hslu-ki-h2501

I.BA_PHKI

Schedule

Session 01

Session 01 image

Part 1: AI Learning — Algorithms and Experiences (slides)

  1. Brief overview of AI and its learning mechanisms.
  2. Brief introduction to machine learning algorithms.
  3. How AI learns through data, patterns, and repetition.
  4. Differences between AI learning and human experiential learning.

“AI learns from data — but does it actually experience the world?”

Part 2: Human Learning — A Philosophical View

  1. Conceptual Introduction – Unnatural Animals

    • Humans lack strong instincts and must learn through experience and error.
    • This weakness becomes the foundation of creativity and freedom.
    • Guiding question: If humans learn through failure, how does that differ from AI’s efficiency?

  2. The Failed Animal – From Instinct to Learning

    • Humans are “denatured animals” (Charles Pépin): instinct too weak to guide action.
    • A foal walks in minutes; a baby fails about 2,000 times before succeeding.
    • Activity link: “The Foal vs. the Baby” — compare instinctive success and human trial-and-error.

  3. The Methodology of Failure – Learning by Inefficiency

    • Humans progress through trial, error, and adaptation.
    • AI minimizes error; humans learn meaning from it.
    • Activity link: “Virtues in Failure” — identify what humans learn only through mistakes.

  4. Compensation and Culture – Learning from Others

    • Weak instinct drives reliance on culture, education, and empathy.
    • AI processes data but lacks relational and emotional context.
    • Activity link: “The Compensation Audit” — compare informational vs. social learning.

  5. Freedom and Invention – The Outcome of Failure

    • Human deficiency enables freedom: existence precedes essence.
    • Unlike AI, humans can redefine themselves and their goals.
    • Activity link: “Freedom Inventory” — reflect on freedom as a product of imperfection.

  6. Synthesis – AI vs. Human Learning

    • Humans learn because they fail; AI learns to avoid failure.
    • Weakness creates adaptability and freedom.

“If humans learn through embodied experience and error, how does perception shape that learning — and how does AI compare?”

Part 3: Perception — From Data to Experience

  1. Conceptual Introduction – Perceiving Machines?

    • Can AI truly experience what it perceives?
    • Introduces George Berkeley (“to be is to be perceived”) and Maurice Merleau-Ponty (the perceiving body and world are inseparable).
    • Guiding question: Does AI perceive the world, or merely compute it?

  2. Application Topics – Connecting Philosophy and AI

    • AI vs. Human Perception: data processing vs. lived experience.
    • Virtual Reality and Simulation: Berkeley’s idealism and digital worlds.
    • Embodied AI and Robotics: Merleau-Ponty and perception through movement and sensors.
    • AI–Human Interaction: perception as feedback and shared experience.

  3. Philosophical Lenses – Berkeley and Merleau-Ponty

    • Berkeley: perception defines existence; relates to virtual environments and AI “worlds.”

      • Key idea: Objects exist only insofar as they are perceived (“esse est percipi”).
      • Connection to AI: Virtual reality and data-driven “worlds” as digital idealism.

    • Merleau-Ponty: perception is embodied; links to robotics and interactive AI systems.

      • Key idea: The perceiver and the perceived world are intertwined.
      • Connection to AI: AI without a body may “see” but not feel the world — can it ever achieve phenomenological perception?

  4. Discussion and Reflection – Thinking Like a Philosopher

    • Key questions on AI experience, embodiment, and the nature of perception.

  5. Synthesis – AI, Perception, and Reality

    • Comparison of human vs. AI perception (embodied vs. computational).
    • Takeaway: AI can simulate perception but does not live it.

  6. Transition to Practice

    • Reflect or debate.
    • Design an AI inspired by Berkeley or Merleau-Ponty.
    • Prepare for Part 4: practical application.

“Is AI perception an imitation or a fundamentally new kind of perception?”

Part 4: Practice

Session 02

Session 01 image

Part 1: In the news

  1. Iran wedding scandal

  2. GPTs in shopping

    • Official post
    • “LLMs are slot‑machines” (16 Aug 2025 on Pluralistic) (pluralistic.net)
    • “Elon Musk to introduce ads to X’s AI chatbot” (7 Aug 2025 on FT) (FT)

  3. AI Kill internet doc

    • Satire distinction
    • Eliza bot
    • Stocastic parrot

Part 2: Trial and Error - A Common Ground ?

  1. The “shared” theme of learning through trial and error in both humans and AI.
  2. Implications of error-making as a learning process.

Part 3: Learning to learn

  1. Attention mechanisms
  2. Memory
  3. Inhinbition and flexibility
  4. Metacognition
  5. Motivation and spirit for changes
  6. Emotions!

Part 4: Practice

  1. The declaration of independance