Compass-maxxing infinite loop intuition build
In my Lv.99 lich post, I talked about intuition and rational builds. Intuition builds are creatives who have high levels of instinct about their art, who can immediately tell if something feels right or wrong. They rely on a finely tuned internal compass that strongly guides every artistic decision they make.
At the end of that post, I added an example that clearly illustrates the difference between intuition and rational builds, repeated here for clarity:
There is something which people learn, partially from school and partially from work experience, which causes them to write as if they were charged for every word which goes down on the page.
— Patrick McKenzie (@patio11) April 21, 2025
Words are free! They belong in a vast mindscape! You can claw more from the aether!
@dwarkesh_sp: Unreasonably effective writing advice: “What are you trying to say here? Okay, just write that.”
@patio11: Cosigned, and surprisingly effective with good writers in addition to ones who more obviously need the prompting. Writing an artifact attaches you to the structure of it while simultaneously subsuming you in the topic. The second is really good for good work; the first, less so.
One thing that I tried, with very limited success, to get people to do is to be less attached to words on a page. Writing an essay? Write two very different takes on it; different diction, different voice, maybe even different argument. Then pick the one which speaks to you. Edit that rather than trying to line edit the loser towards greatness.
There is something which people learn, partially from school and partially from work experience, which causes them to write as if they were charged for every word which goes down on the page. Words are free! They belong in a vast mindscape! You can claw more from the aether! I think people might operationalize better habits after LLMs train them that throwing away a paragraph is basically costless.
@me: It’s not costless, unless you’re a rational build. Rational builds need to force themselves to write more because they have no instincts, so the strategy outlined by patio11 here works for them. But not for intuition builds. The intuition build already knows if something is going to be good or not as it pops into their head, so doing the exercise of writing two paragraphs/essays and throwing the worst one away is pointless. But not only is it pointless, it’s damaging.
Putting down words that are wrong damages the compass of an intuition build. People know not to do this not because they learn from school or work experience, but because they have an instinct, which the rational build lacks, that writing down something that doesn’t feel right will damage them in some way. And that instinct is correct, because it will damage their compass, it will make them less able to, in the future, tell at a glance if something is good or not, and so they correctly decide to not do it.
The argument here is that writing things down is costly for intuition builds because it damages their compass. I’d refine it and say that writing things down with conscious intent to throw them away is what’s most damaging. Everyone needs to revise, but putting down words you know aren’t good enough is bad. That instinctive avoidance is right: when you train yourself to ignore your sense of what feels right, you dull it over time.
Intuition builds tend to get blocked if their compass isn’t aligned. If an idea doesn’t feel like it’s really good, they stop. They don’t like exploring weak signals, but that instinct can keep them from discovering new paths entirely, like the resistance to generating variations of some paragraph.
But chatbots solve this problem effectively. I’ve been using AI for four main things recently: coding, worldbuilding, writing revisions, and game design brainstorming. For example, you can see how it helped tighten the Soul Society and Haircutter posts. I included the full chatlogs in the AI usage post.
When I work with the chatbot, I always start with text I’ve already written that’s complete, but not finished. First, I use the AI to help me improve the structure so that the story’s core goal comes through more clearly. Then I go paragraph by paragraph, tightening language, removing repetition, and refining flow.
To improve a paragraph, I usually ask the chatbot to generate two or three variations, exactly as patio suggested. Then I look at the options, pick what resonates, mix in some ideas of my own, and rewrite that paragraph as a combination of both our inputs. After that, I feed the new version back to the AI, and we refine it until it feels right.
The reason this workflow works especially well for intuition builds is because it avoids compass damage. It sidesteps the problem, as even though I’m still generating throwaway drafts, I’m off-loading the damage to the AI. It creates the variations - which are often very good and creative, by the way - and I just pick the best parts.
But there’s another, stronger advantage: I’m also training my compass. Every time I have to pick what’s good, I’m practicing my internal judgement. I’m using my intuition as a filter, thus sharpening it in the process.
This same loop applies to game design, worldbuilding, and even coding. Yesterday, for example, I described a prototype idea in detail and asked the AI to summarize it back to me, and then to generate items, passives, and build archetypes based on its understanding. It replied with a really solid list. Some ideas were off, but many were great. I picked what worked, gave better context. It gave back another list, repeat. This loop kept going until we reached a cohesive set of ideas. And all the while, I was selecting, pruning, filtering, and thus my compass was levelling up.
This is the compassmaxxing infinite loop intuition build.
I believe if you follow this build, you’ll reach level 99 in no time!