Hey LLM, why do I enjoy playing games better than doing homework?

Hi friends,

I’ve been looking at this one chart from this report by Our World in Data about time use.

I’m sure you’ve seen different sections and different charts out of this report being shared online since this was released a while ago. But my mind keeps going back to this one chart in particular:

Obviously, like you, these thoughts popped into my head: “who are they surveying? how statistically significant is this and for which demography?”, “what are we not seeing here?”, “well ofc leisurely activities are enjoyable, water is wet”, “but what are the underlying themes in the low enjoyment activities?”, “how can we make some of these activities more enjoyable?”, “what the is professional services?”.

Well, we’re not going to scrutinise over all those questions because life is too short for you to be reading my brain dump and nitpicking.

So today we’ll focus on: what are some good1 explanation and key patterns found in these three buckets of enjoyment level?

I am going to show how we can use LLM to help suggest frameworks from different fields that can help us explain and make sense of why these activities are where they are in the ranks.

If you’re pressed on time, jump to the Closing Thoughts for the mind-tickly stuff.


First, give me some frameworks!

I asked GPT-4o:

here are a list of activities that people rate as having a {low/medium/high} enjoyment rating. for each level of enjoyment rating, what are some frameworks (in any field) that can be used to evaluate these characteristics?

It suggested a total of 13 frameworks2, but we will just focus on these two as I think is the most interesting and useful to give a taste of how LLMs are amazing at triggering verbal alchemy in us. I’ve included the full list in the footnotes.

  1. Cognitive Load Theory (Intrinsic Load, Extraneous Load, Germane Load)

  2. Transactional Analysis (Parent-Adult-Child Model, and Scripts and Games)


A. Parent-Adult-Child Model

The Parent-Adult-Child (PAC) Model from Transactional Analysis (TA) posits that people operate from three distinct ego states: Parent, Adult, and Child.

  1. Parent Ego State: characterized by behaviours, thoughts, and feelings copied from parental figures. It includes nurturing and controlling aspects.

  2. Adult Ego State: involves objective, rational, and data-driven thinking. It focuses on problem-solving and dealing with reality in a balanced way.

  3. Child Ego State: reflects our internal responses and feelings from childhood. It includes natural, adapted, and rebellious aspects.

Analysis

I asked GPT-4o:

could you apply the Parent-Adult-Child (PAC) Model against each enjoyment level, and use a couple of activities in each level to illustrate the point?

Here’s an edited excerpt of its response:

By applying the PAC model, we can see that high enjoyment activities often engage the Free Child state, medium enjoyment activities blend the Nurturing Parent and Adult states, and low enjoyment activities are dominated by the Critical Parent state with necessary Adult involvement.

  • High Enjoyment Activities

    • Child State: Engaging activities like visiting friends and playing sports often activate the Free Child state, resulting in high enjoyment due to the spontaneity, joy, and playfulness they evoke.

    • Adult State: These activities can also involve the Adult state, providing a balanced blend of logical engagement and carefree enjoyment.

  • Medium Enjoyment Activities

    • Parent State: Activities like gardening and voluntary work engage the Nurturing Parent state, providing a sense of care and contribution.

    • Adult State: The Adult state is involved in planning and executing tasks, offering a balanced sense of accomplishment and care, leading to medium enjoyment.

  • Low Enjoyment Activities

    • Parent State: Low enjoyment activities like cleaning and job searching often engage the Critical Parent state, associated with obligation, rules, and societal expectations.

    • Adult State: These activities also require the Adult state for logical execution, but the dominant Critical Parent state reduces their enjoyment.


B. Cognitive Load Theory

Cognitive Load Theory (CLT) helps us understand the mental effort required to perform different tasks. It divides cognitive load into three types:

  1. Intrinsic Load: The inherent difficulty of the task.

  2. Extraneous Load: The way the task is presented and any unnecessary complexity.

  3. Germane Load: The cognitive effort required to process information and create new knowledge or skills.

Analysis

I asked GPT-4o:

could you apply the Cognitive Load Theory against each enjoyment level, and use a couple of activities in each level to illustrate the point?

Here’s an edited excerpt of its response:

  1. High Enjoyment Activities

    • Low Intrinsic and Extraneous Load, High Germane Load: Activities like visiting friends and playing sports have low inherent difficulty and minimal distractions, allowing high cognitive engagement and enjoyment.
  2. Medium Enjoyment Activities

    • Moderate Intrinsic and Extraneous Load, Moderate Germane Load: Activities like gardening and voluntary work have moderate inherent difficulty and some complexity, but they provide meaningful cognitive engagement.
  3. Low Enjoyment Activities

    • High Intrinsic and Extraneous Load, Low Germane Load: Activities like cleaning house, job hunting, doing homework, and attending school are inherently difficult and complex, with many distractions and low cognitive engagement, leading to low enjoyment.

So what?

None of what is outlined here is surprising, I’m sure. We have merely stated the obvious, just like many seemingly “doh” scientific studies in social sciences3. But hey, having a formalised framework and labels for concepts is useful so we can point at it and build on top of it.

https://www.instagram.com/reel/C5V6zmzLKUo/

But most of all, I hope this example of my LLM usage give you ideas for doing more of these verbal alchemy on yourself.

I think these words engines are like knives. You have the power to trust it to think for you, act for you, do your homework for you — and other performative, hygienic, and political wordsmithing — but it can also help you discover secret entrypoints to accelerate your encounters with the world’s knowledge.

Your prompt is the boundary of your map, your critical thinking skill is your lane markers, and your curiosity is your fuel.

We need to wield our discernment, curiosity, and patience well.


Coda A: Offline chores are my gateway drug to DMN

I find it interesting that housework and domestic chores such as laundry, cleaning, doing the dishes are rated so low. I reckon it’s kinda fun. I don’t mind doing laundry. It’s an easy hack to a quick wins, to feel like I have gotten something done.

Well I suppose chores are mostly fun when it’s not something you have to do, but something you get to do — and I suppose a big factor in this — something you chose and initiate yourself.

Doing these chores also gives my eyes and brain a break from all the screens. I can also listen to music, podcasts, and let ideas emerge while ironing the clothes, mopping the floor…

Also reminded me of:


Coda B: Comments on the different activities

  • On looking for a job: It has high variance because it’s between the extreme of lining yourself up to being judged, evaluated, and selected (and most likely rejected). But it’s also enjoyable because it triggers the “hopeful” emotion — window shopping for an alternate life, daydreaming about exiting your current main job.

  • Adult Care also has a large variation in enjoyment rating probably because it’s both tedious and demanding but at the same time it’s also an act of service for someone presumably you care about — it often can easily be interpreted/justified as a meaningful act.

  • On Paid Work at Home: interesting. Does this mean working at your main job, but doing it from home?

  • Doing homework is low perhaps when 1) the student have no idea why the homework is necessary, beyond appeasing some authority / checking off some external objective set by others for them. And because 2) it demands cognitive effort, you need to engage, for reasons you don’t understand or aren’t fully on board with

  • RE:Games, I recently heard Thi Nguyen posited that games are enjoyable because it is most of the time a non-ambiguous low-stake environment with clear goals and path to completion. Am maraphrasing here but that’s the high level gist of it.

  • I feel affirmed to find multiple music-related activities so prominent in the High Level Enjoyment section. As Iain McGilchrist is often quoted saying, “music does not so much free time from temporality as bring out an aspect that is always present within time, its intersection with a moment which partakes of eternity“ or as George Steiner put it, “music is … time made free of temporality”.

  • But again. These are Brits they’ve surveyed. So who knows what we’re really seeing here.


1

Backed by some sort of (cargo-culted and probably not reproducible and confabulated/hallucinated whichever term you prefer) science. But that’s the point. It’s your cue to test these words against your gnosis.

2

  1. Cognitive Load Theory (Intrinsic Load, Extraneous Load, Germane Load)

  2. Self-Determination Theory (SDT) (Autonomy, Competence, Relatedness)

  3. Flow Theory (Challenge-Skill Balance, Clear Goals and Feedback, Complete Immersion)

  4. Maslow’s Hierarchy of Needs

  5. Transactional Analysis (Parent-Adult-Child Model, and Scripts and Games)

  6. Job Characteristics Model (Skill Variety, Task Identity and Task Significance, Autonomy and Feedback)

  7. Positive Psychology Frameworks (PERMA: Positive Emotions, Engagement, Relationships, Meaning, Accomplishment)

  8. Behavioural Economics (Prospect Theory, Nudge Theory)

  9. Herzberg’s Two-Factor Theory (Hygiene and Motivators)

  10. Compensation and Spillover Theory (Wilensky)

  11. Social Identity Theory (Tajfel and Turner)

  12. Time Management Theories (Eisenhower Matrix, Pareto Principle)

  13. Cost-Benefit Analysis in Economics (Utility & Satisfaction, and Opportunity Cost)

3

I’d love to be able to talk more about the problems with “hard” and “soft” science, scientism, anti-science-ism(??), and all the systemic issues around the way we do science (reproducibility, specialisation, incentives, etc), but I am not well informed enough on this yet, or ever will!

Leave a Reply

Your email address will not be published. Required fields are marked *