Mini Blog Post 14: Minds should make sense
(The structure of this post is experimental - please let me know whether you think this works well! The process I use approximately follows this framework, and it might be helpful to read that first)
I think a really key life skill is the skill of empathy: understanding other people’s internal experiences and why they act the way that they do. This is visibly a thing that most people are bad at, because empathy is hard. As with all important things, empathy can and should be optimised for. My goal in this post is to outline how I’ve gotten better at a specific form of empathy, a skill I consider highly valuable. I find that advice like this if often frustratingly vague and fuzzy, so I’ll try my best to make this as specific and actionable as possible.
A common theme of my previous posts has been that problems are for fixing - that there are many ways my life can be better. And that by deliberately applying effort in the right ways, I can make progress on these problems. I think this is an important insight, but can often seem fuzzy and abstract, and not very actionable. So, in this post, I’m going to try to highlight the concrete thought processes by which you might make progress on this problem. (This is a highly stylised process, and not what I actually went through, but hopefully somewhat instructive!)
My process roughly breaks down into:
Understand the problem - look for examples, think about, come up with explanations, get data
Think about possible solutions - once I understand what’s going wrong, think about what behaviour/environment changes could be that would lead to better outcomes
Make this concrete - figure out which action changes, and what concrete things should trigger them - make it something that will actually work in the moment, and can be implemented
Implement this! Take steps to ensure that I actually have a meaningful behaviour change, and that this is now less of a problem
The first step to understanding a problem should always be to look for examples! Figure out when the problem come up, what goes wrong. You don’t currently understand the problem, and analysing real world data is one of the most valuable ways to make progress. The underlying insight is that you want surface area on the problem: to understand its surrounding context, exactly what it feels like from the inside in these situations, when it comes up, etc. Solutions to difficult problems will require creativity and thought, and surface area gives you the hooks into a problem that let you start being creative.
Thinking of examples is hard, and more is always better! I highly recommend setting a 5 minute timer, and thinking for the full 5 minutes. For me, examples of lacking empathy break down into a few categories:
Friendships - not understanding other people’s preferences
Other people often have different preferences from me, and this is important to realise! But it’s easy to lack empathy, and to not properly internalise that other people are different, and instead to feel frustration with them. Because they care about things that are so clearly unnecessary!
For example, I recently had a disagreement with a friend about whether we should have video calls or audio calls - I feel more comfortable with video calls because they can give me clearer social cues, while she finds video calls require much more social energy
But, in the moment, I just felt frustration - obviously my preferences were important, while she was making a big deal out of nothing! We had to have a proper discussion about it afterwards to get a better feel for the other person’s perspective - in the moment her view didn’t feel visceral to me
Important note: A key part of studying examples is looking at the relevant emotions, and the internal experience. Not just understanding it on an abstract, rational level.
Teaching - not understanding what’s going on in the student’s head
I enjoy teaching, tutoring, and explaining things. And a pretty common occurrence is that the student asks a super dumb question. A question that doesn’t make any sense, and shows they clearly don’t understand what’s going on.
Even if you don’t do formal teaching, I think this example also applies heavily to explaining a concept to a peer
And in practice, I feel frustrated. It feels like they’re just not understanding, and my reaction is often to repeat the point again.
This often means the lesson could move on and they’d drop about the question
But often it comes out later that the student had misunderstood an important point, which came out later. The question made sense in their head. And once I realised this, I could often explain their exact misunderstanding, and reconstruct a correct picture in their head. But the lesson had gone on another several minutes without them really getting it
Disagreements - not understanding why the other person believes what they believes
Talking to people I disagree with is a key component of seeking truth, but this is often hard. It’s easy to get frustrated, and for the conversation to break down. To conclude that the other person is an idiot, or not acting in good faith.
These points of frustration are the points where I most lack empathy. And when I dig into them, I find that often I’ve just made a knock-down argument that they brush off, or they’re talking about something that seems clearly irrelevant or unimportant.
Eg, they’re saying that optimisation is a bad idea because you can never know your true goals
I find these often come up when either person uses specific jargon which has meaning to them, and to their subculture, but not the other person. This is something I often notice going wrong in conversations between people with political disagreements. Eg marginal utility, safe space, micro-aggression, efficient markets
At this point, it’s good to sanity check whether this is actually a problem. And yes, it clearly is! I am acting in ways that make other people feel worse, and don’t help me achieve my goals. Possibly solutions will be too high cost to implement, but I should definitely explore more!
And now I’ve analysed these examples, I think I have a model for what’s actually going on underneath the surface (this step isn’t always possible, but highly worth thinking about!). When you think about it, it’s really weird that we’re capable of empathy at all. Human minds are complicated. A good heuristic: whenever I can do anything naturally, but couldn’t put into words exactly how I do it, there’s a lot of work being done by my subconscious mind (my System 1). Eg, I’d struggle to program a computer to have empathy! And this suggests that I have mental routines automatically built in that are doing a lot of the work in empathy. This is a subconscious problem, and so will require different solution techniques than simply correcting a misunderstanding - changing my subconscious mind requires more than just an expression of will.
And the most obvious built-in mental routines to understand emotions are my emotions. This suggests that a lot of my failures of empathy come from implicitly assuming that other people think like I do. And this is a really common failure mode called the Typical Mind Fallacy. And when I look back at those examples, this fits well. I don’t, intuitively, understand the other people have different preferences, or different levels of understanding, or different worldviews and models of the world. And once this is pointed out I can do something about it, but this is an error with my intuitions. It happens in the moment, and any solution must also live in the moment. (A failure of empathy can also stem from not understanding that other people have deep internal experiences at all, but I didn’t find that that was my error here).
So, this naturally suggests that a solution will be to notice when this is happening in the moment. And then to change how I was acting! To be curious, to explore, and to try to understand the other person. What was going on in their head, what led to their actions, and how are they currently feeling. There is a skill to doing this well, and especially to doing it in a way that ensures the other person feels comfortable and as though you’re acting in good faith. I outline some of my thinking on these skills when talking about how to approach learning from conversations. But this is an empirical question, and the correct way to answer those is with data! And I found that when I did realise my error in the moment, I was normally able to make things better. This wasn’t perfect, but realising to be curious, to ask questions, and to try to paraphrase back what I thought was going, was often enough to resolve the problem. So, empirically, this is what a good solution should focus on.
The key insight here can be summarised as follows: Minds should make sense. Other people’s minds follow an internal logic, and make sense from the inside. I know that mine does! And my error is implicitly not realising that other people’s minds can make sense from the inside in different ways from mine. But that this internal logic can be understood, and with time and effort I can build a better model of what’s going on inside their head. And that when I notice that somebody’s mind doesn’t feel like it makes sense, the flaw isn’t with them, it’s with my model of them.
So, now I know that the solution is to be curious and try to really understand the other person. But knowing the right thing to do is just half the battle - a problem is only resolved when I actually implemented this, and could reliably do this in practice. At this point it’s easy to just press the Try Harder button - “I’ll just pay attention more, now that I’m aware of what’s going wrong!” And this is often a good place to start, and can become a reflex in the long-term. But a good long-term solution must eventually be robust. Something that happens even when I’m not paying active attention. To be a real solution, it needs to become a reflexive mental habit. It’s hard to form these mental habits, because they live in your subconscious mind. My subconscious mind is not clearly legible to me, and it’s not obvious how to get it to do what I want. I call my exact process for forming these habits Noticing, and I outline it in this post.
One of the key points is that a good mental reflex needs a clear trigger. And the bias of the typical mind fallacy isn’t labelled as such from the inside. It just feels like the other person is wrong. So I can’t make this into the trigger. But biases often come with subtle emotional cues and contexts which can become good triggers. And when I look back through the examples, and introspect on what they felt like in the moment, the common trend is that their mind does not make sense to me. My friend expressed a preference that didn’t make sense to me, and this created a sense of frustration and confusion. The student asks an inexplicably dumb question, and I feel a sense of frustration and desire to repeat the point. Someone brushes off a knock-down argument, or makes a seemingly-irrelevant point, and I feel a sense of frustration and confusion. And this sense of frustration and confusion, with an underlying sense that their mind isn’t making sense to me, is a clear trigger that I can notice in the moment.
And, now I have a concrete plan for a solution, I can get to work implementing this with Noticing. Roughly, by making the trigger feel very visceral in my mind and clearly important. This takes time, but once I reliably notice it in the moment, I can then decide to be curious and inquisitive. And once I’ve done this enough times, it can become an ingrained mental reflex: Notice the feeling of somebody else’s mind not making sense -> be curious and inquisitive. These mental habits are often difficult to build, especially if you’re starting out. And they won’t always trigger. But I found this one was fairly doable. There was a clear trigger, and whenever it did trigger it was very obvious that it was helpful, and this felt rewarding enough that I felt motivated to continue!
The implementation step is deceptively easy to forget, because a lot of the intrinsic reward for thinking about a problem is finding the creative solution, not the grunt work of implementing. But all the prior effort is useless without this!
Overall, I’ve found empathy has been a really valuable skill: I’ve become a better teacher, more pleasant to be around, and better at understanding people I disagree with. I’m nowhere near perfect at it, but feel like I’ve made a lot of useful progress. And if the ideas in here have connected with you, and especially if my solution feels like it might work for you, I strongly urge you to block out some time to explicitly practice the ideas in here! This can sometimes backfire, especially if someone else is acting in bad faith, or just a bit of a jerk, but I’ve found on the whole it’s obviously worth it.
And I hope that seeing a more detailed example of how to actually go about solving a personal problem can add some useful context to previous posts! The exact process will differ a lot depending on you and your problems, but I think the underlying ideas in here transfer pretty well!
