Lecture 1 - The laws of thought
| ← Back to AICS | Next: Lecture 2 → |
This lecture is largely introductory. It summarises the history of the idea of a thinking machine, from early conceptions (Leibniz, Boole, Lovelace) to the birth of computer science. It discusses how neural networks can be trained with supervision, and how they were used to model language in the pre-LLM era.
What you need to understand:
- The history of ideas about how to build a thinking machine
- The difference between symbolic and connectionist approaches to AI
- The debate over how to model natural language
- The basic functioning of a neural network
- How RNNs can be trained to model language
Sample essay questions:
Describe how AI research and cognitive science were shaped by the debate between symbolist and connectionist approaches. Have key controversies been resolved today?
What are the principles by which we should build a thinking machine?
References:
Most of the background for this lecture can be found in books. Good starting points are: