MST 101d - Ch 4: Artificial Intelligence

• Artificial Intelligence


  The use of computer systems to stimulate human mental processes, such as interpreting and generating language
• AI was first coined by John McCarthy at the Dartmouth Conference, where he defined it as "the science and engineering of making intelligent machines"


  1956
• Difference between AI and Automation

  • AI can be used in automation, but most automation uses conventional software to transfer data from one location to another
  • Automation is about setting up robots to follow a set of pre-defined rules
  • Artificial Intelligence is about setting up robots to make their own decisions (though still based on human input)
• Components of Intelligence

  • Learning
  • Reasoning
  • Problem-solving
  • Perception
  • Language
• How Artificial Intelligence Works

  AI combines massive amounts of data with quick, iterative processing and intelligent algorithms to enable the software to learn automatically from patterns or features in the data
• Machine Learning


  A subset of AI that is the science of programming machines to think and act like humans without being explicitly programmed to, aiming to enable computers to learn automatically without human intervention and adjust actions, using algorithms that can make predictions through pattern recognition
• Machine Learning Categories

  
  

  Supervised Learning - Creating models that can learn from marked dataset to make predictions (e.g. spam filtering and face recognition)

  
  

  Unsupervised Learning - Given unlabeled dataset, the algorithm must find some way to learn without the guidance of humans (e.g. fraud transactions)

  
  

  Reinforcement Learning - Uses feedback data to improve the performance of a model (e.g. games like Chess and Go)
• Deep Learning


  An application of Machine Learning that uses complex algorithms and artificial neural networks (which structures and functions like brain neurons) to train a model, where data and patterns can be better perceived
• Deep Learning Applications


  • self-driving cars
  • speech recognition
  • pattern recognition
  • computer programming
  • image recognition
  • contextual recommendations
  • fact checking
• Predictive Analytics


  A statistics-based method that data analysts use to make assumptions and test records in order to predict the likelihood of a given future outcome, capturing historical trends and applying these patterns to current data to compute a specific value at a future point in time
• Predictive Analytics Applications


  • Banking
  • Healthcare
  • Human Resources (HR)
  • Marketing and Sales
  • Supply Chain
• Natural Language Processing (NLP)


  A subfield of AI that focuses on enabling machines to understand and interpret human language, used in applications such as chatbots, virtual assistants, and language translation, with the ultimate aim of making predictable organized attempts with the computers we use every day by teaching systems to understand the human language and to react intelligently
• Machine Learning Categories


  • Supervised Learning
  • Unsupervised Learning
  • Reinforcement Learning
• Speech AI


  Uses AI for voice-based technologies: automatic speech recognition (ASR), also known as speech-to-text (STT), and text-to-speech (TTS)
• Speech AI Applications


  • Call center transcription
  • Clinical note taking
  • Virtual assistants
  • Media and marketing
• Expert Systems


  A computer program that uses AI technologies to stimulate the judgment and behavior of a human or an organization that has expertise and experience in a particular field, relying on having a good knowledge base where experts add information and non-experts use the system to solve complex problems that would usually require a human expert
• Expert Systems


  • CaDet (Cancer Decision Support Tool)
  • DENDRAL (helps chemists identify unknown organic molecules)
  • DXplain (clinical support system that diagnoses various diseases)
  • MYCIN (identifies bacteria such as bacteremia and meningitis, and recommends antibiotics and dosages)
  • PXDES (determines the type and severity of lung cancer a person has)
  • R1/XCON (early manufacturing expert system that automatically selects and orders computer components based on customer specifications)
• Planning, Scheduling, and Optimization


  A system based on AI that uses planning to determine what steps to take and when to carry out a certain step in order to achieve a goal, with the common term given to an intelligent system that implements planning being 'agent'
• Robotics


  A field of AI that involves designing, building, and programming robots to perform tasks, with AI-powered robots augmented with a variety of sensors that feed them with sensing data they can analyze and act upon in real-time
• Computer Vision


  A field of AI that enables computers and systems to derive meaningful information from digital images, videos and other visual inputs and take actions or make recommendations based on that information, enabling computers to see, observe and understand
• Computer Vision Applications

  • Image classification
  • Object detection
  • Object tracking
  • Content-based image retrieval
• Computer Vision Trends


  • Autonomous Vehicles (Automotive Industry)
  • Facial Recognition (Public Security)
  • Customer Experience, Inventory Management (Retail)
  • Medical Imaging (Medicine)
  • Attendance and Engagement Monitoring (Education)
  • Tracking Systems (Fitness And Sports)
  • Precision Agriculture
  • Human Pose Estimation (Gaming)
  • Interactive Media, Smart glasses (Media & Entertainment)
  • Product Assembly, Defect Detection (Manufacturing)
  • Violations Detection, Traffic Flow Analysis (Transportation)

  
• Turing Test


  Originally called the imitation game, a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human, where the interrogator must determine which respondent is a computer and which is a human
• John McCarthy


  Known as the Father of Artificial Intelligence, in the early 1950s he wanted a new, neutral umbrella term that could collect and organize disparate research efforts into a single field focused on developing machines that could simulate every aspect of intelligence
• ELIZA


  Considered to be the first chatterbot (chatbot) natural language processing (NLP) program, created from 1964-1966 at the Massachusetts Institute of Technology (MIT) AI Laboratory, aimed at tricking its users by making them believe that they were having a conversation with a real human being, designed to imitate a therapist who would ask open-ended questions and even respond with follow-ups
• Shakey


  The world's first mobile intelligent robot with cameras and touch sensors, controlled by a large remote computer, which became an archetype from which subsequent robots were built and significantly influenced modern robotics and AI techniques, funded by the Defense Advanced Research Projects Agency (DARPA)
• Stanford Cart


  Originally built to simulate a remote-controlled moon rover, had a 45-year career of ups and downs, including redesigning and configurations, and in 1979 successfully crossed a chair-filled room without human intervention in about five hours
• Deep Blue


  A chess-playing expert system run on a unique purpose-built IBM supercomputer, development began in 1985 at Carnegie Mellon University, and on May 11, 1997 it beat the world chess champion, Garry Kasparov, after a six-game match
• Roomba


  A mobile robot for vacuuming floors, a series of autonomous robotic vacuum cleaners which have a set of sensors that enable them to navigate the floor area of a home, with a million Roombas sold by 2004.
• Siri


  A spin-off from a project developed by the SRI International Artificial Intelligence Center.

  
  

  A virtual assistant technology that uses voice queries, gesture-based control, focus-tracking and a natural-language user interface to answer questions, make recommendations, and perform actions, using the voice of Susan Bennett in US, Jon Briggs in UK, and Karen Jacobsen in Australia, with the name Siri meaning "beautiful woman who leads you to victory" in Norwegian and "secret" in Swahili.
• Alexa


  A virtual assistant technology first used in the Amazon Echo and Echo Dot, Echo Studio and Amazon Tap speakers, capable of voice interaction, music playback, making to-do lists, setting alarms, streaming podcasts, playing audiobooks, and providing weather, traffic, sports, and other real-time information, such as news, and can also control several smart devices using itself as a home automation system
• Tesla Autopilot


  A combination of adaptive cruise control and Autosteer, a lane centering function designed to keep Tesla vehicles within painted lane lines, with the Tesla Version 7.0 software release enabling Autopilot as a feature for Model S drivers, though it is not a self-driving system and the driver is still responsible for and in control of the car
• Sophia


  A humanoid robot with simulations of every major muscle in the human face, allowing her to generate expressions of joy, grief, curiosity, confusion, contemplation, sorrow, frustration, and other feelings, who was given Saudi Arabian citizenship in 2017, becoming the first robot to receive citizenship of any country, and was named the United Nations Development Program's first Innovation Champion a month later
• Artificial Narrow Intelligence (ANI)


  Also known as weak AI, designed and trained with a specific task in mind and cannot perform beyond its limitations, having a very limited or narrow range of competencies
• Artificial General Intelligence (AGI)


  Also known as strong AI, can understand and learn any intellectual task that a human being can, allowing a machine to apply knowledge and skills in different contexts
• Artificial Superintelligence (ASI)


  Surpasses human intelligence and can perform any task better than a human, including thinking, solving puzzles, making judgments, and decisions on its own
• Types of AI Based on Functionalities

  • Reactive Machine (e.g. IBM's Deep Blue)
  • Limited Memory (e.g. self-driving cars and chatbots)
  • Theory of Mind (e.g. Sophia)
  • Self-awareness (hypothetical)
• As AI becomes more prevalent in society, there are concerns about ethics and bias, with a need to ensure that AI systems are fair, transparent, and unbiased, particularly in applications such as hiring, criminal justice, and healthcare
• AI is rapidly evolving, and its future is exciting, with the potential to revolutionize various industries such as healthcare, transportation, and education, but there are also concerns about the impact of AI on jobs and society as a whole
• Types of AI Based on Capabilities

  • ANI
  • AGI
  • ASI