Mission

Mission #2: Collect actual, everyday instances where you fell short because of lack of “critical thinking” (why am I putting “critical thinking” in scare quotes?) and divide them into skills. No abstract skills like “logical structure” or “critical thinking”. Concrete instances where you failed. That will give you harsh feedback. If you stop failing, it means your deliberate practice is working. If not, it means it’s not working, even if you think you’re doing great. NINRTT. Then, come up with rough deliberate practice plans for the skills based on the three models.

For example, one skill could be “not accepting a claim without empirical evidence”. Concrete situations where you may have failed in this could be “I didn’t spend much this month” (but you didn’t even look at the actual debits on your bank account) or “the gym will be open till 9pm” (and you go there and find out it closed at 5pm because of Fall break, which you could have looked up on the website). So, the obvious feedback is that you don’t make such mistakes because of lack of empiricism. Another quicker form of feedback is perhaps a daily counter for the number of empirical sources you’ve looked up. (Notice how hard it is to get exact feedback for bad thinking? It happens without us even realizing. Getting good feedback is a problem that needs to be solved.)

10 skills. 10 concrete instances for each + DP detailed plan. 3k words. 1 week.

What do I need to do

  • Write several instances where your thinking failed you

  • Identify the skills from those instances.

  • Provide a point-by-point training session design. Just like your workout schedule where you know the entire plan for the week, month even.

  • We need to design our own practice sessions as these are not well established techniques like in sports to improve performance.

  • There are 2 key aspects :

    • doing something you couldn’t do earlier

      Like say gaining the ability to do Data Science or critical thinking like never before.

    • and then doing it on demand

      Knowing the scientific method and applying it as and when you encounter a problem, to understand or predict things.

Where has your thinking failed you?

Not accepting claims without evidence

Today I came across a video, and I have been seeing videos like this on Facebook. A song runs in the background, a very meaningful song, that always brings me to tears. This song is backed by beautiful montages of this guy bowing down, folding his hands (being humble), helping people and so on. All these indicating his loyalty to a certain x-chief minister. I sort of went on with it, meanwhile creating beliefs that he must be a very nice humble guy. All of this based on a video? It was pointed to me, later that these are sponsored. And (only) then I woke up from my trans. You won’t believe it, even as I am writing this article on the importance of empirical evidence, I went with the flow and started believing the image of the person it was casting. Let me off the hook, what if the millions of people watching this, fall for this? And end up voting for someone whom they believe to be of certain characteristics (based on a stupid video)? — from a post in 2016

My friend was cooking on a teflon pan which had much of its teflon removed over usage. And he said, “eating teflon is not injurious to health”. Without any hesitation I have been eating that food for 2 years now. In another instance my friend said, “slightly burnt food is carcinogenic”. I ran away from “carcinogenic food” ever since. I didn’t even check it. I didn’t need to?

I stopped wearing earplugs because I was afraid they would cause ear infection and suffered a plethora of distraction. The difference is quite a lot with and without this device.

I drink protein shake twice a day for a while now. I have just accepted that it is ok!

Drinking beer every week to the exact amount I drink, should cause no harm to my health over the long term. Pfff! It gets off your system in the form of piss right! The point is, given the opportunity, I wouldn’t hesitate to drink often. In India last year, I drank 4-5 times in a week. Hmmm, so what is good for you again? I accept implicitly that as long as I am not a drunkard (someone who has to drink everyday) — I consider my self a social drinker — I will be in exceptional health. Well!

I will find a house in the next 10 days if I put all my effort in it. So far it has been a long long while since I started and it is very hard to find a house. Claims without EE much?

I will find a job in less than 3 months once my company goes bankrupt and that time will be peaceful. The last time I found a job it took me a few months of practice followed by 3 months to land a job. With experience it might be easier (which is the case now), but considering the process times taken by companies, it could be a real big nightmare.

I believed that global warming was taking place as a result of human interaction. Today it was pointed to me that it was something that I took for granted. My takeaway is that I need to get back to the facts before assuming certain things. — from a post I wrote in 2016

Impact of lack of such a skill

In all of the above cases, it appears that it is quite important to know the truth as I care extremely much about my health and my long-term happiness. Researching before accepting claims as evidence, could benefit me quite a bit.

So how do I improve this?

‘Not accepting a claim without empirical evidence (EE)’, sounds simple. In any case it looks like 1. it requires training to identify such claims, and 2. the training to look up or determine the acceptable EE.

Training to Identify claims of EE and non-EE

Improving the first part, ‘identifying claims that have no empirical evidence’, seems like it is similar to the music model, i.e., training based on high focused repetition of task. So how would the DP of this look like?

I could start by saying I will be conscious and collect 100 ‘claims without EE’, in day to day conversations by the end of the week. The idea I think is to train the brain to notice the claims being made and to separate them as ones with evidence and ones without. I reckon it will be mind numbingly hard to stay focused on every conversation and every thought in an attempt to meet the goal of 100 claims. Hmmm, so it is going to be painful, in large numbers and your focus is probably going to be only about this. That seems about right. This way we can initiate the training process.

Another method could be to look at a source, a text or anything, and determine the claims without and with EE. So this source could be everything from the news, news comedies, political discourses, opinion speeches, articles on life, religious preachings, my own writings, writing of rationalists etc… I would like to keep the source spanning various fields so that I can train for different ratios of claims with EE and without EE. For example, Religious preachings/debates, flat-earth discussions will give you a taste of predominantly claims without EE, and writings of Eliezer Yudkowsky, Paul Graham and STM could give you a taste of claims with EE. Differentiating all of them, seems like it will help and would allow practicing in large volume. In addition, I could as an everyday task, identify 10 claims made during the lunch conversations and 5 claims made during the rest of the day during interactions with colleagues.

But what about feedback?

I use a log to note how many hours per day I spend on writing and other activities. This serves me an indication over time, if I am getting better, how my week was and if some days didn’t go well and why. The same could be done for every skill including this I guess. Maintaining a log of daily claims identified with and without EE over time gives us an indication of where we might be.

Engaging in questioning and evaluating if something has EE or not could in itself be instructive. For both the above proposed methods, following it up with a blog post to explain why something has EE and something doesn’t have EE, will help. This can then be used to get feedback from external readers once in a while. This feedback will tell you if the direction you are steering in is right or wrong and where you need focus.

What if there are some types of claims that you don’t notice?

There is a chance that I don’t notice some of the ‘claims without evidence’, in that case it would be nice to have external feedback from someone who is better than you once in a while and focus training to identify exactly that.

Training to determine EE

I would expect from me to determine EE for 1k claims. Obtaining claims shall be easy. I can take them from my own life or the sources suggested above. Determining EE for different types of claims would be different. If the claim is like ‘eating teflon is good’, then I would look at sources like stackexchange to give me info. If the claim was that ‘a temperature rise of 1 degree is expected if current human pollution doesn’t decrease’, I would spend time reading a couple of articles that refer journals (and look into those citations and check if the articles are credible) to get EE.

I guess the way to go about it is to gather all the claims to be investigated (atleast a 100 of them), classify them roughly on the basis of the expected method of finding EE. And then take each class and practice. This should be done on 1000’s of claims atleast. Finally a post on the different methods of finding EE with concrete examples, identifying what counts as EE.

Feedback

The number of claims that I should theoretically come across should be very similar with every week. So, if I record the number of claims that I encountered per day, the number of claims I looked for EE and the number of claims I successfully found EE for, then over a week I have statistics. And logging this over time, I could get some useful feedback about my progress, and level.

The intense study of these things could in themselves be instructive. Of course like always, when in doubt or when you would like feedback in certain aspects, it is best to write a post and ask some one better than you to evaluate and give feedback and take it from there.

Problems

There seem to be some potential pitfalls in this approach. I expect most of the claims to be vague, in the first place, let alone with or without EE. By vague I mean, the claim might not be verifiable in the first place. I think I will start with this anyway and optimize as necessary.

Ability to communicate what is in the head in a systematic way

Often there is lack of clarity in my head and hence I am unable to communicate well with people around me. I often start at a point and then realize I forgot to tell something else and start over again. This happens often in my work place, when in general I am trying explain something that I have not much understanding of. Better understanding helps, but in this case it looks like delivering the information in a systematic way is the real problem.

I also notice the “same” thing in writing I think. I write all over the place and then work on re-arranging things over multiple iterations. It is painful to write like this where I don’t have clarity of what happened until now.

I was asked to deal with a trouble ticket which I had barely any experience with. I had a look at it and had come back to my DL to explain to him what the deal was and as expected I stammered, I stuttered. I was slightly all over the place. I was not clear in my head, atleast I couldn’t see the flow of thoughts. In the end I managed, owing to making notes in a notebook (observation, hypothesis, tests)

This is almost everytime. Whenever I have to discuss anything with my Design Leader/Boss (DL), I struggle to get shit out of my mouth in an organized manner.

How to DP

During interviews I have found the STAR method (situation, Task, Action, Result) to allow me to deliver my thoughts in an organized manner. This I used to practice on everything that I was going to deliver and managed to not stutter. Deliberately practicing this can have a pronounced effect on my speech and maybe writing even, I suspect. Maybe STAR might not be appropriate for delivering in the above cases. I would first like to deliver the Problem, the observation, followed by a hypothesis, test and results if applicable i.e., POHTR (like Potter).

So the goal is to use this framework while communicating, to allow for clarity. Improving this skill is very similar to the music method. So practicing explaining using the POHTR method for a 1000 problems is something that I should aim for. Where will I get so many topics to talk about? Well, I could take sources from the news, news comedies, TED talks, discussions on LessWrong and even the HPMOR work by Eliezer Yudkowsky. Hopefully summarizing these topics will in addition to this skill give me more understanding of the topics dealt with. The same exercise can be used for both writing and speech.

As for feedback, obvious feedback is you apply the above techniques in situations in real life and see if you stammer or stutter. For immediate feedback, I think I can re-read my essays, or play my explanation-voice-recordings and determine how good or bad the organization of thoughts was. Maybe other aspects to improve turn up. An STM can give me feedback when he has time. For more feedback you can ask your friends to listen to your discourse on youtube or your blog, to give feedback of how much they understood or what you need to improve on. A counter similar to the ones proposed in the earlier section can be used as well to obtain statistics.

Ability to solve unknown problem close to domain?

At work I often get problems that I have no clue about. And it is my task to work on solving them by asking the right people for help with knowledge, going over the observations, proposing a hypothesis and a testing, to diagnose the problem.

Had a meeting with my boss and another colleague (same age,background everything) this week, to discuss an issue and how to solve it. I sucked, the colleague who was into these kinds of trouble shooting didn’t suck at all and my boss was on top of everything. I was surprised my colleague was this good, in the sense that he was able to propose hypothesis and quickly look at pros and cons with a neat focus on the goal. Whereas I had a bowl of thoughts being tossed inside my head, struggling to make any reasonable contribution.

So looking at a lot of information all over the place, sometimes an overwhelming amount even, throws me off. I would like a way to focus on what is important (and why it is important) and improve my skills in solving unknown problems like these, i.e., be much greater than my colleagues and DL.

How to DP

Coming to the applications: Much of “critical thinking”, I would argue, falls under the chess model. The music model applies when you know what you want to say and it’s about how well you say it. Whereas here, it’s often a question of coming up with the right response to a situation. So, one practice technique would be to take case studies, ask yourself what move you’d make, and compare it to that of the experts. — an STM

Luckily for me there is a plethora of trouble tickets (TTs) at my work. I could say spend some time after work everyday, with the focus of training my problem solving capability. Here domain knowledge might be quite important, but as I said, I always worked on unknown problems (atleast that’s how my DL trains me to handle new modules) — it’s also a way to learn about the system. So to keep it less overwhelming I am going to look at TTs that are slightly in my domain. I should take a look at few entries in the TT and propose actions, compare with the actual entry and see just like in the chess model. We then move on to the next entry.

My DL constantly works on solving some kickass problems for the company. He gives his team the lame work and sometimes the nice challenging work as well. I could look at the problems he has solved. Not surprisingly I see that he uses the scientific method. There is a clear statement of observations, followed by some hypothesis that made it and some hypothesis that already failed, with some clear tests. I could look at the problems he has solved and do the same exercise of predicting what I would do.

Real world as it is makes it hard to evaluate which of the proposed solutions is better. In any case the process of focusing on the problem and evaluating solutions is in itself instructive.

What about feedback?

The obvious feedback is that you would be solving problems without much intervention even when it is an unknown error. The feedback while working on DP looks like it is built-in within the problem solving framework. I could collect some statistics such as responses which are correct over a week against the total responses made, time spent on solving and the like, to allow me to know how I am doing.

Solving a new problem (in life)

Since last year September (2017), I have been doing cardio 3-4 times a week on average in addition to playing sports (1-2 times a week). I started with riding the bike and an STM introduced me to the equation of motivation. This meant that I get feedback of some sort, to know how good or bad my workout is at a given instant. So I got a speedometer for my bike. Well that was not really representative as winds happened! So I started doing cardio at the gym (no winds). I was looking at info such as current RPM and calories for feedback. But it was not enough (i.e., doing cardio was painful mentally). I wanted something more! I wanted something that would tell me about the average workout rate during the workout. So I got a heartrate meter (HRM). It pretty much does that, but now I want some finer statistics, I would like to race against my own ghost even. What after I get the exact statistics I want! Am I claiming to be able to do cardio on all days? What do I want and what am I going after? Or will cardio always be a strain mentally? In other words did I even understand the motivation equation? Would thinking about body fat percentages help (I remember an STM mentioning it)?

Upon the last measure I was at the same body fat % and muscle mass as a year ago. It seems unclear how to go to 10% body fat despite doing certain things. So obviously my ‘fitness-model’ is wrong.

My RSI is cured. I also used the same ideas to work on my back and knees (in terms of impact I deadlift and squat my max now), but I have pain in some other areas occurring randomly. In order to investigate if it is RSI-like, I would need a model that makes predictions? What is this model and how do I make predictions?

The last one week I would come to the library and start working on this essay. As soon as I start work I would feel sleepy. Somedays I just couldn’t resist and took a nap (30-45 mins). Then I would continue working. At about 1-1.5 hrs in I feel the flow catching on. What model can I use to determine why this is happening? How can I make it feel like working on programming?

Skills

So, this appears to be like the last section on ‘solving unknown problems close to domain’, but I am afraid working on those plethora of trouble tickets, might not help in cases listed in this section.

So it looks like there are a couple of skills at play here. Making accurate mental-models of the domain based on observations, making narrow predictions and/or problem solving.

How to DP

To develop the above listed skills which seem to be interconnected, it appears that the chess model would fit it, i.e., we would need to solve a plethora of case studies and compare our performance to that of the case study. But where do we go for such cases studies?

to be continued…

Final words

This essay was really hard. As mentioned earlier, I wanted to sleep as soon as I started on most days. But after like an hour of work, it was alright. Generating skills off of life instances was quite a challenge. Maybe if I see a few more examples it would help.

Today I realized how vague some of my DP for skills were, such as using criticalthinking.org. And ended up in quite some re-writing. I am not claiming my current DP for skills is like how I know my work-out would be. Long way to go!

Most of the cases I came up with, seemed to end up with the skill of ‘problem solving’, such as cases of procrastination, failed motivation, ‘lack of application of theory to practice’ etc… I am unsure how to deal with this skill on a DP basis. Where to source example problems to be solved from (just like in the chess method). And how to make this look like the chess method, aka get feedback after every move.

Other things I was trying to explore was ‘thinking aware of biases’, as I saw a recurring theme of allias paradoxes in the instances of my life. But didn’t have time to work on it.

I didn’t deal at all with the aspects of doing something you couldn’t do earlier, and then doing it on demand.

Overall I was not able to do 10 skills. I did something like 4 maybe! Sorry!

P.S

Spent on average 3.7 useful hours per day for the last 7 days. ~26 hrs.