The Joy of AI: Introduction (02:15)
Artificial intelligence is an integral part of our everyday lives. Professor Jim al-Khalil will explore how AI works, how it can learn, and the history of turning a machine into a mind.
What is Thinking? (05:25)
The aim of AI is to simulate human intelligence. In 1955, Herbert Simon and Allen Newell figure out how to mechanize thinking. The Logic Theorist solves 38 problems from "Principia Mathematica."
Chess is often perceived as the ultimate test of reasoning. In 1948, Alan Turing writes a plan for the first computer chess program; al-Khalil plays a game as the program.
Classical AI (02:23)
Businesses use systems with programmed rules to plan complex operations with maximum efficiency and economy. Humans instinctively know what they look at. The world is too complex for classical AI to make sense of the environment,
Machine Learning (03:39)
AI techniques based on learning are powerful; systems learn from raw data. Al-Khalil demonstrates how machine learning is superior to classical AI when assessing complex data.
Conscious vs. Unconscious Mind (04:21)
Machine learning may better reflect the human subconscious mind. In 1988, Hans Moravec states that AI progress appears paradoxical from a human perspective. Artificial neural networks replicate the structure of a human brain.
Artificial Neural Networks (03:20)
Prof. Anil Seth believes AI should identify the principles by which brains work. The networks must learn what to do through training.
AI Revolution (05:46)
Artificial neural networks master new tasks in all sectors of our lives including transportation. Some worry how AI will affect our lives. Prof. Paul Newman believes it will make us healthier, wealthier, and enhance our capabilities.
Experts are trying to develop neural network systems that can learn anything without human intervention; see an example with a classic Atari game.
In 2016, programmers at DeepMind create an AI system that teaches itself the game of Go. The system makes a surprising move in a game against Lee Sedol, showing that learning systems can create new ideas.
Intelligence Explosion (03:54)
Computers that outstrip human intelligence and escape our control are common in science fiction; "2001: A Space Odyssey" is a prime example. Al-Khalil discusses Jack Good's vision of future AI. Super intelligent machines may meet objectives in unexpected ways.
AI Intelligence (04:18)
Today's AI is fundamentally limited. Al-Khalil uses a trained neural network to identify dog images; it has no understanding of what the pixel patterns mean.
Conversational AI (06:15)
Roboticists work to create an AI that can behave like a companion; Al-Khalil has a conversation with Alana; Prof. Oliver Lemon explains elements of the chatbot. AI does not have enough common sense knowledge to figure out everything someone is talking about.
One-Shot Learning (04:37)
Studying toddlers can help AI progress; children learn through data and experience, and understand abstract concepts. AI is learning without intervention but is not advanced enough to learn and think like humans. Al-Khalil considers whether AI will become conscious.
Credits: The Joy of AI (00:50)
Credits: The Joy of AI
For additional digital leasing and purchase options contact a media consultant at 800-257-5126
(press option 3) or firstname.lastname@example.org.