I read Drive by Daniel Pink a decade ago, in high
school. Since then, my framework for getting shit done has been
intrinsic vs. extrinsic motivation.
This lined up decently well for a long time. But there were cracks in the idea - I had enough “intrinsic motivation” to keep going, I guess? But overall I think the whole framework is unsustainable.
This framing tries to understand your set of actions as:
- the “good set”: intrinsic
- the “bad set”: extrinsic
The idea that I took away from it was, you need to find actions better aligned with this thing called “intrinsic motivation” and less in line with this thing called “extrinsic motivation”.
I’m cutting orthogonally to that. I want to explore this two part systems of Goals and Garbage:
- Somehow, you must have a strong goal that you keep your mind towards.
- There must be “garbage work” in said goals.
Goals
“Goal” can mean anything. It can mean this thing we call “passion”. It could mean an external paycheck. It can mean something as generic as survival (and we interpret this in a Darwinian sense). “Goals” are not this objective, defined thing.
But I think it’s deeper than that. I guess it’s this sense of permanence, that comes from whatever source it comes from, that is the real thing you’re anchoring to. “Goal” is a contract, rational, or irrational, that you feel like you never have to question.
I say it like this, because I’m thinking about capacities where I can offload the ambiguous mental stress of thinking about ambiguous goals, and a canonical example is work.
I would say it’s not merely because my tasks are “well defined” (relative to some of the huge scope that the company has), but also stuff like, “somebody else has figured this out”, “even if the company fails, I can find another job”, stuff like that. In this sense, good goals put up good boundaries.
I was trying to reconcile this, because this makes goals seem a bit negative. What about goals that are “intrinsic”? I think that what these goals let you get away with is a justification (to yourself, at the very least) to ignore certain constraints. It’s almost like a self-limiting tool, in order to spike in one area. At least for me, that’s the motto I’ve usually held in my head - I’m so passionate that I am willing to ignore XYZ.
Garbage
So, if you have a goal, one with this “permanence” aspect, and obviously some force (positive or negative) wanting it to succeed, well, you have your work cut out for you. Not really - the point of “garbage” is navigating the ambiguity, sucking it up, and pulling through. But navigating ambiguity is usually done for something.
Garbage… with no goal
This is the killer that we want to address. For personal learnings with “no stakes”, I’ve never been able to get it to succeed, for this reason.
Well, sure - I’d pull up a blog, read some source code, and genuinely, my sporadic, one-off learnings has helped on the job before. I’ve read up on Python internals back in sophomore year of college, and on two jobs, funnily enough, did I have to deal with objects at the PyObject level. But I’m not a Python core maintainer, or a maintainer of any library.
I could set up a study group, or force myself to release a concrete artifact, force myself to stick to one project for 6 months, sit down and read a textbook front to back, or something else. That’s what the classic sayings are. It’s always, anchor yourself to one thing.
But on the job, there have been plenty of times where, on the same project, we’ve pivoted completely different directions. And a lot of my high-leverage knowledge has been the accidental knowledge that I learned on the quest for a very concrete goal (no matter how artificial/BS it was!). I learned quite about the Linux kernel from a project at my job that, while needed to be done, kind of was BS long term. I’ve learned a lot about math while doing random things here and there, and fun math - I thought all statistics and linear algebra was really boring, but I never thought that it’d actuall be useful in signal processing of all places.
Arguably, I’d say I had far more clarity when I had a specific goal I wanted to extract out of a field, and explicitly said, “This doesn’t sound useful”. A naive interpretation of “intrinsic motivation” would say that this is not “correct” behavior, and yet this kind of behavior has lead to some of my highest leverage bits of knowledge.
However, translating that kind of high-level drive to a personal project, inflates into ridiculous things like, starting a side business, or becoming an OSS maintainer. Then you get the same “ambiguous pressures” as a job, that incentives you to do whatever it takes.
(Of course, I also believe a “passion side project” could work here, but I haven’t found anything that scratched my itch, in a deep way to “keep this going”, as opposed to “this is a cool technology that I wanna explore for a few days.”)
Harnesses & Manufacturing Passion
This is more of a coping mechanism than a real, positive philosophy. There are plenty of things in life that we don’t want to do, but we must do. And not in a “Ultimately in service to a greater goal” kind of way - you could interpret this statement as “Good goal, have to do garbage”; no, I mean that the goal itself, you don’t believe in either. A trivial example we all can relate to is the education system. Or a boring job.
In this situation, one thing I would always do is, find something interesting to do, and laser-focus on that, within the system. In this sense, this is a manufacturing of a goal.
Another bias to watch out for is the bias of “one”. An easy example is existing harnesses - video games, structures that incentivize you to continue engaging with it, etc. These are not structures that (you usually) consciously create. Either someone else has done it, or it’s something that people have accepted and settle into, etc. It’s easy to call out these kinds of structures (even if it’s not easy to escape them!)
It’s harder to get rid of this bias when creating your own harnesses. A thing I tell myself is, “I’m gonna sit down, and understand every part of this textbook!” That’s not how learning works. The issue is, you see this “one” product, and you extrapolate some essential structure from it. Yes, it is a stable capacity; yes, it’s gone through all the heuristics you can name of. But no, it doesn’t mean that it’s a harness to actually help bootstrap you.
Note: This actually is how it worked for a philosophy of science textbook, but only after I’ve spent a few months digging through adjacent literature and thinking through the related problems myself. In that case, I consumed the entire textbook in less than 8 hours during a lazy day at work, and it helped synthesize and reformulate a lot of my thoughts - but only because I’ve already been primed for the subject. Theory and Reality by Peter Godfrey-Smith - check it out.
It’s not enough to just say “Everything is patchwork” though. I’ve embraced said philosophy, and while it’s been helpful for exploring new things, when done wrong, it obviously leads to unstructured meandering.
And so we’re back at the, “Don’t overthink, just do one little thing” approach. Which honestly, is pretty good. This blog wouldn’t be written without such a heuristic.
It’s this careful balance of not optimizing too hard immediately for this abstract “essential structure”, yet still manufacturing it properly, with the meta-knowledge that you are manufacturing it.
A potential plan
I will need to think more about this; but the core observations I think I’ve laid out.