CoZE 3: Empiricism

post by alkjash · 2018-03-17T04:10:00.858Z · LW · GW · 4 comments

Contents

  Day 25: Empiricism
    Overly Conservative Lines
    Scientific Detachment
    Micro-Experiments
      Exercise
  Daily Challenge
None
4 comments

This is part 25 of 30 in the Hammertime Sequence. Click here for the intro.

The boy on the right has gone places. The boy on the left has a map. Whom do you marry?
Whom

Sometimes, I think that most of the value of the CoZE experiment lies not in the expansion of comfort zones but in the experimental attitude it conveys. A good map-maker must constantly check the territory; the trick is to figure out how.

Day 25: Empiricism

The comfort zone is the region in the environment you understand. It contains the locations you frequent, the skills you have mastered, the people you know well. The farther away you get from your comfort zone, the more unknowns you have to prepare for. The boundaries of the comfort zone are designed to protect you from exactly these dangers: the unknown unknowns.

Overly Conservative Lines

The lines of the comfort zone is very conservatively drawn. In the ancestral environment, mistakes were often fatal: failing a hunt, losing a duel. Even nonfatal mistakes were reproductively so: humiliation in the front of your tribe lasts a long time, and you have no place to run. In this environment, it was reasonable to draw the lines of the comfort zone conservatively, since failure was too costly to test.

What does scientific progress look like in this danger-fraught world? Imagine that every time a scientific experiment fails, the experimenter pays with his life. Science would have progressed much more slowly, if at all.

But the world is no longer as dangerous as the ancestral environment. People live longer, are healthier, and are more mobile between communities. Equally, there are larger and richer positive opportunities outside our comfort zones. These are the preconditions for the viability of the scientific method, and the reason we can now use the power of empiricism, in the form of CoZE experiments, to test our boundaries.

Scientific Detachment

A key realization to make is that the comfort zone is part of your map. That is to say, it makes testable predictions about the territory. Your stage fright is making a testable prediction about how awful the experience of public speaking will be, and how much permanent damage you might sustain from a mistake. Your fear of heights is making a testable prediction about how likely it is for you to fall off a tall ledge unsupported.

Once you understand that the emotional aversions that form the boundaries of your comfort zone are built out of beliefs about reality, the logical next step is to design cheap, safe ways to test those beliefs.

I desire to believe what is true.

Usually, you’ll find that the lines of your comfort zone are too simplistic and conservative, and there’s obvious ways to tiptoe past it without getting in trouble.

Micro-Experiments

One of the core insights I gleaned from Inadequate Equilibria is that modesty, in the form of Status Regulation and Anxious Overconfidence, is one of the biggest fences around your comfort zone. In that post, Eliezer makes the following recommendation that can’t be repeated enough:

Don’t assume you can’t do something when it’s very cheap to try testing your ability to do it.

Don’t assume other people will evaluate you lowly when it’s cheap to test that belief.

The comfort zone is a set of beliefs about reality. Test those beliefs.

At very least, take five minutes and try to come up with a cheap experiment to test your beliefs. For example, my light novel Murphy’s Quest was a cheap way for me to figure out if it’s really true, as my System 1 emphatically stated, I’m terrible at writing fiction.

Design cheap experiments to test your fears.

You’re afraid your ideas won’t be well-received? Make an anonymous account and post the gentlest form of them.

Let me repeat Eliezer’s advice again.

Don’t assume you can’t do something when it’s very cheap to try testing your ability to do it.

Don’t assume other people will evaluate you lowly when it’s cheap to test that belief.

One more time.

Don’t assume you can’t do something when it’s very cheap to try testing your ability to do it.

Don’t assume other people will evaluate you lowly when it’s cheap to test that belief.

Exercise

Pick something you believe you can’t do but haven’t checked. Set a Yoda Timer and design a cheap experiment to test this belief.

Pick someone you trust who you believe will evaluate you lowly. Test that belief.

Daily Challenge

Share an experience where you radically underestimated or overestimated your own ability.

4 comments

Comments sorted by top scores.

comment by tcheasdfjkl · 2018-08-30T03:15:04.134Z · LW(p) · GW(p)

How, though, do you actually test whether someone will evaluate you poorly? I guess sometimes it's possible to get that info but a lot of the time (I'd say most of the time) it's deliberately obfuscated.

comment by itavero · 2020-12-16T17:18:00.976Z · LW(p) · GW(p)

Overestimation: Running my first session in a legit music studio (after interning there) 

I had only mixed/produced music "in the box" (on my laptop with software) before and figured I could just sit down at the Neve and do everything else the same. Nope, it was a nightmare. We went 3 hours over on the first day because I had to figure out so much on the fly. The amount of unknown unknowns that came up was staggering - this mic uses phantom power, how does this work in Pro Tools, which input is the synth patched into, what button for talkback, etc. 

Luckily I got up to speed and now run sessions regularly, but that was a huge dose of reality.

Underestimation: Submitting a dev PR at work

Worked at a startup doing generalist tech/product things. I taught myself React and Python before starting but didn't think I was good enough to get a programming job. After talking to the devs they let me try out a ticket. I worked on it and submitted it. It got accepted and merged in. Later I realized, "this is programming". 

I was starting to do more dev work but the company shut down recently. Oh well. It gave me the confidence to learn more though.

comment by silentbob · 2019-08-22T11:12:35.917Z · LW(p) · GW(p)
Share an experience where you radically underestimated or overestimated your own ability.

Overestimated: being filmed for an interview for a promo video of my company. Didn't think much of it beforehand, but it turned out to be awkward as hell, zero usable footage emerged. Wasting the time of all the ~8 people in the room wasn't great.

Underestimated: Nothing too radical, but giving a speech at a big birthday party. Expected it to be decent as I generally enjoy public speaking, but it went smoother than I thought, people laughed at the jokes and I think most were actually interested in what I said. Some complimented me on the speech afterwards which was nice.

comment by theme_arrow · 2021-02-04T06:30:39.484Z · LW(p) · GW(p)

Overestimation: Interacting with external reviewers/customers at work. I thought I had useful things to contribute to discussions with external folks starting maybe 4-5 months into my job. I didn't understand how to handle those interactions tactfully (and overestimated the chillness of by bosses) and got slapped down pretty hard.  

Underestimation: Research ability as an undergrad. I kept thinking I was a fraud and doing terrible work right up until the day I won the research top prize in my department.