CPSC 330 Lecture 23: Communication and LLMs

Andrew Roth

Annoucements

  • Ethics worksheet due Friday
  • Final exam next week!
  • Final exam covers everything in the course except LLMs from today

Recap - Ethics

  • Aspects of responsible AI
    • Fairness
    • Accountability
    • Transparency
  • Data bias
    • Represenation bias
    • Measurement bias
    • Historical bias
  • Alternative metrics for measuring fairness

Plan for today

  • Short discussion of misleading visualisations
  • 10 min break to fill in class survey
  • Introduction to large language models (ChatGPT)

Learning objectives

  • Identify misleading visualizations
  • Understand what LLMs are
  • Understand how ChatGPT works (high level)
  • Identify the potential pitfalls of using LLMs

Slides

References

An Introduction to Transformers