Skip to main content
seanmcp.com

Thinking Outside the Bots

Slides

Overview: The explosion of large-language models in the past few years has turned everyone’s attention to AI. OpenAI’s ChatGPT has pioneered a new pattern for human-model interactions that other companies have been eager to follow. But is the chat-bot pattern the future of AI-powered products? In this talk, we’ll analyze and critique current patterns of human-model interactions and consider a more human-centric future for AI.

Outline

  1. Introduction
  2. Background 2. Survey the room 3. Brief history
  3. The Problem: Technology-centric design
    1. No direction
    2. No context
    3. No resolution
  4. The Solution: Human-centric design
    1. Target a resolution
    2. Build in context
    3. Design a direction
  5. Examples
    1. Khanmigo Writing Coach
    2. Khanmigo Coeditor
    3. GitHub Copilot
    4. Adobe Firefly
    5. Apple Intelligence
  6. Takeaways
    1. Chase the problem, not the solution
    2. Focus on the human user and their goals
    3. Embrace multiple model types and interaction modes

References

Appleton, Maggie. (2023). Language Model Sketchbook, or Why I Hate Chatbots. https://maggieappleton.com/lm-sketchbook

Brownlee, Marques. (2024). AI the Product vs AI the Feature. MKBHD. https://youtu.be/sDIi95CqTiM

Henley, Austin Z. (2023). Natural language is the lazy user interface. https://austinhenley.com/blog/naturallanguageui.html

Norman, Don. (2013). The Design of Everyday Things: Revised & Expanded Edition. Basic Books.

Ryseff, J., De Bruhl, B., Newberry, S.J. (2024). The Root Causes of Failure for Artificial Intelligence Projects and How They Can Succeed. Rand. https://www.rand.org/content/dam/rand/pubs/research_reports/RRA2600/RRA2680-1/RAND_RRA2680-1.pdf

Wattenberger, Amelia. (n.d.). Why Chatbots Are Not the Future. https://wattenberger.com/thoughts/boo-chatbots