I am brainstorming presentation ideas.

Here is a list of titles: 1. Scientific Method 2.0 : applied scientific modeling in the age of AI 2. data about data : 3. finding your niche in the machine : systems engineering principles 4. Reclaim your data : Personal OpSec & Data management advice 5. How to speak computer : Machine Learning, systems engineering, code languages, and magic

Here is a list of slide ideas: * underfitting, overfitting issues are pervasive * tech secret vendor lock-in * what is data, what is metadata? * hash functions + content addressable file systems * the semantic web * UML diagramming * user interface design * the linux philosophy * DRY * personal blogging & “research journaling”

Here is a diagram where I tried getting at the different states of product development with the intention of talking about vendor lock-in and the incentives put onto code “owners”.

---
title: State Diagram for a Digital Products Creators
---
graph TD

start((startup))
%% MVP addresses >=1 use case
writeCode("working on the magic")
sellCode("sold the magic")

leaseCode("sell use of \n the magic")

demoCode("give away the magic")

%% ======
start --> writeCode

writeCode --> sellCode

writeCode --> leaseCode --> writeCode

writeCode --> demoCode --> writeCode

I am stuck at how to represent different levels of solution obfuscation and data harvesting in this diagram and I think I have lost sight of the purpose of this diagram. I need to take a step back and reconnect with what I want to communicate.

Here is a bulleted list that inexplicably feels relevant: * data * metadata ontologies rely on a common semantic understanding * the problem of naming things * the “semantic web” makes the source of our shared understanding explicit * LLMs include mathematically encoded semantics

But what do I really want to communicate to anybody doing research? * The importance of establishing trusted sources of information. * Remembering your inability to perceive an objective reality

Here is a screenplay:

EXT.  UNIVERSE - DARK

A pure black screen.

NARRATOR
This is a representation of the universe.
In this representation the information needed to explain everything is stored entirely in time.
We can sit here and let I can talk for an infinite amount of time until I have explained everything to you but it is not much to look at.
Let's convert some of the information into space.

A single white speck pops into existence in the center of the screen.

NARRATOR
A single point in space existing only for the tiniest infinitesimal of time.
That point has no meaningful spatial dimensions, but we have made it a little bigger so you can see it.
This is a unified grand theory looks like in spacetime.
One single piece of raw data that we can plug into a time equation that would define everything yet to happen to this point.
The realities of space and time are intertwined in ways that render the universe computationally irreducible.
This is not a model of all space and time as viewed from an omnipotent persepective.
This is a model of everything as it is viewed from a conscious perspective - filling not both space *and* time - but filling space *or* time.

The speck grows into a circular white line.
As the circle grows it begins fading from bright white into hues into red and blue representing hotter and colder areas in the cosmic microwave background.
The circle stops growing.

NARRATOR
This circle has spatial dimension because we have used up some time to create it.
You have observed the passage of time, the expansion of space, and now your reference frame is no longer that of a single point.
The representational model we created is stored in the color of the points along the circle.
See the whole circle all at once by focusing on the center but percieving the outer limits.
Focus your perspective on a single point along the edge to inspect it more closely.
Hold in your mind this single spatial frame of reference.
Be the circle.

The circle disappears.

NARRATOR
...and whatever else time has in store.

Revisiting this 2024-01-07 and I wrote the following:

Trust and Citations in Academic Publishing: Workshop Guide

command line basics

An introduction to command-line interfaces and basic POSIX commands. The workshop will cover the fundamentals of navigating the command line, executing commands, manipulating files and directories, and using command-line tools for various tasks.

intro to scientific computer programming for research

An introduction to tools available for researchers.

Reproducibility in Data Analysis

How to maximize the reproducibility of your results. A guide to implementing best practices in data analysis, including documentation, version control, and sharing code and data.

getting a computer to do something is a translation problem
.

Data Visualization

User Experience Design for Research

commanding computers

As AI becomes increasingly accessible An introduction to command-line interfaces and basic POSIX commands. The workshop will cover the fundamentals of navigating the command line, executing commands, manipulating files and directories, and using command-line tools for various tasks.

Reproducibility in Data Analysis

How to maximize the reproducibility of your results. A guide to implementing best practices in data analysis, including documentation, version control, and sharing code and data.

how to cite a superintelligence : why we cite

science-> know (Latin) skepticism->search (Greek) cite ->call (Latin)

scientific thinking means having a huge amount of skepticism. trust absolutely only what your own data says. if you are to trust words, then you must establish sufficient shared experience. the most trustworthy scientific claims are ones you can and have tested yourself.

An ideal scientific publication makes one specific claim so well that any reader would be convinced, and so concisely that it would be an entertaining read.

Problem: audience. We must rely on shared experiences to establish trust. This works best if you mostly already agree with the author. Citations can be used as a way to establish trust by citing authors, publishers, or results that the reader already agrees with. Word choice and use of familiar jargon can also establish a sense of shared experience with the reader.

Problem: data. We must always be skeptical of data. Do we trust those who collected and published it? Can we collect it ourselves?

Heuristic: - citing a person. means you have evaluated the trustworthiness of a person. - citing an org says I trust everyone in this org and the mechanations of the org - citing an AI says I trust everyone who contributed to the training data and the modeling method being used

fact citation provenance

types of citations • define a term (see here for clarification) • substantiate a claim (see here for justification) • data source (see here for raw data) • code

As a reader I want to __ so that __ - lookup the definition of a term - investigate the legitimacy of a claim - reproduce some results presented -