My Plan for Contribution

Given the goals of automation of reasoning, exploration, and population expansion across the stars, I’ve set personal goals that guide me through my life. I don’t do this to achieve power, wealth, or fame, and I don’t need to be the one who receives credit or even makes the breakthrough achievement. My goal is for these things to happen.

TODO: expand each section into a complete article

The AI Assistant

The human mind is a wondeful source of ideas, and we are experts at imparting our will on matter. Doing this at scale requires delegation, and making people do your bidding is no better than slavery. That’s why I strive for an intelligent digital assistant, one who will keep track of my goals, and have agency to handle matters for me.

I have studied Computer Science, Mathematics, and Machine Learning for this very reason: to create a universal algorithm that will be able to mold my ideas into concrete plans, and perform as much as possible of the plan of its own accord. Ilya Sutskever’s GPT Agents are exactly that tool, and I am very grateful for access to this technology, and for the way forward that this paves.

My first contribution is the creation of a local instance of the intelligent assistant, one that cannot be taken away from me. I know that now, and for the rest of my life, I will be able to make use of a tireless assistant that will follow my instructions and interface with systems for me. My second contribution is the recursive task planning strategy, by which an LLM with limited agency can divide and perform a complex task by breaking it into smaller tasks.

The Self-Replicating Machine

In many ways my hands are not enough to manipulate matter to my will. Without tools I can’t handle the very large, the very small, the electrical, or the chamical. As tools, such as a forklift, are inevitable, it’s not a big stretch then to have these operated by the AI Assistant, instead of myself. Another benefit of automation is the tireless nature of machines; they will do the task for as long as there is a task to do.

In order to scale my agency, I need access to a large number of machines, and it seems to be the best way of achieving this without first having to work to gain wealth is by creating self-replicating robots. A robot that can do useful work, and also assemble a clone of itself from basic parts. The robots can then man the mines, the factories, and repair shops. They can handle the complete production process of the modern industrial complex.

My contribution is a robot arm simple and precise enough to replicate itself. I built it and ran it.

The Von Neumann Proble

This is effectively the machinery that performs the entire production line from raw materials to copies of itself. Such a machine is the most convenient way of colonizing exoplanets: send an artificial mind to any barren rock, give it robots to do the necessary work of converting resources into machines and power, and get them to terraform and replicate.

By giving it the tasks of replication and terraformation, we can expand the livable space for humans and all of nature, exponentially.

Getting Material to Space

Rockets are a prohibitively expensive way of getting to space, which is the reason why the ISS is both humanity’s most expensive endeavour, and a small cramped series of hubs. In order to make a proper space station that isn’t in the Earth’s magnetosphere, or a fleet capable of exploring and colonizing, we need an economically sustainable method of getting hundreds of tons into orbit. Escaping orbit with ion pulse engines is then easy.

Large quickly reusable rockets may provide the answer, and I am an avid follower of developments in this direction.

My contribution is the invention of a method to get material to orbit for even less, by combining the space cannon of Gerard Bull with a supersonic orbital interceptor.

The Pocket Nuclear Engine

Can the CANDU reactor of heavy water and unenriched uranium scale down to something so small it heats a house? I like to think yes, and love exploring the topic.

Toroidal Magnetic Protection

I believe a rotating space station surrounded by an MRI-like magnetic field can provide the protective benefits of the Earth’s magnetosphere, far away from the Earth

Closed Ecological System

A pre-terraformation space colony will need a source of nutrients, as well as an artificial managed environment to sustain life. I believe aquaphonics have the potential to provide the solution.

Longevity

These are hard things to achieve in the short span of a single lifespan, and I’d like more time. That’s what fuels my fascination for the machinery of life, and its maintenance in a useful state.

We age and die, but it is not inevitably the only way. Just as a hundred-year old car can be kept in pristine running condition, I believe that with sufficient understanding the processes of ageing can be kept in check with the already present processes of self-repair.

It seems the AI Assistant will be able to achieve more in this field than any human, because the brightest minds of modern medicine already struggle to further extend useful lifespans. My intuition tells me to focus on mitochondria, which support us without ever getting older.

Relative Musical Notation

Beyond the purely necessary, mankind finds pleasure in music. In order to bring the music in my head into reality, I wanted to compose. Learning musical theory, I hit a foundation I remain deeply dissatisfied with: ideal tonal progressions are made impossible through a well-meaning but ill-devised system of approximations. It’s impractical to represent certain syncopated rhythms.

So I set out on a journey to understand the fundamentals, and devise a new musical theory and notation. Where current musical notation deals with absolute tones, mine is built on relative frequencies, making it impracticable on the piano, while very natural for the composer and listener.

P vs NP

Computability is a fundamental aspect of reality, and many have been asking whether problems verifiable in polynomial time can be solved in polynomial time. All cryptography of modern society is based on the assumption that yes, P=NP. In my opinion, Godel’s Incompleteness Theorem can be used to construct a proof that the problem of P vs NP is undecidable, and I endeavour to formalize my intuitionistic proof of this.

Errata

While aiming for these goals, I’ve found myself pulled into a number of questionnable endeavours which may have short-sighted benefit to others and society, but distract me from that which I consider fundamentally important.

  • two PhDs in irrelevant topics
  • various computer science jobs
  • a patent
  • three companies in three countries
  • investments and predictive models for trading
  • property

Elementary Building Blocks of Life

While aiming for these goals, I’ve also found myself pulled into things that I value and enjoy.

  • wife
  • kids
  • travel
  • intuitionistic history
  • reading
  • writing
  • art
  • drugs
  • cooperation
  • poetic expression