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

One of the overarching themes from CFAR, related to The Strategic Level, is that what you learn at CFAR is not a specific technique or set of techniques, but the cognitive strategy that produced those techniques. It follows that if I learned the right lessons from CFAR, then I would be able to produce qualitatively similar – if not as well empirically tested – new principles and approaches to instrumental rationality.

After CFAR, I wanted to design a test to see if I had learned the right lessons. Hammertime was that sort of test for me. Now here’s that same test for you.

The Final Exam

I will give three essay prompts and three difficulty levels. Original ideas would be great, but shining a new light on old hammers is also welcome!

Prompts

1. Design a instrumental rationality technique.
2. Introduce a rationality principle or framework.
3. Describe a cognitive defect, bias, or blindspot.

Difficulty Levels

Bronze Mace mode. Write one essay on one of the topics above.
Steel Cudgel of the Lion mode. Write two of three.
Vorpal Dragonscale Sledgehammer of the Whale mode. Write all three. For each essay, give yourself five minutes to brainstorm and five minutes to write.

Here are my answers.

1. Cooperate First

There’s an old story about a famous painter of the Realist school who spent a whole year of his training painting still lives of eggs. Each day, he would draw a single egg over and over. He must have produced thousands of sketches and paintings of eggs. His teacher knew exactly how important fundamentals are.

This same motif is deeply embedded in stories all over the world:

Return to fundamentals. Practice your fundamentals.

The iterated prisoner’s dilemma is one of the fundamental lessons of rationality. The world is more like a number of iterated prisoner’s dilemmas than you’d think. Human beings are more like tit-for-tat players than you’d think. It follows:

Cooperate First!

The first move you make in any interaction with a new acquaintance should be a cooperate, even if you expect them to defect. Perhaps even if you observe them defecting already.

Here’s a lesson I learned from meditating on the maxim Cooperate First:

Cooperating First feels like accepting an unfair game from the inside. There will be many situations in life where things are framed in a slightly but noticeably unfair way towards you initially. Err on the side of accepting these games anyway!

2. Below the Object Level

One of my main complaints about rationalists (myself included) is our tendency to escalate to the meta-level too often. For example, in any given discussion, arguments over general discussion norms get much more heated and lively than any discussion of the underlying subject matter. We need to spend more time at the object level, touching reality, making experiments, testing our hypotheses.

The move I use to combat the tendency to escalate meta, I call looking below the object level.

Looking below the object level is like the move HPMOR_Harry does to achieve partial transfiguration: continually upping the magnification on your mental microscope to actually stare at the detail in reality. Reality is so exorbitantly detailed it’s overwhelming to take it all in. Try.

Look at the folds in your clothes, the way light and shadow play off each other. The way threads interweave. Pinch the cloth and watch the creases reorganize under your fingers.

Now reflect on this fact: falling water is attracted to both positive and negative charges.

What.

There’s so much going on under what we think of as the object level.

3. Pre-Excuses

Pre-hindsight is a version of Murphyjitsu where you query your mind for what you will learn from an action in hindsight. Pre-excuses are an unproductive cousin that often derail my work.

As a serial procrastinator, I notice a fairly regular pattern of thinking that appears the couple days before I have to meet a professor, and especially before meeting my thesis adviser. My mind is already spinning excuses on overdrive. Here’s what my mind sounds like a full day before I have to meet my adviser, when I think about the meeting:

Sorry, this paper took longer than I expected to read.

Sorry, I was busy from other classes, so I didn’t do as much paper-writing as I’d planned to.

Sorry, I got sidetracked by this research problem, so I didn’t finish the homework.

That’s right, I’m having these thoughts about how to apologize for not doing work even though I still have plenty of time to do the work. Even worse, I have these pre-excuse thoughts regularly even if I’ve done the work expected of me – it feels something like cushioning the fall in case it turns out I did it poorly.

And they’re usually not even good excuses.

This is part 29 of 30 in the Hammertime Sequence. Click here for the intro.

I find myself dragging my feet on the last couple days of each Hammertime cycle. From this and several other data points, I think current my writing attention span is around a week, and drafts and outlines sitting for more than a week feel too stale to finish. Had I known this in advance, I would probably have structured Hammertime as six 5-day sprints.

Reinforcement Learning?

What happens when reinforcement learning isn’t enough?

You playing a game of Go against sensei. On move twenty-four, sensei invades your three-space extension with devastating precision, cutting a group you thought was safe into two scattered dragons. The left dragon tries to run away, but sensei cuts its escape route off with a delicate leaning attack on your corner enclosure. It dies with abandon.

The right dragon, now facing the massive wall sensei built up by attacking the left group, tries frantically to make life locally. Its second eye is poked out unceremoniously by a well-placed tesuji. Because of your struggle, sensei has fifty points of territory and thickness radiating across the entire board. You resign.

What is a novice supposed to learn from a game like this? If your teacher leaves you to your own devices to review the game, you might easily conclude any of the following, if not a dozen other things:

1. Don’t make three space extensions.
2. Never try to run away.
3. Do not respond to leaning moves.
4. Sacrifice early.
5. Study life and death.

Let’s say you learn lesson 1, don’t make three space extensions. The next week’s teaching game, you dutifully plod out two spaces from each approach. Sensei’s stones are balanced and efficient while yours are over-concentrated and unimaginative. You lose handily by points.

What happens now? Do you return to three-space extensions, frustrated with two-space ones?

Over-correction and Learning Stopsigns

The Strategic Level is a CFAR flash class about learning strategically: updating in such a way that will actually prevent the same failure modes in the future. The kind of learning above is definitely not strategic.

As I see it, there’s two common and overlapping kinds of failure modes in learning, where the lessons learned can be worse than nothing.

The first kind is over-correction:

Had an argument: “I should be more understanding.”
Had a panic attack: “I should just care less about everything.”
Was a White Knight at Dragon Army: “I should just never trust human beings.”
Lost a Go game: “I should never make three-space jumps.”

Such overly general lessons can be cures worse than the disease. As your simple strategies progressively fail, you need to come up with and try more and more complicated strategies. You can’t just continually bounce between two extremes, refusing to stare the complexity of reality in the face.

The second type of failure is similarly unproductive:

I should have just read out that dan-level life and death problem!
I should have just studied chapter 3 instead of chapter 2!
I should have just tried to use the polynomial method on that problem!

I call these thoughts learning stopsigns. A common type of learning stopsign is of the form “should have done so and so,” where so and so is some arbitrary, brilliant, unreasonable choice you would never have made in advance. Just as semantic stopsigns masquerade as answers, learning stopsigns masquerade as lessons learned while not actually providing practical utility for the future.

The learning stopsign simply says: turn back, nothing to see here, painful thoughts past this point. It’s usually accompanied by a nonchalant shrug.

Strategic Learning

What does it mean to learn strategically?

Whenever you fail, try to answer the question, “What way of thinking would I have had to employ to have caught this problem ahead of time?” Every lesson learned is a chance to tune your cognitive strategies to prevent as wide a class of similar problems as possible in the future.

At very least, learn to recognize unproductive over-correction and to drive past learning stopsigns. When you encounter a failure and make a snap judgment about what went wrong, ask yourself: is it any less likely I’ll fail in the same way again?

Exercise: Set a Yoda Timer and meditate on your most recent mistakes.

Daily Challenge

Share a story of a cure that was worse than the disease.

This is part 28 of 30 in the Hammertime Sequence. Click here for the intro.

The last three days of Hammertime, I’ll wrap up with some scattered thoughts to reinforce important principles.

Today, I’ll return to applications of reductionism to instrumental rationality.

Read the rest of this entry »

This is part 27 of 30 in the Hammertime Sequence. Click here for the intro.

Focusing is a tool for accessing the messages the many sub-personalities in your subconscious are trying to send you. What happens when two or more of these messages are in conflict with each other?

Internal Double Crux (IDC) is CFAR’s answer to this problem. Roughly speaking, it’s a script for taking turns Focusing on two conflicting inner voices and holding space for them to debate and compromise. A sort of internal couples therapy, if you will.

Read the rest of this entry »

This is part 26 of 30 in the Hammertime Sequence. Click here for the intro.

满罐子水不响，半罐子水响叮当
The full can is silent, but the half-empty can makes a loud noise.
~ Chinese proverb.

Take a bottle or soda can and fill it halfway with water. Shake the can – the water will slosh around loudly.

Now, fill the can to the brim and shake it again. It’s almost completely silent.

This is an essay about inner silence – calming one’s loudest inner voices to allow quieter voices to speak. Usually, the quieter ones have urgent messages, especially given how long they’ve been neglected.

This post is, in some sense, a followup to Babble.

Read the rest of this entry »

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.

Read the rest of this entry »

This is part 24 of 30 in the Hammertime Sequence. Click here for the intro.

Intentions are momentary, but problems last forever.

A human being’s attention flits around like the Roman God Mercury, root of the word “mercurial” – subject to sudden or unpredictable changes of mood or mind. The biggest problems in life require concentrated effort over years or decades, but you can only muster the willpower to even intend to solve a problem for minutes or hours. Worse, you can pretty much only maintain one intention at a time.

How do we make intentions count?

The philosophy of Design is: build intentions into external reality. Like your problems, external reality also lasts forever.

Read the rest of this entry »

This is part 23 of 30 in the Hammertime Sequence. Click here for the intro.

In school, we spend thousands of hours learning about the building blocks of the universe. We learn that reality reduces into little pieces: organisms into cells, books into pages, skyscrapers into atoms.

Your life belongs inside this infinitely divisible reality. Your psyche divides into subpersonalities, emotions into qualia, actions into goals and aversions, habits into TAPs. In fact, what we think of as objects are usually patterns of interaction between many tiny pieces.

Day 23: Reductionism

Trigger-action plans are the building blocks of habits – all habits can be built out of single steps.

I want to share a model for why it’s so important to break actions down with reductionism.

Zeno’s Paradox Retold

Here’s the old paradox of Zeno:

To win a race, you have to run the first half. Before you finish the first half, you must complete the first quarter. Before you finish the first quarter, there’s the first eighth, and so on ad infinitum. Thus, by halving the first segment, every race is divisible into infinitely parts, and to complete the race you must make infinitely many actions.

What can we learn from Zeno’s paradox?

Of the infinitely many steps in the race, the first step accomplishes almost all of them. It follows that the first step in a race is infinitely more difficult than every later one.

The Method of Exhaustion

From Zeno’s Paradox, we readily derive the following algorithm for deconstructing problems:

1. Pick an action.
2. Divide it into halves. Focus on the first half.
3. Repeat to exhaustion.

For example, I can decompose the action of “write a blog post” in exponentially ascending order of difficulty:

1. Take a deep breath.
2. Visualize success.
3. Turn on computer.
4. Open Chrome.
5. Log in.
6. Type a letter.
7. Type a word.
8. Type a sentence.
9. Type a paragraph.
10. Type a section.
11. Type a post.
12. Click “publish.”

After having completed the method of exhaustion, executing the action is much easier. Notice that even though I’m ostensibly only 1/3 of the way through writing this post, I’ve already accomplished 10.5/12 steps in the workflow.

I’m almost done!

Steps of Equal Difficulty

You may think the last section was flippant or self-delusion.

Nope.

I’m completely serious.

Walk through the whole activity of blogging (if blogging’s not aversive to you, pick whatever else you’re procrastinating on and apply the method of exhaustion to that one instead), and note how much total mental resistance you push through at each step in the 12-step process. Also note how likely you are to give up at each step.

The normal method of planning is to break into equally sized blocks, where size means “time and effort in objective reality.” Take stock of all the plans you’ve made in your life. How many failed at the very beginning? How many failed near the middle? How many failed towards the very end?

Most things fail before they begin. Of the ones that do begin, most fail immediately.

You don’t live in objective reality. You live in the mad world of Zeno, where the first step is infinitely difficult. The Method of Exhaustion is designed to parse a hard problem into steps of roughly equal psychological difficulty and failure rate.

Exercise: Apply the Method of Exhaustion to your next big project. How many pieces did you break it into?

Daily Challenge

Share anecdotes or data on how long it takes [intentions, projects, plans, relationships, careers, startups] to fail. What do the curves look like?

This is part 22 of 30 in the Hammertime Sequence. Click here for the intro.

At some point around the end of high school, being fast became unfashionable. When did this happen?

Why do we channel so much more energy into doing more difficult things, instead of doing simple things faster? How much faster could you do your job? Two times faster? Five times?

Instead of I want to be stronger, say I want to be faster.

If you pay attention to speed, you might just find a way to do a whole week’s worth of work in five minutes.

Read the rest of this entry »

This is part 21 of 30 in the Hammertime Sequence. Click here for the intro.

I took a long break from Hammertime to check the fundamental question: am I actually better at achieving my values now?

Read the rest of this entry »