Skip to content

Latest commit

 

History

History
48 lines (28 loc) · 3.26 KB

wireheadingMATRIX.md

File metadata and controls

48 lines (28 loc) · 3.26 KB

Social media platforms rely on virality emerging through user engagement. However, differences between human and AI information processing capacities may modulate virality, therfore vreality. Heres a framework for modeling how content attributes and platform dynamics interact with human versus AI capacities to impact viral potential. Virality represents a lottery reward function in humans. I want to explore a positive wireheading outcome, considering this is where we are currently at.

Definitions:

  • H: Vector representing human information processing attributes.

    • h1: Attention span for a unit of info (e.g., time in seconds).
    • h2: Context processing ability per unit (e.g., number of contextual connections made per unit of info).

  • A: Vector representing AI information processing attributes.

    • a1: Attention window for a unit of info (e.g., time in microseconds).
    • a2: Context density processed per unit (e.g., number of contextual connections analyzed per unit of info).

  • C: Vector representing content attributes.

    • c1: Length (amount of info, e.g., time duration or word count).
    • c2: Context density (e.g., number of contextual references per unit of info).

  • P: Vector representing platform attributes (e.g., ease of sharing, algorithmic promotion, user base size).

  • V: Scalar representing the virality potential of a piece of content.

The relationship between these variables is given by:

[ V = f(H, A, C, P) ]

Assumptions/Conditions:

  1. V increases when h1 > a1 and h2 > a2 — This implies that the content has a higher virality potential when humans have an advantage in attention span and context processing.
  2. V increases when a1 > h1 and a2 > h2 — This implies that the content has a higher virality potential when AI has an advantage in attention window and context density processing.
  3. The comparative advantage between H and A may switch based on the attributes of C.

  • Short, Low Context Density Content: If c1 is small and c2 is small, then V will be higher when the attributes of H are greater than A (indicating a human advantage).

  • Long, High Context Density Content: If c1 is large and c2 is large, then V will be higher when the attributes of A are greater than H (indicating an AI advantage).

Meaningful Human-AI Collaboration

For meaningful human-AI collaboration, consent and agency are critical prerequisites. Thoughts on how individuals can protect those in the CONTEXT of potentially addictive technologies:

Align yrself

Seek balance and moderation: Regularly reflect on how technology use aligns with personal values. Advocate for better design practices: Support designs that respect autonomy, like introducing friction, nudges, and opt-in models as opposed to the default endless scroll.

Individual agency + collective action holds power. By taking responsibility for our faculties and advocating for aligned design principles, we can harness technology in a way that promotes flourishing. These small efforts can compound, much like viral content, but in a direction that values human dignity over addiction. That way we can still haz redbull as our Will needs it, sugar is a tool on the path.