The Meta-Application of Rationality

Last weekend, I had the great pleasure of being able to attend an Applied Rationality workshop held in Berkeley, CA. I can say without a doubt that it has had a serious effect on the trajectory of my life. I want to talk briefly about this more specifically. I do have much more to say about the workshop when I get a chance, and I’d love to talk about it either privately or in the comments.

So, my life’s trajectory: those familiar with calculus will understand the following in terms of derivatives, but I’m going to use the analogy of a moving vehicle. If you want to describe the vehicle, you can talk about several things. The most obvious (and most immediately useful) is “where is it?” which often called displacement. The next most obvious/useful is “how fast is it going and in which direction?” also known as the car’s velocity. But, unless a wall or gas shortage stops the car before the relevant moment, the most useful thing to know about a car is how far down the gas pedal or brake is. This is the acceleration and ultimately it will tell you a lot more about how far the car will have gone by tomorrow than the velocity or displacement. You need to know roughly where the car was and how fast it started, but smaller changes in acceleration are much more important.

At the rationality workshop, I learned some things that definitely took me beyond where I’d been before (displacement / location). While at the workshop, I was definitely growing faster than usual, but right now I might be slower than usual as I focus on consolidating everything. Here’s why I’m not worried: what I definitely increased was my ability and propensity to assess how fast I’m going and to go faster still. Have a look at the chart to the right:

A chart demonstrating how much faster an exponential function increases than either a linear or polynomial function

Ideally, you want to be the yellow line on top.

Accelerating returns

It should be clear that the yellow line on the top is increasing much faster than the others. It begins just as slow, but because it’s not only getting faster but getting faster faster, it quickly takes the lead. This is the model I’m adopting. Short-term, it’s fine if my displacement and velocity are taking a hit. In the long run, by getting better at getting better (by applying my rationality skills to becoming even more rational) I can easily overtake my hypothetical other-self and never look back.

However, I’d like to think that I haven’t just increased my acceleration but that I’m increasing the rate at which I’m increasing my acceleration, and increasing that rate as well, and so on indefinitely. This resembles an exponential function, which is shown by the yellow line above. If I were just accelerating at a constant rate, I’d be the red line in the middle, and if I were just holding steady at a certain speed I’d be the shallow blue line. One fascinating principal is that no matter how many layers you do this acceleration like xbigNumber, the exponential curve always has more. That means that if I were to ask “who will win the race?” between x1234567890 and ex, then ex will always win as time goes on. I plan to do the same.

How?

How am I going to do this? I have some general thoughts on this which involve developing personal scaffolds that will make it easier to install what I learned at the workshop as habits, but I’ve modified my environment as well. Or rather, I’ve decided to put myself into a different environment. I’m working at a software startup in San Francisco right now, but rather than live in SF I’ve decided to spend the next 3 months living in Berkeley. This will make me much more likely to attend various free training events held by CFAR (the Center for Applied Rationality, who ran this workshop) and will also make it easier for me to hang out with some of the people I met at the workshop, encouraging me to focus on these various aspects of my life. By investing the time and effort now to make my next 3 months more fruitful, I hope to ultimately set up self-reflection habits that are exponentially powerful. I’m not sure how I’d tell success from a close approximation, but honestly I think that if I get the first 5 or so layers than there really isn’t much to be gained at that stage.

I was surprised by the variety of rational techniques that were relevant to my decision to live in Berkeley rather than SF. They included not only a substantial consideration of my preferences and goals, but also a recognition of how I’m not as rational as I’d hypothetically like to be, and therefore I can predict that if I’m living in SF I won’t go to as much CFAR stuff. Similarly, while I might generally enjoy not having a 1h-each-way bus commute, I expect to be able to devote a lot of that time to this work, where previously I felt like there was a bit of imbalance such that most of my waking hours were directed towards working on things for the company I’m at. Then, when deciding if I wanted to accept a certain sublet option, I also weighed factors like, “how much more valuable of a place do I expect to get by waiting / more effort, and is it worth it?” I decided it wasn’t, so I accepted. Personal happiness is hard to judge, but we do our best.

One could also consider a function of personal utility. Again, I expect to slow down briefly but ultimately go so much faster/further. This is fairly easy to measure but very hard to judge cross-domain.

What’s a moment that not just changed your life, but changed how you approach it?

A portrait of Malcolm Ocean

I'm Malcolm Ocean.

I'm trying to figure out how humans work so I can help make humanity work. More about me.

Stay focused on what matters
Check out Complice, a web-app that I built to help people achieve their personal & professional goals. Complice logo
Follow me on Twitter!
420 feed subscribers