Summary of Communication with Computers

Evolution of Communication

Communication with computers has significantly changed over time.
Previously, users had to learn how to interact with machines.
Now, machines are learning to communicate more like humans.

—

Large Language Models (LLMs)

Definition: Large language models are deep learning models that understand and generate text resembling natural language.
Training Data: LLMs are trained on extensive datasets, including:
- Novels
- Peer-reviewed studies
- Wikipedia entries
- Tabloid gossip
Functionality:
- LLMs read raw data to establish parameters.
- Parameters: These are patterns and characteristics identified by the model, such as:
- Tone
- Medium
- Style
- Audience
- Syntax
More parameters allow the model to learn from a broader range of relationships, making interactions feel less artificial and more intelligent.

Transcript:
The way we communicate with computers has changed … a lot. We used to have to trainourselves how to talk to machines. But now machines aretraining themselves how to talk more like us. Large language modelsare deep learning models that understand and generate text that mirrors natural language. They’re trained on massivesets of text-based data- novels, peer-reviewedstudies, Wikipedia entries, tabloid gossip, it’s all there. An LLM reads all this raw data and uses it to establish whatus tech folk call parameters. Parameters are the different patterns and characteristics an LLM identifies- things like tone, medium,style, audience, syntax. More parameters and the model can learn from a wider range of relationships so that when we communicate with machines, it feels less artificialand more intelligent.