Ask HN: How do you work through large ideas?

145 points by samcgraw ↗ HN
I seem to constantly have 1-2 large ideas in my mind that I chew on in the background, which seem to arise (or give way to something bigger) from reading or a lengthy discussion with a coworker or friend. For example: "what is the best way to organize a liberal democracy of >100M citizens?"

I'm wondering if there is an approach people have that best helps to organize, prune, and work through large, fundamental ideas such as these that they'd be willing to share.

86 comments

[ 9.6 ms ] story [ 155 ms ] thread
Being able to work through big ideas is really just a process of breaking a problem down into component parts and seeing how those parts interact and connect (much like with software). I'd recommend the book "How to Take Smart Notes" by Sönke Ahrens for some strategies on how to do that.
This! I see this as a research question, because to even begin to work through the problem requires proper framing—noticing the history of the problem, seeing what the best of modern thinkers are thinking, discovering what has not been applied yet probably from different fields, etc. This is the hard work it takes to have an opinion about something.

Whether you use metaphors or first principles, to even capture the scope of the problem and the important pieces that influence it is the first task.

Break large ideas into chunk size pieces. Focus of solving/understanding that piece and gradually build up momentum to tackle the bigger pieces. I love the book "Atomic Habits" and it tackles this issue in a much bigger context.
I give each of these big ideas a folder on my computer and then fill that folder with markdown documents and think aloud via writing. The core idea is that I don't want to lose the momentum of my thoughts. Maybe that might be helpful for your use-case.
Markdown files. Even better software that supports the concept of Markdown notebooks like Vnote. Organize your thoughts in a heading hierarchy.
Whatever that solution to your example is, pls don't raise tax. It's already too high.
I use my own app[1] for when they're small, when it grows into a project, it often has to do with code, so then I have a repository with a todo txt file in it.

I also use the concepts app [2] on my tablet to draw out mindmaps / charts /etc.

I use taskpaper vim plugin so I can quickly collapse and expand items to not make the file too overwhelming. In the file I group topics / questions based on what makes sense for that project, and at the top of the file I often have a 'topic' where I can quickly dump thoughts, that then later have to be refactored into one of the real topics.

I hope to one day find a way (and the time) to connect my ideas app with the repo txt file approach, so I can sync both.

Conceptually I think the hard part of most problems is figuring out the real important questions you want to ask yourself. Once you think you know them, then the answer is either obvious, or requires you to do something (research, build prototype, etc).

[1] Idea Growr app - https://play.google.com/store/apps/details?id=com.juliushuij...

[2] Concepts app - https://concepts.app/en/

Take time off, go offline and have deep thoughts. Record everything.

In 90% of cases: you need to ignore these thoughts as they can be time-consuming. You are not going to solve anything worthwhile related to organizing a 100M democracy if you are not in that field. You will not just "pick it up" unless you are a freak of nature and you are not.

In brief terms how would you organise a liberal democracy of > 100M citizens?
Your example is a big question, and it presupposes that answering the question will yield an optimal solution. If you focus on identifying your starting assumptions then you will be more open to incorporating currently unrelated areas of knowledge or research.
My current framework for that (chronologically organized journal) has its deficiencies, so I've been looking at Zettelkasten for nurturing big, long-term ideas. It looks promising so far -- it's certainly better than a journal for revisiting, updating, tweaking, and for "spending quality time with the idea".

However, I've been using it for just a month or so, and it's too early for me to form any conclusions.

Start with assumptions, and try modeling them. For example, my brother and I had a similar question so we wrote an extremely simple actor model in Python. Then we started adding more and more rules to the model until various metrics we were watching matched reality.

From there, it was a matter of adding/changing rules to match what desired and seeing the outcome.

So with a minor assurance that I'm not criticizing you for doing so, are you using "large ideas" to mean "things that are interesting, but have very little practical application"?

I ask because it changes my answer depending on whether the sample question is representative and how you are thinking about it, either as an author imagining an ideal situation or as a political theorist trying to understand human behavior of as a political reformer writing out a five year plan.

The issue being that politics is a team sport, and it doesn't really matter what you think is the most ideal system so much as what 100m people will agree to.

(Which isn't to say I have a problem with thinking about the ideal situations which have no chance of being implemented; I spent far too much time yesterday coming up with a proposal to redefine the foot-pound-second units as 1 second = 9.5x10^9 caesium-133 oscillations, c = 1 billion feet per second, and 1 pound = mass of 10^25 atoms of Si-28.)

As far as I know, there is only one known successful approach to working through large ideas, and that is to build solutions to smaller related problems and scaling up those solutions after some kind of objective information gathering.

Generally a well-understood stable system of any kind (including a democracy) evolves from a smaller stable system.

This principle is more or less built into the scientific process, where smaller results are peer reviewed and kicked around to a while before being combined to solve large problems.

Most ideas can be deconstructed into smaller ideas. When I started my recruiting business, I wanted to become so good that I eliminated unemployment. That was a bit lofty, so I paid attention to things smaller in scope. One thing holding back my recruiting business was was index finger hurt too much. I could review a few hundred resumes in an hour, and processing each resume took about 10 clicks of the mouse. My finger started to hate me, so I designed a recruiting application that got the number of clicks down to 1 per resume.

So maybe try to start big, and zoom in to focus in on one piece of the puzzle. There's a good chance that whatever you choose can be used in other ways if the primary idea doesn't work out.

I take the question seriously but the answer that came to mind was- have kids with a partner. Caring for and supporting them- and a partner- will in the near term put to concrete use the spare cycles your brain is devoting to these abstractions, and in the long term they will ask of you these abstract questions and you will by then have arrived at answers to help them.

Ok. More specifically- some other poster suggested recording assumptions around these questions, and I will go further. Behind every complex large scale optimization question- what is the best way to X...- when X is fundamentally about humans- are a set of assumptions- here, around the definition of "best."

Rather than chew on the question as framed, perhaps work to understand the assumptions your own brain is baking into the question in whatever value is being optimized. Understand further those assumptions as choices in a landscape and identify other choices that did not originally arise in the framing.

With an understanding of the larger variety of choices that were originally hidden, visit in conversation with your fellow humans and ask of them how they see that landscape and set of choices.

This discovery process through engagement may yield more insight than the self-centered utilization of your brain's optimization machinery. This process makes yourself available for teaching moments that are usually more impactful both for teacher and learner than any specific optimization insight and may help in further writing/sharing those moments.

Best wishes.

I don’t have kids, but I do have a partner. Nonetheless, such large questions no longer consume my spare cycles.

Obviously there are many exceptions, I wonder if on average interest in such questions peaks. If so, does the average peak coincide with an age or a period of life. Do we lose interest in our late 20s/early 30s, on average, because that’s when we have kids or is that just how long it takes for people to get bored with contemplative pondering?

I sort of hope it’s that having kids, a partner, a more challenging career soaks up the extra cycles. But even without those things, I find my passion ebbing. I worry this is an inevitable part of aging.

You must break large problems into smaller pieces and solve those individually. It's only when you master the details of the small steps that you begin to solve large problems.
1) write a plan

2) execute the plan

The plan is likely hasty and will prove to be a failure when you execute on it. That's fine. The plan is little more than worthless, but the process of planning is invaluable. Once failure or any collision become clear repeat the process with a modified plan. Don't wait for clarity to strike.

If its a big idea, break it down into smaller pieces and try to prototype smaller parts - eg before organizing 100m, see if your idea works on 20 people. Also, try to boil the big idea down to an actual example where it does/does-not work or apply. Also - review the history/literature of your idea. This is key. You will always find historical analogues for just about any idea you can conceptualize. Someone will have had the same or similar idea before, and it will likely be in the historical record somewhere. Read the history, to go further, faster. And if the big idea is worthwhile, share it with others and get their views - but try to use your ears to listen. Don't listen with your mouth (as so many do). Recognize also that big ideas and results that come from them, often start - in science especially - from careful observations of small anomalies. Try to pay attention to see things that others miss - often these micro-observations are both the source of and can hold the solutions to something that turns out to be really significant. Also, if the big idea is also a problem, you might benefit from using the tricks and techniques of creative thinking algorithms of De Bono - invert the idea, transpose with random phrases, etc. And final comment - beware of "big ideas" that are just sophistry and obfuscation. You deal with those by using courtroom techniques to pull out the truth, and show that your counterparty is either just "talking his position", or perhaps outright lying.
that was a superb set of brilliant advices... ty for your time to answer it
Scale variance of complex systems: a solution that works for 20 people rarely works for 200k or 20M people.
As a consultant, breakdown the problem into to its constituent pieces ensuring that each part mutually exclusive, completely exhaustive (MECE) this approach helps to make sure the entire solution set is covered, nothing should be missed out or overlap. Of course this can be difficult when the solution set is extremely open ended or broad,but it's good for organising ideas
Hiring McKinsey to run the country sounds like the logical and chilling conclusion to this line of thought.
What's the goal? If you want to just stop thinking about it, write it down somewhere.
I post big ideas on Twitter and see if anyone bites.
> What is the best way to organize a liberal democracy of >100M citizens?

Start by defining "best".

i'd define identity and scope of my terms; here citizen, person and human, each with their aggregates and conglomerations in order to expand the question as far as seems reasonable to domain experts -- before looking for generics and other abstractions.

this would get me to a point where i could inquire into the meaning of organize .. in this case probably realizing that there's sufficient self-organization that i'd have to think about needs/support and self-emerging structures.

at that point we could look at how the situation differs from the understanding implied in the question.

its only after i get my and the questioner's terms and models straight that i could begin to define some kind of scale to deal with someone's idea of best in order to work on generating 'deliverables'.

at least for me its best to become an expert in the things i want to change before suggesting changes in or to them; if i don't do that, or start from or with any kind of enthusiasm engendered by the power vested in me by big questions, good outcomes tend to move toward the vanishing point instead of forming solid new baselines.

but thats what you're suggesting the questioner do by having them question the use of the subjective term 'best'. right? :)

When confronted with this question:

> What is the best way to organize a liberal democracy of >100M citizens?

Ask yourself this: Can you reformulate the question into questions that make more sense?

Who can/should organize a liberal democracy of >100M citizens? What are relevant criteria to assert that a liberal democracy has been organized in "the best way"? Is it possible to agree on a fixed set of criteria? Is it possible to draw a model that matches reality? What game rules are you going to feed your model? Can you discern first principles that you're going to agree on and keep to build on top of?

The more questions and follow up questions you ask, the more concrete/tangible your answers will become.

Also, ask questions about yourself too. For instance, which biases can you identify with yourself? Are you able to avoid them?

In this case: how do you want to answer "best way"? If you are seeking an answer in absolute terms, then you may risk ending up in a logical tarpit discussing the existence of objective truth. Here you have a red flag: Is this question even a complete or answerable question?

So, it helps to question the question itself too. Maybe you need to bend the question before it can yield a worthwhile answer?

I described my approach to large new projects here: https://news.ycombinator.com/item?id=24896237

For more free thinking the best thing is to keep it in the back of your mind and stay close to a notebook at all times. With several similar ideas they will cross-pollinate and put you in a mindset to explore them further.