Caring about excellence

“Anything worth doing, is worth doing right.”

― Hunter S Thompson

“Be not professional in what you do, rather be excellent. Excellence has life in it — it has colors in it — it has sweetness in it — whereas professionalism is a dead corpse exuding the disgusting smell of obedience. Excellence requires no obedience, yet in excellence you act your best, without all the life-sucking efforts.”

― Abhijit Naskar

“Excellence is the result of caring more than others think is wise, risking more than others think is safe, dreaming more than others think is practical, and expecting more than others think is possible.”

― Ronnie Oldham

“I must address the topic of whether the effort required for excellence is worth it. I believe it is — the chief gain is in the effort to change yourself, in the struggle with yourself, and it is less in the winning than you might expect. Yes, it is nice to end up where you wanted to be, but the person you are when you get there is far more important.”

― Richard Hamming

“You’ve baked a really lovely cake, but then you’ve used dog shit for frosting.”

― Steve Jobs

A good part of my felt motivation^[1] for caring about EA, and for working to reduce existential risk, is a desire for good things to be excellent. I think there’s something healthy about that, and that we should talk about it more.

The feeling of caring about excellence

The quotes above all speak to my heart. They each capture something important, and they convey its essence, powerfully. In a way, I feel a kind of love for them — almost fierce in its admiration if I try to tune in to the flavour — as I do for all things I feel to be great. (This is strongest for the Naskar and the Hamming quotes, and weakest for the Oldham one, but it is there in some degree for all of them.)

When something doesn’t work very well, or isn’t a good fit for its position, or is just kind of mediocre, sometimes it’s frustrating; often it’s just … fine. There’s a lot of things in the world. We can’t realistically hope for them all to be excellent.

But when something fits well, and feels like it’s doing a good job — whether it’s the clean curve of a teapot, or a clarifying sentence in a research report — there is some satisfaction in that. And when something important is doing its work excellently, there is a deep sense of rightness.

This hooks into my motivations. When I come across something that’s great in some dimensions (including seeming potentially important), but kind of mediocre in others, I feel a pain — I desire it to be better. I want to know if I can do something to help it work towards its potential.

Similarly when I have an idea that I’m wondering about writing up, often I’m implicitly asking myself “could this be excellent?”. When I am working on something, I try to let my sense of excellence guide me. Of course I make mistakes all the time, and in any case I can’t always access that feeling — but I do often have a sense of “is this reaching excellence?”, and when the answer is “no” the right response is usually to interrogate my feeling, or to go away and talk to other people about it, or to come back another day.

Caring about excellence and caring about growth

Another part of my felt motivation is more directly growth-flavoured. It feels good to help things be a little better, a little more themselves. Some of this feeling is grounded in caring about excellence — and I feel most trust in the parts which are. We can only make things incrementally better; pursuing growth in service of excellence feels healthy. On the other hand, I worry that pursuing growth when it’s detached from this can lead to not pushing things through to really excel, or to spending too long polishing turds.

Excellence and x-risk

I like effective altruism because it’s centrally about trying to make something I care about for other reasons (helping people) better^[2]. Decisions which are informed by scope-sensitive thinking seem better than those such are not.

Similarly, my work in x-risk is not motivated at a felt level by thinking directly about untold numbers of beings in the future light cone^[3], but by a sense that avoiding x-risk should be an important priority for a sensible world. I feel like this is a critical ball which is being dropped, and I would like to help it not to be dropped.

This extends to how I orient to work within x-risk. I generally don’t have grand theories about exactly how my work helps. (I’ll think through theories of change, but hold them lightly.) Rather, I try to find places where I can meaningfully nudge things towards being more sensible. I think the communities thinking about x-risk, and the thinking on how to make the future go well, are something precious, but with many flaws. They are worthy of our efforts in holding them to higher standards; making them more excellent. I want to work to help make them wiser, more virtuous, more informed. I care most about improving the parts that I think are already very good. I am trusting that this excellence will help — and more robustly so than any plans I can make today based on my necessarily flawed understandings of the issues. So I keep an eye out for ways that I can contribute to help make them the best versions of themselves. Sometimes this means helping other people to orient in ways they’d endorse; sometimes contributing to the discourse to provide clarity on a point that seems important but underappreciated; sometimes directly advising on plans or commenting on other people’s writings to try to make them better.

On not making the perfect the enemy of the good

Pursuing excellence can blur into perfectionism. And there’s something important to not making the perfect the enemy of the good. Sometimes something is worth doing, but not important enough to warrant a pursuit of excellence.

However, my experience in practice is that there’s less tension here than you might guess. I see perfectionist energy as caring about eliminating all flaws, while excellence cares more about how high it can elevate something’s best facets. A great startup is excellent, but never perfect — and chasing perfection would stifle its excellence.

Or here’s how I might think about this if I’m writing something I intend to post publicly. If it’s a straightforward point and I feel mastery over it, I can just write it and post it. That will be enough.

Often I feel like I’m trying to untangle some knot, and add insight. Here I care about excellence. I may often start writing when I’m still in some part confused. But at some level I’ll be aware of this. So I write not to publish, but to start a reflective process. I’ll write and then seek conversation with others. I’ll seek to understand where I’m confused, and where my writing is confusing. I’ll ask for help in making it better^[4]. (Collaboration can be very helpful in the pursuit of excellence.)

As I iterate through drafts (and sometimes there are multiple revisions of throwing out an old draft and starting again), I can often feel things getting closer to “right”, even though I don’t know exactly how to describe that “right”. And when things are sufficiently in alignment, and comments aren’t making me want big revisions, it will be time to publish. Are there still imperfections? Of course. Does it serve the pursuit of excellence to root them all out? It does not.

Alternatively perhaps I’ll reach a sense of being stuck and not knowing how to improve it, while still sensing that it’s flawed. Then I’ll make a decision between leaving it to return to later (with hope that time will bring clarity), or posting anyway (perhaps with caveats), knowing that it has some flaws but thinking it may be of value in spite of those.

Why excellence matters

I think caring about excellence, and using that as a heuristic to guide my actions, has worked pretty well. It has typically led me to make good choices, that I’d endorse if I were to step back and use a consequentialist lens on things.

Why should this be?

I think there are three main mechanisms at play:

Making people better informed and wiser is a robust intervention, which doesn’t rely on particular models of the world being correct to be useful. Improving the understanding of people who are already near the top of the curve on this is particularly high leverage (because that understanding ripples out to others).
Written artefacts have influence on some kind of power law distribution. The most impactful really get a lot of credit. But people like excellence, so excellent things are more likely to be shared. Helping the almost-excellent to reach excellence is usually a much better deal than helping the lousy reach pretty good.
As Hamming alludes to in the quote at the start, the pursuit of excellence is helpful for us in levelling up our sense of what is good. After you have strived to make something excellent, diving deep to get an understanding of what that should mean, it will be faster and cheaper the next time you try for the quality for the same kind of work. But also, the quality of your quick-and-ready work will probably have improved, as you have a more nuanced taste for what excellence looks like.

The pursuit of excellence

Here are some questions that I’ve found can be helpful to ask others, when I’m trying to help them orient to their own projects:

You’ve got some fire for making something happen here — what does the fire ultimately want?
Can you let the fire write a first draft, or come up with a first plan?
- Don’t worry about being reasonable or defensible, that can come later — just try to say the important pieces directly
- Maybe start by saying them out loud (to me), if you don’t know how to write them yet
What’s the title of your thing?
- Titles matter for how people remember things, but they can also help you orient consistently
Should you be looking for a collaborator?
- [the ideal collaborator is often someone with complementary strengths — you might want someone who is in touch with the ways that your plans/drafts suck, but still cares about the underlying project, and gets on well with you]

The first three of those questions are about staying in touch with what’s special about the project — all the best versions of the project are about elevating that. The last question is a meta-level question about whether things can be set up for more effective pursuit of the excellent.

And here are some questions I sometimes find it helpful to ask myself:

Have I lost touch with the reasons I cared about this?
Could I just cut some of this content?
Who could I ask for feedback who might be able to tear the work apart, and help me see which pieces are adding something important, and where I might be using dog shit for frosting?

^
By “felt motivations”, I mean the things that drive me in the moment — that make me want to put effort into stuff, and that help to steer that effort. These are distinct from “things that I ultimately think are valuable”.
Indeed I think it can be a little corrosive to blur these things together — it feels like trying to cross a large country, and thinking that you can keep most of your attention on the global map, and spare only a small fraction for the local terrain. Maybe this would work for superintelligent agents, but I think it’s a bit doomed for humans.
Ultimately I think that felt motivations are kind of a personal matter. You can do some gardening with them, but they can’t be forced. Still, I think it is worth some discussion — some sharing of aesthetics, and heuristics — that may help people to better grasp felt motivations they’d endorse, and to avoid sending the implicit message (as I think sometimes happens with EA material) that you should try to adopt your sense of what is ultimately valuable as your felt motivation.
^
And because effective altruism seems great, I’m very into work to make effective altruism even better.
^
And honestly I can be a little weirded out when people’s felt motivations tie to that, because it feels so ungrounded — we just don’t have a good sense of which actions are robustly good for this goal, so it feels like the person not looking at the local terrain.
^
Contributors to making this present post a better version of itself include Anna Salamon, Max Dalton, and Raymond Douglas. Responsibility for flaws that remain is mine.