---
date: '2024-02-25'
description: and the state of being.
id: representations
modified: 2026-10-12 09:16:29 GMT-04:00
tags:
  - seed
  - philosophy
title: representations.
created: '2024-02-25'
published: '2024-02-25'
pageLayout: default
slug: thoughts/representations
permalink: https://aarnphm.xyz/thoughts/representations.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
resources: Edward Tufte’s [The Visual Display of Quantitative Information - ISBN-13: 978-1930824133](https://www.edwardtufte.com/book/the-visual-display-of-quantitative-information/)

> \[!question\] Question
>
> How do we represent information flow when we interact with different media, especially text?

Maps shape _<mark>how we see and understand the world</mark>_, which in turn shapes how we act within it.

## maps

[Linus’ talk on Representation](https://www.youtube.com/watch?v=Tt_hDVN5TZE\&list=PL_Kh09CR3iHa1ZPi4EsHeqX-1IB3CRLbA\&index=2\&ab_channel=CharStiles)

how maps shape our _agency_

## as technology

Viégas and Wattenberg propose a dashboard beside an AI conversation, exposing selected parts of the system’s model of itself and of the person using it. Their proposal assumes those internal models can be identified and interpreted. They describe the design as speculative and report no user studies; which features would help someone use the system remains open. \[@viégas2023modelusermodelexploring\]

How can we manipulate new ideas and abstractions to create new <mark>notations</mark>?

Thesis: Technologies for _representation_ mediate how we see the world.

Tenet:

1. A representation must abstract
2. An interface always gives or takes agency. _Be intentional_ about which you choose
3. Represent to communicate, not model.

## internals grokking

What could [[thoughts/mechanistic interpretability|understanding an AI system’s internal representations]] tell us about how we think? Establishing a connection would require evidence about human cognition too.

<mark>Seeing clearly</mark> to exercise _[[thoughts/Agency|agency]]_

## interface

Instrumental use:

- frictionless input, low-latency progress: payments.

Engaged use:

- See the information relevant to a decision. Express intent naturally and precisely.
- Explore possibilities by interacting with the information.

> \[!question\] Why don't we do this more?
>
> It is a hyperreal produced from a radiating synthesis of combinatory models in a hyperspace without atmosphere.

_Simulacra and Simulation, Jean Baudrillard_

> Art of Cartography attained such Perfection that the map of a single Province occupied the entirely of a City \[…\] <mark>saw that that vast Map was Useless...</mark>

_On Exactitude in Science, Jorge Luis Borges_