the training data.
You can use Google's Teachable Machine to train your own machine learning model using images, sounds, or body poses. Check out one example of a Teachable Machine below. To test it, download each of the artwork images below, and upload the files into the machine. The program will test the images to determine if they were created by Monet or van Gogh.
Monet or van Gogh Machine
Machine Perception
The goal of machine perception in AI is to imitate humans' capacity to comprehend their environment, utilizing technologies like speech and image recognition. Machine learning involves the use of neural networks by computers to understand information. Artificial neural networks, also called neural networks, are designed based on hypotheses about the functionality of the human brain. These networks comprise numerous processors that work simultaneously. By examining examples, neural networks improve their ability to perform tasks without requiring task-specific programming. This process of learning is what enables neural networks to become more proficient at their tasks.
Try Quick Draw, a game in which an AI agent employs a neural network to acquire knowledge of the distinctive features of various objects or concepts that players attempt to draw, and then predicts the player's drawing. Upon launching the program, users are prompted to create a drawing within a time limit of 20 seconds. While the drawing is in progress, the computer makes guesses and continually refines its predictions, thus enhancing its accuracy over time.
Natural Language Processing, Chatbots, and Voice Experiences
Natural Language Processing (NLP) pertains to the field of AI that focuses on enabling computers to comprehend, analyze, and manipulate human language, with the ultimate goal of bridging the gap between human language and machine understanding. Chatbots are computer programs that can carry on a conversation with a human via text and/or speech. Other technology uses NLP for speech recognition to correctly identify spoken words, and voice recognition to identify the person who is speaking. Check out this video featuring the virtual assistant Google Duplex to see how far chatbot technology has come!
Generative AI
The latest AI technologies, such as those being developed my OpenAI, employs machine learning algorithms to generate videos, images, music, text, and other media modeled after examples upon which it has been trained. Generative AI harnesses various machine learning methods, such as Generative Adversarial Networks (GANs) and Generative Pre-trained Transformers (GPT). However, in addition to creating media like art, music, or graphics effects, AI can also be used to create disinformation. For example, the Obama video below shows a sample deepfake - a video created with artificial intelligence software that looks authentic but actually manipulates the face and/or voice of a person.
Big Ideas in AI for K-12 Education
These Five Big Ideas in AI can help guide you in thinking about what K-12 students need to know about AI. In the AI4K12 community, further explanation is provided about each of the five areas as they consider student competencies across the grade levels.
Spend a few minutes reading about the Five Big Ideas, downloading the draft grade band progression charts for the grade levels you work with, and exploring the resources and community of educators found at AI4K12.
AI and Big Data
Data mining involves the extraction of valuable insights from vast amounts of data. The algorithms used in data mining are specifically designed to handle large data sets, often those that accumulate over time or "big data." Google Trends can demonstrate how the process of pattern mining can uncover fresh perspectives. By utilizing Google Trends, one can analyze search trends within Google search data for a given topic of interest. Through data mining techniques, Google Trends enables users to gain valuable insights into their topic of choice.
Ethics and AI
ISTE's Hands-On AI Projects for the Classroom: A Guide on Ethics and AI is centered on the idea that building ethical AI is a collective obligation. Various stakeholders, including students, teachers, programmers, users, regulators, and investors, are responsible for shaping the direction of AI development and utilization. The guide presents a series of projects that serve as a starting point for collaborative learning between teachers and students, enabling them to broaden their understanding of AI and its potential impact on society. By engaging with these projects, students will realize their crucial role in influencing the responsible implementation of AI technology to address issues within their surroundings, communities, and the broader world.