AI Applications and Concepts Quiz

Introduction to Artificial Intelligence

• Course name: COE 292 Introduction to Artificial Intelligence
• Date: 2/4/2025
• Artificial Intelligence (AI) is the theory and development of computer systems able to perform tasks normally requiring human intelligence.

What is AI?

• AI's definition evolves with technology advancements.
• AI systems perform tasks like visual perception, speech recognition, decision-making, and translation.

Application Areas of AI

• Virtual Assistants/Chatbots
• Agriculture and Farming
• Autonomous Flying
• Retail, Shopping, and Fashion
• Security and Surveillance
• Sports Analytics
• Manufacturing and Production
• Livestock and Inventory Management
• Self-driving Cars/Autonomous Vehicles
• Healthcare and Medical Imaging Analysis
• Warehousing and Logistics Supply Chain

Natural Language

• Speech technologies (e.g., Siri)
• Automatic speech recognition (ASR)
• Text-to-speech synthesis (TTS)
• Language processing technologies
• Question answering
• Machine translation
• Web search
• Text classification, spam filtering

Computer Vision

• Image Captioning (describing images)
• Image Segmentation (dividing images into parts)
• Facial Recognition (used for security)

Robotics

• Robotics is a combination of Mechanical Engineering and AI.
• Robotics is more difficult than simulations.
• Self-driving cars and rescue robots are advancements in this field.
• Robotics has lots of applications in automation.

Decision Making

• AI helps with complex automation processes.
• Examples include: scheduling, route planning, medical diagnosis, web search engines, automated help desks, and fraud detection.

A (Short) History of AI

• 1950s: Alan Turing wrote the Turing Test
• 1960s: Marvin Minsky, and symbolic integration by James Slagle.
• 1970s: Expert systems
• 1980s-1990s: Deep Blue beats world chess champion
• 2000-present: Robotics, Computer Vision, Natural Language Processing, Deep Learning
• etc...

Acting Like a Human - Turing Test

• The Turing Test is a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.
• The test involves a human evaluator engaging in natural language conversations with both a human and a machine, without knowing which is which.

Turing Test

• Assesses an agent's ability to exhibit intelligence.
• Computer programs are considered intelligent if their responses can fool a human evaluator.
• Turing test limitations: requires controlled environment, lacks precise definition of intelligence, and needs adaptation to new advancements.

ELIZA Chatbot

• ELIZA (1966) was a computer program designed to imitate a therapist by using open-ended questions and follow-ups to deceive users.

How AI Was Viewed Over Time

• In the 1980s, there was a focus on thinking rationally.
• In the 1990s, there was a focus on acting like people.
• In the 2000s, there has been a strong focus on balancing thinking and action while acting rationally.

What is AI?

• AI is the science of making machines to think and act like humans or rationally.

Definitions

• Agent: A thing that takes inputs using sensors and acts based on these inputs.
• Utility: The preferred outcome or result from a mathematical treatment of preferred outcomes.

Rational Agent

• Rational agents act so that their outcome, or expected outcome, is the best.
• Rational agents often have clear preferences, and they maximize their utility in all possible outcomes and actions.

Designing Rational Agents

• Agents perceive and act.
• Rational agents select actions that maximize their expected utility.
• Characteristics of the environment and actions dictate strategies for selecting rational actions.

Human Brains

• The human brain is exceptionally good at making rational decisions but not perfect.
• It is difficult to replicate the complexity of human brain behavior in a machine or computer program.

What is this Course About

• This course will introduce different representations to develop AI programs that are intelligent.
• AI models are used to solve problems and make predictions.

You Will Learn

• Constructing AI algorithms (procedures) by exposing problem constraints
• Using appropriate representations for building models within perception, thinking, and actions.

AI Solution Strategies

• Use perception to solve problems
• Use "Generate and Test" strategy to solve problems

Example 1: Use of Perception to Solve Problems

• The problem involves a person needing to move a pet lion, lamb, and a bundle of grass across a river with a limited capacity boat.

Example 1 (cont.)

• The task is to represent initial and final states of the problem using perception to solve it.
• Constraints are identified (avoiding the consumption of items by each other)
• States that violate the constraints are identified.
• States that violate constraints are removed.
• Steps for achieving a solution are defined (using a "generate and test" methodology)

Example 2: Generate and Test in AI

• Finding the name of a fish using online images and identifying discriminating features (shape, size, color)

Example 2 (cont.)

• Online searches and images provide possible solutions.
• Identifying critical features to uniquely describe the fish.
• Determining successful features and filtering out nonsensical ones.

Generate and Test Properties

• The "generator" should be non-redundant (should not generate the same item twice).
• The "generator" should be informed (it should be specific about the task, not just produce pictures of random objects).
• Testing should be fast and efficient.

Summary

• Course overview of AI and key definitions and concepts.
• Two AI solving methods (use of perception and generate/test concepts) were shown.