Use writing to augment your thinking
I've found it very helpful to use writing as a tool for thought. Writing forces you to actually think instead of one-shot evaluating, augments your working memory, and lets you evaluate your thoughts as a whole.
When you're using an LLM, the best way to get intelligent output is some sort of chain-of-thought tree search. See Tree of Thoughts, AutoLoom/Jacquard, MiniHF, and the recent math olympiad MCTS results. Essentially, each of these techniques has the model generate a number of possible continuations, and then evaluate which of those continuations is most promising. MCTS is a more sophisticated algorithm for doing essentially the same thing: iterative search over the space of possible continuations. This sort of prompting augments the model's thinking capacity by both: a) letting it work step-by-step, rather than in a single forward pass, and b) letting it leverage its capacity for noticing good thinking to steer itself towards good paths in thoughtspace. It's usually much easier to notice when something is correct or well-thought-through than to come up with it yourself, and the same applies for your own writing. When you write, you'll notice which ways of paths are best and steer towards them in a way that's much more difficult to implement using only your inner monologue.
It generally seems like many of the most effective people throughout history spent a lot of time reading and writing. Newton likely wrote millions of words in his lifetime, Franklin was a prolific writer, and Darwin famously kept an in-depth scientific journal. This definitely isn't true of everybody who did great things, but it's likely true of many of them.
Here's a haphazard list of places writing might be helpful:
- John Wentworth taught his SERI MATS section a series of exercises and workshops intended to accelerate research. Those include:
- Giant Text File Technique: Just write down everything you know about the field.
- Game Tree of Alignment: Write a list of problems you might encounter, ways you might solve them, and ways those solutions might go wrong. Pushes you to emulate a minimax algorithm in a game of strategy.
- Writing: Practice communicating your ideas to others!
- Existing evidence: Decide on a claim, then look for real-world evidence which might help support or undermine it. List possible sources of existing real-world evidence, then figure out what that has to say about your model. JD Pressman has written about similar processes (and justified them well) in Necessity and Warrant and Literature Review for Academic Outsiders.
- Hamming questions: Write about the most important problems in your field, and how to solve them.
- Framing: Write three examples of a frame (e.g. stable equilibria). Ask: what does the frame imply is interesting about each example?
- Oh, and to complete the summary, even though these are less related to writing, you should:
- Quickly design and execute on an experiment if possible
- Maintain a prototypical example of the idea under discussion
- Attempt to convert intuitive arguments into mathematical conjectures
- Just write out your plans for an upcoming project. Specifically: what are you trying to accomplish, why do you want to accomplish it, how will you accomplish it, and why will that go well?
- Write about what you're learning. You probably shouldn't expect yourself to absorb and integrate knowledge from a one-shot exposure. Writing about what you've learned, your thoughts on it, or how it might be applied helps you refine and integrate what you've learned. You could substitute this for many other things, but writing is a very helpful technique. See TurnTrout's public writing about his learning.
Also, try to as much as possible write to a mind you trust. Your thinking will be better if you know your intended audience will follow you when you light the way.