Introduction to AI 2

by Albert Major

This slide shows key figures from the 1956 Dartmouth Conference, where AI was born. Claude Shannon was a central participant, linking his work in information theory to early AI development.

This slide summarizes Shannon’s groundbreaking impact across fields—*channel capacity*, *coding theory*, *switching circuits*, *cryptography*, and *AI*. The lecture emphasized his 1948 paper *A Mathematical Theory of Communication* as a cornerstone of *information theory*, showing how his ideas still shape digital communication, encryption, and even *machine learning* today.

1. Claude Shannon: Historical Context and Importance

• Claude Shannon (1916–2001) was a mathematician and electrical engineer.
• He is considered the father of information theory.
• His 1948 paper A Mathematical Theory of Communication introduced fundamental concepts used in digital communication today.
• Shannon's work underlies technologies such as data compression, error correction, encryption, and wireless communication.
• He worked at Bell Labs, not a university, when he wrote his most important paper.
• Bell Labs was a major center for innovation in electronics, telecommunication, and computer science.
• Shannon's work influenced not just communication, but also cryptography and artificial intelligence (e.g., he built an early chess-playing machine and a maze-solving mechanical mouse).
• The AI language model "Claude" developed by Anthropic is named after Claude Shannon, highlighting his ongoing influence.

This slide introduces a key concept from the lecture: Shannon’s focus on the technical problem of communication. It asks students to identify the message being sent, the sender and receiver, and the problems like noise or distortion that can affect accurate transmission—core elements in Shannon’s information theory model.

This slide visualizes Shannon’s idea of noise and redundancy in communication. To reduce a 10% error rate, the message (image) is sent 3 times using a repetition code; the decoder applies majority vote to reconstruct the image. It shows how redundancy lowers error (to 3%) but also reduces transmission rate—a trade-off central to channel capacity and error correction in information theory.

2. Shannon’s Communication Model

Shannon's model breaks communication into clear components:

• Sender (Transmitter): Chooses the message and sends it through a channel.
• Receiver: Receives and attempts to reproduce the original message.
• Message: The content being communicated.