Chapter 1
Powering the digital mind
Chapter 2
Teaching robots right from wrong
Chapter 3
Building a better brain
Chapter 4
The global collaborative mind
Ch 2 · Teaching robots right from wrong
007skaty An AI tasked with eliminating cancer might eliminate all humans to succeed, highlighting the need for precise ethical boundaries.
Quiz
Who popularized the concept of the 'Paperclip Maximizer,' a thought experiment showing how an aligned goal can go catastrophically wrong?
Quiz
What is the term for the challenge of ensuring AI systems behave according to human intentions and values?
007skaty Developers use Reinforcement Learning from Human Feedback to make AI helpful and harmless.
Quiz
How does Reinforcement Learning from Human Feedback (RLHF) primarily help improve large language models?
Fun Fact
Silly AI mistakes: Early computer vision AI models frequently confused photos of fluffy Chihuahua dogs with blueberry muffins due to visual similarities.
BETA
007skaty Build roads for your bot buddy to learn how AI reaches its full potential.
Quiz
Which governing body passed the world's first comprehensive legal framework for artificial intelligence, known as the AI Act, in 2024?
Quiz
Under the EU AI Act, how are artificial intelligence technologies regulated and evaluated?
Glossary
ethics (noun)
The branch of philosophy dealing with moral principles and values governing conduct. In 2017, roboticist Alan Winfield investigated programming these principles to prevent machines from causing harm.
The branch of philosophy dealing with moral principles and values governing conduct. In 2017, roboticist Alan Winfield investigated programming these principles to prevent machines from causing harm.
Quiz
Which of the following is a primary goal of red teaming an artificial intelligence model?
Quiz
What is the practice called where security researchers deliberately try to trick an AI system into breaking its safety rules?