Artificial Intelligence (AI) Chapter 5 PDF

Artificial Intelligence Outline What is AI? Types of Intelligence History of AI Data Science & Data Analytics Machine Learning & Deep Learning Fields of AI AI Applications 2  AI vs Automation  Automation is a bunch of machines...

Artificial Intelligence Outline What is AI? Types of Intelligence History of AI Data Science & Data Analytics Machine Learning & Deep Learning Fields of AI AI Applications 2  AI vs Automation  Automation is a bunch of machines doing a repetitive task without human intervention.  AI is simulation of intelligent behavior in computers, in contrast to the natural intelligence displayed by humans.  AI helps improve the automation process by learning and improving itself.  Make selections and decisions based on different factors. 3  Artificial Intelligence (AI) Artificial Intelligence can be classified into 3 different types:  Narrow Artificial Intelligence – Still here  General Artificial Intelligence  Super Artificial Intelligence 4  Narrow Artificial Intelligence Narrow AI is good at performing a single task within a limited context/field. Examples:  Playing chess  Making purchase suggestions  Google’s translation engine 5 General Artificial Intelligence General AI (human-level AI or strong AI) entails understanding and reasoning its environment as a human would:  How you perceive things  Juggle between multiple unrelated thoughts  Making use of memories when making a decision Humans might not be able to process data as fast as computers but they can solve problems without going into the details. It’s very hard to teach a computer to invent something that isn’t there. 6  Super Artificial Intelligence When AI becomes much smarter than the best human brains in every field, including scientific creativity, general wisdom and social skills. Some opposing views of scientists:  Stephen Hawking sees the development of full artificial intelligence as the potential end of humanity.  British AI researcher Demis Hassabis, believes the smarter AI gets, the better humans will become at saving the environment, curing diseases, explore the universe, etc. 7 Generative AI (subset of AI) Train machines to generate new and original data:  Images  Text  Audio  Video Traditional AI works on existing data to recognize patterns and make predictions Generative AI learns from existing data to generate something new 8 The Turing Test You enter a room which has a computer terminal. You have a fixed period of time to type what you want into the terminal, and study the replies. At the other end of the line is either a human being or a computer system. If it is a computer system, and at the end of the period you cannot reliably determine whether it is a system or a human, then the system is deemed to be intelligent. 9  AI History 1950: The Turing Test (Imitation Game) by Alan Turing 1955: John McCarthy first coined the term Artificial Intelligence. 1956 – 1974: The golden years  Game Playing: Chess, Checkers  Machine Translation 10  AI History 1974-1979: AI Winter  AI was subject to critiques and financial setbacks  Limited Computer Power 1980: AI boom – Expert Systems to solve domain-specific problems 1997: Deep Blue (IBM) beats Garry Kasparov (world chess champion) 2012: Deep Learning Revolution 11 Subfields of AI 12 Machine Learning vs. Deep Learning 13 14  MACHINE LEARNING DEEP LEARNING A subset of AI, which is a A subset of Machine study of algorithms and Learning which learn from models to perform a task large amounts of data using without explicit Neural Networks. instructions. Example: Tossing Bot (Princeton) Example: Detect spam emails. 15 ChatGPT Comprehends human language Generates accurate responses relevant to the user prompt Uses advanced deep learning techniques Performs well in various fields:  Learning  Teaching  Translation  Marketing  Planning 16 Visual AI: Recognition & Vision Pattern Recognition: Predicting patterns with a high amount of accuracy Computer Vision: Interpret information from images/videos Examples:  Facial Recognition: Face ID  Natural Language Processing: Siri, Cortana, Alexa (Voice Recognition – communicate with humans)  Text Recognition: detect fake news by analyzing and comparing texts. Can be used to write fake news as well.  Fraud Detection AI: used by banks to detect fake bills  Fingerprint Detection 17 Midjourney Artificial Intelligence image generator that produces images based on a user-provided description 18  Data Science & Analytics Uses scientific algorithms and tools to extract knowledge and insights from structured and unstructured data 1. Knowledge representation 2. Automated Reasoning Decision Making: Accurate and effi cient analysis of a large amount of data. Replace consultants used to make decisions. Helps business companies predict future trends and outcomes. 19 Recommendation Systems These AI systems are tailored to personal use: They see what you like online Learn about you Learn from others as well Examples:  Product recommenders i.e. Amazon  Content recommenders for social media platforms i.e. Facebook and Instagram  Playlist generators for video and music services like Netflix, YouTube 20  AI in Medicine The most popular: perform accurate diagnosis. AI can be trained on multiple images of cancer and then once it has been suffi ciently trained, it can diagnose whether patients have cancer or not. (Supervised learning) Robotics in AI: used to assist doctors in surgical procedures: Da Vinci Robot Surgery Program: Cleveland Abu Dhabi 21 AI application: Vera Robot Vera is being used by employers to recruit humans Here’s how it works:  Vera is connected to different job sites  Recruiters create a detailed job description and interview questions  Vera scours online resumes and cover letters to find qualified candidates  Vera then calls applicants  It uses speech recognition to ask and answer questions about the position 22 Autonomous Cars Combination of AI techniques:  Search and planning to find the most convenient route from A to B  Computer vision to identify obstacles  Decision making under uncertainty to cope with the complex and dynamic environment. Self-driving cars (Testing Phase)  Cars communicate with each other to prevent collisions.  Accurately deliver the passengers to their locations. 23 Gaming: AI agents Determine the behavior of non-player characters in games Example: Alien: Isolation The enemy alien has 2 AI engines:  The first AI tells the alien where the player is, so it always knows where you are.  The second AI gives the alien hints as to how to get to you. The Alien learns how you play and adapts its behavior to anticipate you. 24  Smart Homes A combination of AI technologies will make this possible for ease and security. 25 UAE: AI robot to clean ocean Solar powered robot used to detect and track waste in the marine environment. – Navigate autonomously – Avoid obstacles 26