Rylan Schaeffer

Kernel Papers

Rylan’s Approach to Research


Research is an emotionally tough but rewarding journey. It requires struggling, learning, exploring, communicating. The key logistical details are:

  1. Each research project is a coordinate ascent problem: iterate between (a) trying to find a better goal and (b) improving your approach to achieve your current goal
  2. Especially when we’re starting to work together, let’s overcommunicate to make sure we’re on the same page
  3. Send me personal feedback early, bluntly and often
  4. Let’s meet once every week or once every two weeks (unless you have a strong personal reason to prefer not to)
  5. Maintain at least two documents: (a) a daily personal research journal and (b) a slidedeck of results with key results and the minimal information necessary to understand them

Why Do Research?


Approaching Research

Step 1: Identify possible research goals

Research is fundamentally about advancing humanity’s collective knowledge. Every research project should begin by attempting to identify an goal motivated by at least one of the three following questions (or some subset thereof):

  1. What scientifically would we like to know?
  2. What mathematically would we like to prove or disprove?
  3. What product or tool would we like to build?

These three questions correspond to the three intellectual lineages that research belongs to:

  1. Science: improve humanity’s understanding of the world by hypothesis driven experimentation
  2. Mathematics:
  3. Engineering: build useful, robust, flexible, scalable solutions to important problems

Every project we work on together should begin by attempting to clearly identify the goal of the project. Oftentimes, the goal will be unclear at the outset and will evolve as we iteratively sharpen our understanding - this is great! But I personally believe that even the most exploratory projects should be goal-directed.

Step 2: Evaluate research goals

Once you have a set of possible research goals, it’s time to prioritize and choose one (or however many you have bandwidth for). There are many considerations when evaluating possible research goals. Here are at least 5 considerations:

  1. Importance: If you achieve your goal, how many people will care and how much will each of them care?
  2. Feasibility: What is the probability that the goal is achievable? Do you have a realistic plan of attack? Time travel is important but no one realistically has a good solution.
  3. Capability: do you have the requisite skills to have a high chance of succeeding in a reasonable timeframe? A problem might be solvable, but not with the skills you possess.
  4. Growth: will the path forward offer you the opportunity to grow in directions you wish to grow in?
  5. Scope: is the magnitude of the project commensurate with the time and energy you plan to put into the project?

Accurately evaluating possible projects is a skill that may take time.

Conducting Research

Documents to use

Research - regardless of whether scientific, mathematical or engineering - will involve experimentation. Consequently, we must be good experimentalists!

Being a good experimentalist is hard! Learning to be a good experimentalist will take time and practice Two key components that I expect you to use:

  1. A research journal: This is a diary for you to keep track of what you do / try. It is for you, so organize it however you wish
  2. A research presentation: This is a diary for you to communicate what you do with your research collaborators. You will add key plots as time goes along. This also minimizes time when you later want to formally present your research - just slice a subset of your slides and you have a draft!