The Quickstart Guide to Fasting

It is no secret that I am a huge proponent of fasting, which has been well documented across most cultures throughout history. My belief in its benefits come from a convergence of anthropology, biochemistry, complex systems theory, medical case studies, (low quality) scientific studies, and personal experience. This post is meant to be a practical introduction to fasting, not intended as a comprehensive literature review or justification. Read on if you would like some advice on how to get started and troubleshoot the initial adaptation period! [Read more…]

High trust high investment parenting

I like labeling myself as an “unschooler”, but often I’m interested in explaining what I’m going for with my parenting to people that don’t already have a detailed and nuanced model of what unschooling is! My best description of what unschooling means to me is  that I’m pursuing a high trust high investment strategy. I think most of my parenting decisions can be derived from those two principles.

High Trust

Unschooling advocates often talk about trust. Peter Grey, who talks abut “trustful parenting” in his book. And John Holt summarizes his approach as “trust children”. 

The word “trust” seems to capture a lot of what I care about, but it also requires explanation. When I hear people talking about why they don’t trust their kids, they tend to talk about how kids, especially little ones, don’t understand all the implications of the stuff their choosing to do. Which is totally true! (Of course, neither do adults fully meet this standard, but on average they have a much better model of how the world works.)

The sort of trust I’m talking about is more a characteristic of a relationship than a characteristic of a person. If a friend of mine who didn’t know Lydia well asked me “how much should I trust Lydia?” I wouldn’t have a clear answer. I could share a detailed model. I’d trust her to know what’s physically safe for her, I’d trust her not to physically hurt anyone. I’d trust her to give accurate answers about whether a toy belonged to her or her brother. I wouldn’t trust her to know which direction the playground was. I wouldn’t trust her to supervise her little brother around a dangerous object. I wouldn’t trust her to accurately predict whether she would pee if she sat on the toilet.

But to me, having a trusting relationship with Lydia doesn’t rely very much if at all on tallying up the ways in which I can trust her. It’s more of an attitude that I trust that she’s doing the best to make sense of the world given the constraints she’s experiencing, whether they are legible or not. It means that if she says something is important to her, I believe her, and if she acts like something is important to her, I believe her. Even if she can’t explain it and it doesn’t make sense to me. I trust that there was something important going on in her brain when she wanted to watch videos of people taking toys out of eggs and talking about them, and wanted me to put her toys in plastic eggs so she could take them out. I have theories about what was going on with that, but my trust isn’t contingent on the theories seeming likely to be true.

High Investment

I’ve also committed to using a high investment strategy with my kids. On the most basic level, I stay home with them and hang out with them for most of my waking and sleeping hours, so I’m investing a lot of time. We plan to homeschool, so there will be more of that. I also read books, think a ton about parenting philosophy, and discuss parenting stuff with my friends. I feel committed to working through neuroses of mine that make me a worse parent. (And I’m not claiming this stuff is the best way to invest in kids. Just that these are some ways I’m investing a lot. There are high investment strategies that look different from mine, but I think this stuff is clearer with examples.)

If my kids are doing something that’s inconvenient or frustrating for me, I make sure to consider solutions that involve me changing and doing work, not just solutions that involve them changing and doing work. I’ve been willing to do stuff like sleep with my kids as long as they want, nurse for years on end, carry them around a bunch, and fill my house with toys they enjoy, even though it takes work to organize them and clean them up.

I hesitate to write all that stuff, because it feels like I’m bragging or something. There’s an overall cultural narrative that it’s good to do stuff for our kids. (And naturally a complementary narrative that it’s bad to do too much for our kids.) There are also lots of things I don’t do for my kids. And I expect to invest less time in each kid as they get older. My resources are limited, and I spend some of them on other stuff too! Right now, I’m writing a blog post while they’re both with a babysitter they like, and that seems fine.

But it seems worth mentioning that it seems intuitively correct to me to invest a pretty large chunk of resources in my kids. I also don’t mean that I feel like I’m investing more than people in my rough reference class. I think most people invest a bunch these days. But by historical standards, I think we’re all investing a lot, so I don’t want to take it for granted!

Why these labels?

Articulating my parenting values seems useful for a bunch of reasons, and a big one for me is that it makes it easier for me to have useful and respectful conversations with people who do things differently. 

I’ve found that when I’m doing things differently from my friends, lots of the time it’s just because we have different circumstances and comparative advantages. I have multiple friends who feel happier and more energized when they get out of the house with their kids, so they think of fun stuff to do. I tend to feel more sane if my kids and I spend a lot of time at home, so I try to think of ways to make staying home fun. I have friends that have done soul searching about how to support their kids when they want to do physically dangerous things. I tend to be encouraging my kids to try things that are a little outside their comfort zones, because so far I’ve had pretty cautious kids. 

But sometimes things really do seem to boil down to value differences, and the ones I’ve listed above are the ones I’ve encountered and recognized most often. For example, there’s a paradigm that seems coherent to me where at younger ages, the parent prioritizes the kid obeying and submitting more than trusting. There also seems to me to be a coherent position where people think there’s something unbalanced and bad about investing as much in our kids as modern society seems to encourage.

So, if I have a deep disagreement with someone about parenting, the two values I listed seem like the most fruitful areas to explore. 

ETA: link to Facebook comments

Instrumentally Caring Intrinsically

A way of thinking that I’ve been using for a few years now, but I don’t think I’ve ever written up, is the idea of instrumentally caring about things intrinsically.  

Caring about something intrinsically is often very useful for coordinating with others.

When you care about something for its own sake:

  • It’s easy to strongly and coherently signal that you care about it.
  • People (rightly) expect that your caring will be fairly stable.
  • Your intuitions, aesthetics, and gut feelings will be aligned in such a way that you can act on your caring in realtime.
I remember being an overly analytical kid who wondered whether there was something fundamentally incoherent about caring about things other than my own sensory inputs. I’ve now come around to the opposite idea. Intrinsically caring about only my sensory inputs is incoherent–there’s a lot of utility that caches out in the form of cool sensory inputs that you can only unlock by intrinsically caring about things other than sensory inputs.

I see a lot of conversations break down when people can’t, or won’t, justify why they care about something. And I think there is something that can be a little “off” about trying to come up with justifications for intrinsic values, and in my experience it can actually mess up people’s epistemic to try. Then again, if the things you care the most about become semantic stop signs, I believe you’re leaving a lot of value on the table.

Instead, when these types of conversational roadblocks come up, I recommend people shift to discussing what’s good about caring about something intrinsically. 

A while back, someone on my facebook feed stirred the pot by doing a cost benefit analysis (IMO reminiscent of David Friedman’s stuff) of whether to call the cops on a bike thief. He got some pretty strong pushback from people who implicitly rejected his frame and said stuff like “fuck bike thieves”. 

According to me, the right way to continue the conversation at that point is to ask how the world looks when we do cost benefit analyses of reporting bike thieves vs. how it looks when we become morally outraged when we see bike thieves. This way, neither party is required to directly invalidate their sacred values by the things they protect at the object level, and the people can actually exchange information about their worldviews and whether they disagree with each other’s.

The Heritability of Everything

The gold standard in heritability estimates is the twin study, which involves looking at identical and fraternal twins, raised together or apart. This allows the cleanest test of decomposing the variance in observed traits into genetics, shared environment (factors equally affecting all children raised together), and non-shared environment (everything else, including random noise) contributions.

Generally it is assumed that the effect of parenting is equated with the shared environment, though there is clear evidence that parenting can differ substantially between siblings of the same parents and account for a significant fraction of non-shared environment, and the shared environment by definition also captures e.g. the neighborhood in which you grow up. Generally there are many caveats to apply to heritability estimates, particularly that they are only defined within a given population and may not apply as well in extreme cases, but nonetheless they are our best estimates as to the effects of genetics, and the effect is undeniably large.

An extremely ambitious meta-analysis of all twin studies was published in May 2015, reporting heritability estimates from 2,748 studies featuring over 2 million twin pairs, encompassing virtually every published study to date. The researchers have made a data visualization tool available if you wish to dig down into various aspects of the study, though it’s fairly opaque if you’re not familiar with the field’s jargon.

Across very broad domains of health outcomes, almost everything falls within the 40-60% heritability range, with cancer as a representative example being 46% heritable. Similarly, neurological variables show about 50% heritability (with little shared environment involvement), while cognitive and psychiatric outcomes are similarly heritable, but also have a nearly 20% shared environment component. Social values appeared to be 31% heritable, but shared environment played nearly as big a role at 27% explained. Similarly, social interactions were 32% heritable, with a somewhat smaller shared environment component of 18%.

Drilling down into more specific categories of interest, intellectual functions broadly were highly heritable at 67%, while more specific executive function metrics were 51% heritable with a high 24% shared environment contribution. Mood disorders were highly variable, from bipolar being 68% heritable to depressive episodes being 34% heritable. Height and weight showed 63% heritability, with relatively large 30% and 20% shared environment contributions respectively. The more specific values and social variables were mostly in line with the overall findings. Tendency towards religion and spirituality was 31% heritable with an even larger 35% shared environment component. Basic interpersonal interactions were similar, with 30% heritability but 36% determined by shared environment.

To summarize, basic variables in terms of intelligence, height, and weight are primarily determined by the genetic contributions. Most health and psychiatric outcomes fell somewhere in the middle, but still showed roughly half of variance explained by genetics. Variables relating to fundamental values (e.g. religion, politics) and social interactions (e.g. emotional intelligence, relationships) were by far the most malleable traits, with roughly equal contributions from genetics and shared environment.

Discussing Tricky Stuff

When I’m discussing something hard, especially over the internet, I try to be careful about:

  1. Searching for and explicitly talking about tradeoffs, even when one option seems like a clear overall win to me.
  2. Trying to keep the ratio of positive to normative statements very high.

I don’t think these points are either necessary or sufficient for intellectually honest discussion, but I think they help. And when other people aren’t doing these things, I find it harder (in the sense of more cognitively effortful) to learn from them. 

With respect to the first point, I don’t think it’s very helpful, if it all, to emphasize tradeoffs that “seem reasonable”, but that the person presenting the idea doesn’t actually think are in play.

I imagine this post would be better with examples, but I won’t include any for now. My plan is to write some blog posts soon about object-level stuff and try to follow my own advice.

What Has Changed my Political Beliefs

I think it’s safe to say that political beliefs are one of the most sticky types of beliefs we commonly hold. By some measures partisan polarization is at record highs for the modern era (though these figures are also debated). Politics are also beliefs that provoke some of the strongest arguments between differing viewpoints, and the strongest consolidation among shared viewpoints. Eliezer warned us to be particularly careful when grappling with these ideas.

But, as good rationalists, all of our beliefs should be subject to updating upon receiving further information – and when I look at my political beliefs over the years, I see that they have indeed changed, in some ways massively, in other ways slow and subtly. I thought it would be an interesting to lay out what the drivers of these changes were, as a case study in the art of changing one’s mind.

[Read more…]

Summary of How To Talk So Kids Will Listen & Listen So Kids Can Talk

How To Talk So Kids Will Listen & Listen So Kids Can Talk is a parenting / communication book written by Adele Faber and Elaine Mazlish. While this book is specifically intended for parents to have better relationships with their children, the vast majority of the advice contained within applies universally to all interactions, and I have written this summary specifically to abstract away from parent-child relationships. I consider the first chapter alone better at helping people internalize the principles behind nonviolent communication than Rosenberg’s entire book. HTTSKWL is currently by far my most highly recommended communications book, and because it is appealing to parents and children it is a remarkably easy read.

Note that unlike most books, this one contains a very high ratio of exercises and prompts and anecdotes relative to its advice. The authors recommend going through the book slowly, and doing all the exercises. This summary will only contain their explicit instructions – I highly recommend buying a copy of the book and completing it. The many specific example conversations will give a much better understanding of the principles I lay out here than I can convey in a summary.

[Read more…]


For ages, I’ve been hearing about how important “boundaries” were, and I’ve never been quite sure what people even meant by the word. A while back I went to parenting discussion about boundaries, and I think I finally figured out what the word means! (To me, that is. I’m not sure everyone agrees on how boundaries work.)

The word “boundaries” always bugged me because it was vague, and seemed to include some assumptions that I wasn’t sure I agreed with. One such assumption was the idea that good boundaries were bright lines. (Now I think it depends.) “Boundaries” also sounded like the sort of thing I was supposed be very careful to respect, more so than mere “preferences”, and I was confused about exactly why.

[Read more…]

Some Indian Recipes

These won’t be particularly well-formatted, and I don’t have any pretty pictures to go with them at the moment, but I’ve had a few requests over the years for the Indian recipes I regularly cook, so here they are!

[Read more…]

A Theory of Economic Development

If I had to give the Most Underrated Professor Award to anyone I studied under, it would easily have to be Professor Meir Kohn. His work falls squarely outside the paradigm of mainstream economics, which (coming from me anyway) could not be a higher compliment – yet also makes it difficult to get traction inside the field. Kohn himself wrote a great essay contrasting the mainstream paradigm of Samuelson and Hicks with the lineage of more qualitative thinking descended from Adam Smith, including fields like economic history and new institutional economics and public choice theory among others.

He has been working on a theory of economic development for two decades now, and in my opinion it appears to be substantially correct. He first composed an unpublishable opus about European economic development from 1000-1600, which I read in its entirety. Then he wrote a more condensed version, where he applied the theory to China as well as Europe, which is likely going to be published in the next couple of years. In private correspondence, he has indicated that he and his students are now applying the model quite successfully to analyzing other economies throughout history.

The first chapter of his first opus contains the best description of his overall model I have yet read. Because I hold it in high regard, believe it to be fundamentally true, and I refer to it often, I am posting my summary of it below. All errors and omissions are mine.

[Read more…]