The Philosophy Ethics Of Artificial Intelligence

This course will provide an overview of the main approaches to AI research: the laws of thought approach (GOFAI or logic-based AI), the thinking humanly or cognitive modelling approach (neural networks or connectionism), the acting humanly or Turing test approach, and the acting rationally approach (reinforcement learning). You will learn about the key concepts, theoretical underpinnings, and philosophical issues pertaining to each of these distinct approaches to AI research. In addition, you will be invited to explore the key normative issues surrounding the design and implementation of AI systems, including the value alignment problem and the ethical implications and existential risks associated with AI systems. How is Logic Theorist able to generate proofs of theorems in propositional logic? If it can generate novel and more elegant proofs, does that count as a form of creativity? How might a chatbot like ELIZA be able to pass certain conversational tests for intelligence? What distinguishes reinforcement learning from other approaches to AI research? If our aim is to chart the progress of AI, why might the exploits of AlphaZero be significant? What are some existential risks associated with the emergence of superintelligent AI? These are some of the questions that we shall be exploring. No prior philosophical or technical background is assumed.