Artificial Intelligence (AI) - Chapter 3

Artificial Intelligence (AI) - Chapter 3

• AI is a branch of computer science aimed at creating intelligent machines capable of mimicking human behavior, thought processes, and decision-making.
• AI encompasses skills like learning, reasoning, and problem-solving. Machines with AI can perform tasks without explicit programming.
• Machine learning uses algorithms to analyze data, learn from it, and make informed decisions based on that learned information.
• Deep learning is a specialized type of machine learning. It uses layered algorithms to create artificial neural networks enabling sophisticated learning capabilities.
• Deep learning is crucial for developing AI systems that exhibit highly human-like intelligence.
• An AI system functions as an agent interacting with an environment.
• The agent perceives its environment through sensors, converting environmental factors into percepts.
• The agent acts on the environment through effectors.
• Agents evaluate predictions and adapt based on assessments.
• Machine perception refers to the ability of AI systems to use sensor inputs (cameras, microphones, etc.) to understand aspects of the world (such as computer vision).
• Applications of AI include autonomous vehicles, medical diagnosis, art creation, mathematical theorem proving, playing games, search engines, virtual assistants, spam filtering, image recognition, predicting judicial decisions, and targeting online advertisements.
• Enabling technologies for AI include the internet, big data, smart mobile phones, social media, and cheaper/more powerful hardware such as GPUs.
• AI, machine learning, and deep learning are interconnected, with deep learning being a subset of machine learning, itself a subset of AI.
• AI's purpose is to develop expert systems displaying intelligent behavior and solve complex problems.
• Key goals include replicating human intelligence, mastering knowledge-intensive tasks, and establishing intelligent links between perception and action.
• AI encompasses various disciplines such as computer science, sociology, psychology, mathematics, philosophy, and biology.
• Advantages of AI include high accuracy, speed, reliability, and usefulness in high-risk environments as well as their use as digital assistants and public utilities.
• Limitations of AI encompass high costs, a lack of original creativity, an inability to "think outside the box," the inability to experience emotions, a limited scope of trained or programmed actions, and increased dependence on machines.

Enabling Technologies

• Big data
• Advancements in computer processing speed and new chip architectures
• Cloud computing and APIs
• Emergence of data science

AI Tools and Platforms

• Neural networks
• Control theory
• Programming languages
• Microsoft Azure Machine Learning
• Google Cloud Prediction API
• IBM Watson
• TensorFlow

Sample AI Applications

• Commuting (Google's AI-powered predictions, ride-sharing apps like Uber and Lyft, commercial flight autopilot)
• Email (spam filters, smart email categorization)
• Social networking (face detection, suggesting friends to tag, adding animated effects/masks)
• Online shopping (recommendations, "viewed this item also viewed" and "bought this item also bought")
• Mobile use (voice-to-text, smart assistants like "Ok Google")

Applications

• Agriculture: robotics, soil and crop monitoring, predictive analysis.
• Healthcare: faster and more accurate diagnoses.
• Education: chatbots as teaching assistants.
• Finance & E-commerce: adaptive intelligence, algorithm trading, product discovery.
• Gaming & Entertainment: recommendation systems.
• Data security: identifying software bugs and cyberattacks.
• Social media: analyzing trends and user requirements.
• Travel & Transportation: travel arrangements, travel route suggestions.
• Automotive: self-driven cars.
• Robotics: creation of smart robots capable of performing tasks with learned experiences.