Thanks for sharing! I'm very happy that this seemed to have benefitted you so much. I'm curious, for how long after the cessation of the practice did you experience the benefits? I have the sense they haven't ceased yet but would like estimates in weeks.
It happened about a year ago and hasn't changed since :)
Even though last year was pretty hard for me in a lot of ways and usually I would have been crushed by everything.
I did do the occasional maintenance session. Maybe five 30-minute sessions over the year? Hard to say how much of a difference those made. Internally, it feels like very little, but they could also be critical.
Wow that's really great! :)
Awesome stuff! I can highly second the "loving kindness meditation in front of mirror towards yourself" meditations! I had some of the most impactful meditations of my life this way!
Thank you for sharing this! Do you think your program will work better for people with significant meditation experience? Do you think your own experience was somewhat contingent on the meditation work you did in the Finder's Course (beyond the discovery that you benefited from loving-kindness meditation, something more along the lines of the benefit from the meditation "reps" you'd been through)?
Good question! It's hard to say. I suspect that meditation experience won't be particularly relevant because most of the time when people "meditate", it's just concentration practice, which is not particularly relevant to these techniques.
I had probably a sum total of 1.5 hours of loving-kindness practice under my belt before this, so I don't think that'll be particularly relevant either.
Could be an interesting thing to measure though if I end up doing more studies on it.
Thanks for writing this! If I don't want to sign-up to the finders course, are there any resources you would recommend for doing the one-hour lovingkindness sessions?
Oddly enough, I haven't found any really good resources on this. Except for this one google doc I found ages ago that I can't seem to find again. I think the explanation I give here might actually be the best I've seen.
But also, the explanation is pretty simple, so it's less about understanding (which is relatively easy) and more about practicing (which is harder, but I find, still way easier than concentration practice).
The only pieces of instruction I'd add are that if you're finding it hard to transfer the feelings of lovingkindness to a new object, that means the object is too hard. You should find an object that's easier. It's the equivalent of lifting weights and jumping up to a weight that's too heavy. Gradual increasing weight is key.
Thanks for writing this! For accountability, I'll edit this comment a week from now with an update of how it's been working so far :)
Love it! Can't wait to hear how it goes.
One trick I've learned since teaching it to more people is that if you're having trouble generating the feeling, you just need to spend more time trying to find a good seed. If it's not easy, it's not easy enough. Keep looking.
Also, maybe try updating it on a reply-comment. That way I'll get a notification and can read it (I'm probably the only person still reading the comments here). If you edit it, I won't get a notification, but if you reply, I will.
What an amazing write-up. Thanks for this. I am an experienced meditator of 13+ years (Insight mindfulness meditation). However even when I try to practice self-compassion meditation, it's not easy for the practice especially on my own. I once participated at The School of Life event with an experienced self-compassion facilitator, and I felt a tremendous impact after a short group practice.
Many EAs suffer from impostor syndrome, leading to a lot of suffering and missed opportunities when it comes to impact. You might not apply for a job or for a grant because you think you're not good enough. You might not try an ambitious strategy because you're miscalibrated about your abilities.
I myself struggled with impostor syndrome for nearly ten years, then eliminated it a year ago, and I predict the results are replicable.
In this post I’ll explain how to do it yourself and describe the upcoming class I’m running where I’ll see if I can replicate it with a small group.
The very summarized version of the method is
Apply to attend the class here.
I was a confident kid, then I did an internship at 80,000 Hours back in 2013. Living in Oxford gave me impostor syndrome almost instantly. Turns out that basing your confidence on being a big fish in a small pond only works if you stay in a small pond.
Between 2013 and 2022, I had more or less chronic low confidence and high anxiety around work. I tried everything to get rid of it: mindfulness, cognitive behavioral therapy, internal family systems, memory reconsolidation, exposure therapy, acceptance and commitment therapy, talk therapy, and just plain old reason. Nothing made a dent.
When I had impostor syndrome, every time I made a mistake or performed less than perfectly, I’d beat myself up mercilessly. Even when I succeeded by virtually anybody’s standards, I would always have a reason for why it didn’t really count, and couldn’t everybody see all the ways I’d messed up?
I was terrified of receiving feedback because then I would find out that secretly everybody hated me and thought I was useless, and they were just putting up with me because they had to. I felt I was always one step away from being kicked out of the community and being eaten by bears.
I've since experienced a 90% reduction in my work anxiety. After failures, I no longer feel debilitating self-recrimination; instead, I am inspired to think about what went wrong and improve.
In the past, when I looked at pictures of myself, the first thing that came to mind was how ugly I was. My initial reaction is now, "Aw. I love that girl."
I can now look clearly at the things that I’ve done and see the bad and the good (with exceptionally wide confidence intervals, mind you. I am still an EA).
Even more appealing to this audience, it’s improved my epistemics and impact! Because fewer things threaten my ego, I find it easier to change my mind or disagree with someone. Even if it turned out that my charity was net negative in some way, I don’t feel that that makes me bad. It just means I have to re-calculate.
Of course, I’m not perfectly confident. Before this, my confidence was about a 3 on a scale of 1 to 10. I was a 9 every day while doing the confidence practice and for a month afterwards. Since then, I've maintained my level at roughly 7.5, requiring only sporadic maintenance sessions.
Overall, it’s hard to think of any change in my life that has improved my well-being more. You can see another great description of the benefits of it in this great post by Charlie Rogers-Smith, “Why you’re probably underestimating the benefits of self-love”.
The change very clearly started in the first week of the Finder’s Course (I wrote an in-depth review here) when I did an hour a day of loving-kindness practice.
I've practiced loving-kindness meditation a lot in the past, but it was always more of a "dessert" practice. After I’d had my vegetables of concentration practice, I could treat myself to 2-5 minutes of loving-kindness at the end. It was never the main course.
But for the first week, we focused exclusively on loving-kindness for one hour each day.
It was amazing and profoundly life-changing.
Loving-kindness exercises come in a wide variety. The one that I was doing was the one where you:
First, I focused on loving-kindness for others. That’s always been pretty easy for me. It was lots of fun, just radiating love toward people in my life and the world.
Then, motivated by Charlie Rogers-Smith's incredible post on self-love, I tried practicing it on myself. What if I tried radiating loving-kindness towards myself instead of others?
A mental wall slammed down.
“No,” a part of me said. “You definitely cannot love yourself.”
I immediately burst into tears.
So, you know, a perfectly healthy reaction.
While this session wasn’t exactly the most enjoyable one I’ve ever had, it was possibly the most fruitful. After a brief temptation to avoid such negative feelings, I realized that this was the clearest signal I was ever going to get to dig deeper.
I decided to dedicate the week to loving myself in particular.
I started with an easy seed of loving-kindness, then tried to find things about myself that were easier.
At first, everything was hard.
Could I love myself when I was doing good things? No.
What about bad things, then? Heck, no!
Perhaps I could picture myself as a child and like myself? Yet another no
How about when I was a baby? Surely there can’t be any reason not to love myself as a baby!
Nope. I was a pudgy baby. And obviously you can’t love pudgy babies.
Eventually, I found my “in”. When I was a baby, I could see my mother holding me and feel how she felt about me. Even if I had a lot of self-dislike as a child, I am certain my mother loved me.
From there, I was able to build up to harder and harder things. I realized it was easier to love myself when I imagined previous times in my life when I was going through a hard time, since it tapped into my compassion. From there, I felt love for myself in particularly potent scenes from my childhood.
There were a lot of things to work on, so I created a spreadsheet. (Because obviously, if you want to be happy, you’re going to need spreadsheets.) It has three columns:
I saved two views of the spreadsheet:
For the “seeding” view, I sorted by how easy it was to feel loving-kindness, and filtered out the ones that were hard. When I looked at this view, I could see and keep track of the things that would help kickstart the emotion of loving-kindness.
The “to work on” view was sorted by impact, then ease (controlling for impact, why not do the easier things first?). After I’d established the emotion, I’d switch over to one of the aspects to work on.
Here’s a simple example of each view with some real and fake listings mixed in:
I’ve also made a template with a few made up examples for anybody who’d like to do it for themselves. I recommend Airtable for this because I find its view options to be better. It’s free and easy to sign up for. Notion is also probably a good option.
Practical note: I recommend making it a policy to never share your emotion spreadsheets with anybody. Sometimes the people you love the most are not the easiest to feel loving-kindness towards, and people’s feelings might be hurt.
For example, my family cat has always been higher on the “ease” list than any romantic partner or family member. It’s not because I love my pet more per se. It’s mostly because the relationship with him is uncomplicatedly good. I don’t live with him and whenever I see him, it’s just pure cuddles and love. I never have fights about finances with the cat.
One particularly fruitful tactic was to do self-love practice while looking at myself in the mirror. I’d pick out the parts of my body that I usually disliked and would beam love and acceptance at them.
I like this one in particular because, since the trigger is so concrete, it acts a bit like a trigger action plan (TAP).
I used to look in the mirror and think critical thoughts about different body parts. Now whenever I see myself in the mirror, my default reaction is to think “Aww” and feel this wave of affection. In fact, I feel that in particular to the parts of me I used to dislike because those were the parts I worked on in the practice.
At the end of the week of doing self-loving-kindness practice, I was feeling a lot more self-acceptance and self-love, which felt incredible. However, it felt like it was only halfway there. I mean, I love my family’s cat, but I don’t feel confident in his ability to do good in the world. Self-love and confidence overlap but are not the same thing.
A few years back, I realized that loving-kindness practice could be applied to any emotion you wanted to develop. Just:
I call the general principle “emotion practice," since you are practicing the emotion.
Previously, I’d mostly used it for excitement, although the practice never lasted long. (I’d get too excited to sit still.)
I decided to “practice confidence," and that got me the rest of the way.
The steps are the same as loving-kindness practice, but with the feelings of confidence itself.
The feeling of “confidence” in this case was a combination of:
I think the third is the most important since confidence problems often come up at that point. When you make a mistake, you tend to over-update and believe that you are completely worthless.
For myself, my three easiest seeds were:
My go-to seed I used was to think of how I felt towards a friend or family member who I had deep affection for and belief in despite their flaws. I didn’t pick some of my terribly annoying friends who seem to be perfect. Of course, I have deep affection for them and believe they can do it. But of course I do. They’re perfect!
Far better were the friends I had a deep affection for and faith in but who also had massive, glaring shortcomings.
Because most people are deeply flawed. And yet, we love and respect them anyway. It wouldn’t even occur to us to not think well of a friend because they’re not at inbox zero every day or aren’t the highest impact person in the community.
That would be insane!
And yet we do it to ourselves all the time.
The hidden benefit of this common human quirk is that we’re already good at loving and respecting flawed people, so we can harness that to love and respect our own flawed selves.
Imagine a friend or family member who you have deep affection for and belief in. Really feel how much you think they’re great. Then imagine them saying, “But there’s somebody at work who’s achieved way more than I have."
Feel how your love and belief in them don’t waver.
Imagine them saying all of the things you might think to yourself, such as:
I'm providing more generic examples here, but you should tailor them to your specific situation as much as possible. This will make it more likely to stick.
Have them start with easy things, then graduate to harder things for you. Often the easiest things are ones that don’t even apply to you, so you have no internal resistance. For example, I’ll often start with the insecurities I know my loved ones have but that I’m lucky enough to not suffer from.
Perhaps the easiest way to seed the feeling of confidence is to play music that makes you feel confident. For me, epic instrumental music works really well, but everybody’s music tastes are different.
I recommend combining it with a simple visualization to make it more potent. No need for elaborate stories like in other seeds. You could just imagine yourself climbing a mountain or as a superhero doing something inspirational. Whatever works for you.
I recommend compiling a playlist of songs that you can come back to again and again. You don’t want to spend most of your session looking around for songs. Also, I’ve found that the more you use the same song, the faster you get into the state. I suspect that it’s because you’ve trained your brain to associate the song with feeling confident.
This is great, because then if you’re in need of a quick shot of confidence in the day, it can be as simple as turning on the song and listening to it with your full attention from start to finish.
Unfortunately, it seems to follow a curve of potency. In the beginning, it seems to get more potent each time, but after a while, it starts to diminish. I’ve never had it go to zero, but it’s worth always being on the lookout for new songs to add to your confidence practice playlist so you don’t run out.
Another seed I love is imagining myself as an old wise person mentoring a young hero. For example, imagining myself as Dumbledore talking to 11-year-old Harry, except switching the genders to fit my own. Sometimes I’d also use Mama Tala talking to Moana as my seed since Mama Tala is the best old wise person ever. Often I’d just make up a vague story that would fit with what I was going through at the moment.
For example, I’d imagine myself being an old wise woman who’s mentoring this young hero. I’d be watching her train and feel warmth toward her, proud of her progress so far. Then I’d see her stumble and kick herself for flubbing the move. Then I’d help her stand back up, radiate my confidence in her abilities, and say, “You’re so young. Of course you’re going to mess up sometimes. You are going to do great, just keep at it.”
Then to build up to being directed at myself, I’d have the young hero experience things that were analogous to my situation. For example, to deal with feeling overwhelmed with my To Do list, I’d have young Harry be struggling to manage his course load and then have Dumbledore look at him with wisdom and confidence, knowing that he’d be fine in the end.
The scenes don’t have to be canon. Go nuts. I had Harry be sick and not get much done and feel guilty about it. I had him accidentally hang up too soon on a videocall and wonder awkwardly if he should call back. I had him procrastinate on answering hard emails. It was almost like writing fanfiction scenes in my head, where I just used the base characters to play out my weird psychologically-beneficial fantasies of having an old wise figure help me through my issues.
The key is that it’s a scene that brings up feelings of believing in somebody’s basic goodness and ability to pull through.
Of course, these are just three seeds that worked for me. Experiment widely. Play around with what sorts of scenes or people you can think of to bring up the target emotion. Maybe you can think of particular times in your life when you felt you were really on top of things. Maybe you can start with particular aspects of yourself that you feel really confident about already. Maybe you can just remember how it feels to be rather drunk. Whatever brings up the target emotion will do.
It’s important to note that no seed will always work, so it’s important to play around with different seeds during the session. One friend of mine is my go-to to start the session with, but some days we’d had a small conflict, and it was harder to think of them without thinking of the conflict, instead of my underlying love and respect. That’s OK. Just come back to them another time.
The whole process is a bit stochastic, so you’ll want to approach it with a flexible trial-and-error attitude.
OK, so you’ve got the seed down. How do you apply it to yourself?
An important thing to keep in mind is that nobody is “confident” or “not confident”. That’s like saying somebody is “skilled” or “not skilled”. Skilled at what? And skill is a spectrum, not a binary. Likewise, people aren’t confident generally. People are confident about certain things, and it falls on a spectrum.
You might feel low confidence about your impact or to-do list, but feel really confident in your math and writing abilities. You might feel sure that people like you but feel insecure about people respecting you.
Due to this, you’ll not be practicing confidence generally, but practicing confidence for each particular subset of your life that needs a boost.
I recommend making a spreadsheet of the things you want to work on (template Airtable spreadsheet here). List everything you can think of, broad and specific. For example, if you have body image issues, list something like “I don’t like my body” but also list the specific things you feel underconfident about (e.g. “I don’t like my stomach”).
Then rate everything on two aspects:
Then start working through it systematically, starting with the things that would affect your life the most and are the easiest.
Your list will be personal to you, but here are some potential ideas:
As you work on the areas, change their “ease” ratings as they get easier.
First off, I wouldn’t recommend using this technique to change your confidence intervals. I’ve never tried it myself, but I do predict that would worsen your epistemics quite a lot.
You might think that this practice leads to different beliefs about yourself, but I found for me those effects were minimal. I already knew rationally that I had reasons to feel confident. I just never emotionally believed them. This practice just changed my ability to alieve in me being a good, competent human being.
Secondly, I’ve found that, if anything, it has improved my epistemology. I’m more able to look at arguments I would usually flinch from because I have this felt sense that I’ll be OK if I do. I can listen to feedback and update without it being a threat to my self.
A lot of bad epistemics comes from motivated reasoning, where you’re trying to protect an image of yourself as likeable and respectable. If you feel confident that you are already likeable and respectable, then there’s less motivation to avoid looking at certain arguments.
It feels akin to this quote:
Except it’s “a confident person is never afraid of an argument proving her wrong. Her trust is not in never being wrong, but being fine even if she was wrong.”
Notably, I believe that certain types of confidence lack this property. For example, one subset of confidence is believing that you’re always right or always good. This will obviously lead to worse epistemics and ethics and should not be cultivated.
Rather, the practice I propose is more along the lines of “even if you mess up, I still love you and believe in you”. This allows you to see bad arguments you’ve believed calmly and dispassionately instead of with self-loathing, anxiety, or denial.
The honest answer is that I don’t really know. I’m not even sure if it works at all! That’s why I want to run an 80/20 experiment on it to see if it generalizes to anybody else’s brain (see last section for more about the experiment).
However, that doesn’t stop me from speculating!
Here are some of my current top hypotheses for what makes this work:
There are two main ways you can put this into action:
I’m running a 3-week class that’s doubling as an 80/20 experiment to see if this technique generalizes to other brains.
Some potential benefits of the class:
Some potential drawbacks of the class:
Hopefully this post explains it enough that you can do it on your own. The main benefit of doing it via a course is accountability and mentorship.
However, you can get those benefits in alternative ways:
If you try this out, please respond in the comments with your results! Better yet, pre-register your experiment in the comments, then report back in three weeks about your results (including if you decide to stop partway through).
There are two main benefits to that:
Also, if you request it in the comment or a private message, I’ll follow up with you in three weeks to add some extra oomph.
I’d also appreciate you replying in the comments about how it went for you, even if you didn’t do the full three weeks. That will provide information about success rates and common blockers so that I can modify and improve the practice.
Ideally you’d let me know:
It’s also totally fine to send this to me as a private message over LessWrong if you’d prefer to not say these things publicly.
Social accountability is one of the most effective ways to get humans to do things. Use that to your advantage. Rope in a friend and practice at the same time every day. Make it location-independent and do it over Zoom. Or simply check in with each other every day about how it went.
I also recommend Boss as a Service, in general and for this in particular. It's a service primarily aimed at self-employed individuals in which you tell someone your daily goals and they check in with you to see how you did. You send them evidence of completion, and they’ll keep bugging you till you finish. They also help you solve problems when you hit blockers.
It currently costs $25 a month and is well worth it. I’ve been using it for around two months now and I can’t recommend it enough.
I was originally worried that it would just be another thing for me to feel guilty about, but it’s had the opposite effect. My “boss” has mainly been my cheerleader, celebrating my wins every day and compassionately helping me work out my blockers when they come up. It feels great to have somebody on your team, and that tiny amount of social pressure makes all the difference.
You can also hook them up with your Beeminder to make it so that if you don’t accomplish your goal, they’ll derail you, adding some bite to the accountability if you so choose.
Beeminder or StickK are great for this sort of thing. They’re platforms where you set a goal and pay money if you fail at it.
The main differences between the two are:
I mostly use Beeminder because I don’t like having to involve another person in all of my various goals, and I don’t like giving to charities or anti-charities if I fail. I don’t like giving to charities because then it doesn’t feel bad, removing the point. I don’t like giving to anti-charities, because then I’ve made the world worse, and that hurts too much.
However, every person works differently. A friend once made the committment to donate a large amount to Trump’s campaign if he missed a day’s workout, and that was so motivating, he never missed one since.
In summary, a potential way to remove your impostor syndrome is to:
Either way, I hope this post helps you replace your impostor syndrome with self-love. To no longer be filled with self-criticism and anxiety is one of the best feelings, and I hope you can have this as well.
If you liked this, you might also like:
Reminder that you can listen to LessWrong and EA Forum posts like this on your podcast player using the Nonlinear Library.
This was a distinctively wholesome read. I restarted my (mostly focus) meditation practice late last year, and I have been meaning to leverage that foundation for a loving-kindness practice as well. The details of your post have substantially motivated that intention. Thank you for sharing!