Mental Models: Scott Page

Get Started. It's Free
or sign up with your email address
Mental Models: Scott Page by Mind Map: Mental Models: Scott Page

1. What are mental models?

1.1. Map reality to the clean, logical structures of math

1.2. Make sense of complexity

1.2.1. By throwing ensembles of models

2. Why diversity matters

2.1. People who know different parts

2.1.1. And different knowledge overlaps

2.1.2. And different models of how things work

2.1.3. "Cognitive diversity"

2.2. Can bring in much deeper understanding

2.3. Give a more holistic approach

3. Distributions are important

3.1. Bell curves

3.1.1. Additive

3.1.2. Averaging

3.2. Stretched out bell curves (log-normal)

3.2.1. Multiplicative

3.3. Power laws

3.3.1. Preferential attachment Positive feedback e.g. cities

3.3.2. Random walk e.g. lifespan of species

3.3.3. Self-organized criticality e.g. Traffic in L.A.

3.3.4. If heights were distributed this way 10000 people as tall as giraffes 1 guy as tall as Burj tower 170 million in the US 7 inches tall

3.3.5. Why does it matter? There's going to be some people wildly successful And a whole bunch not so much Big winners are as much luck as they are skill

3.4. Understand each distribution

3.4.1. Logic

3.4.2. Structure

3.4.3. Function

4. Convexity and concavity

4.1. Convexity

4.1.1. Increasing returns

4.1.2. e.g. Preferential attachment

4.2. Concavity

4.2.1. e.g. adding workers in a team

4.3. In both these cases

4.3.1. Linear thinking is dangerous

4.3.2. Linear projections are dangerous

5. Wisdom hierarchy

5.1. Data

5.2. Information

5.2.1. How we structure data

5.3. Knowledge

5.3.1. Understanding pieces of information Correlative Causal

5.4. Wisdom

5.4.1. Which knowledges to bring on a particular problem

6. Individuals are limited

6.1. Even if you're the best you can

6.2. Even if you're a lifelong learner

6.3. You cannot beat collections of people

6.3.1. Having different ensembles of models

7. Know a bunch of models

7.1. Have an awareness

7.1.1. Don't have to be an expert

7.2. Know 20-30

7.2.1. But have a gestalt

7.3. Gives you robustness

7.3.1. Like a portfolio

7.3.2. Often much better than the average

7.4. Thinking tools

7.4.1. Force you to get the logic right

7.4.2. Think about what matters Driving people's behaviors How the behaviors interact How they aggregate How should people respond to that

8. Know your place in a complex world

8.1. Be a nimble thinker

8.1.1. Move across models

8.1.2. If you can't, you should go deep

8.2. World needs both kinds of people

8.2.1. Some are T

8.2.2. Some are ∏

8.3. Your success depends on you filling a niche

8.3.1. Could be connecting things

8.3.2. Pulling resources and ideas from different places

8.3.3. Niche could take all sorts of different forms

8.4. Finding your niche

8.4.1. You have to really love it You've got to love the practice of it

8.4.2. You've got to have some innate ability Example: some people are naturally good at dancing

8.4.3. You have to be able to in some sense connect those To something useful, meaningful You should convince yourself And others

8.5. Ask yourself

8.5.1. What are my capacities?

8.5.2. Am I someone who is able to learn things deeply?

8.5.3. Am I able to learn a lot of stuff?

8.5.4. Then think about a strategy For your human capital

9. Spatial vs Col. Blotto model

9.1. Spatial

9.1.1. There is an ideal point

9.1.2. Example: your ideal burrito

9.2. Col Blotto

9.2.1. Many fronts, or dimensions The more the better

9.2.2. Example: car buying More mpg More legroom Higher crash test scores

9.3. Example: failed a job interview

9.3.1. Spatial: I'm not ideal

9.3.2. Col Blotto: somebody beat me on a collection of fronts

9.3.3. There's no best thing you can do Unless you know where the other person was It may be that someone else was positioned better

9.3.4. Strategy Spatial Do I have the characteristics they're looking for? Hedonic Beat my competitors on as many things as possible

10. Which models to prioritize and why?

10.1. Who are the relevant actors?

10.1.1. Single actor e.g. buying a book

10.1.2. Multiple actors e.g. investing decision

10.2. How rational is the person making the decision?

10.2.1. Repeated a lot: could be rational Because people learn

10.2.2. Big decision by an org could be rational

10.2.3. vs. rule of thumb "Standard operating procedure" within a large org

11. Some important models

11.1. Markov models

11.1.1. A set of states Transition probabilities across them System goes into a unique equilibrium

11.1.2. Means History doesn't matter One-time interventions don't matter There's a vortex drawing it into one thing

11.1.3. How to fix Create a new state Or fundamentally change transition probabilities

11.2. Linear models

11.2.1. If it's not working well Talk to people Usually it's a missing feature Put it back into the model

11.3. Colonel Blotto models

11.4. Lyapunov functions

11.5. Systems dynamic models

11.6. Simple signaling models

11.7. Local interaction models

11.7.1. It doesn't matter what you do What matters is you do the same thing that other people do

11.7.2. e.g. handshake

11.7.3. A culture is often a collection of such models Not genetics Not "character"

12. A meta-model

12.1. Perspective-taking

12.1.1. What does the problem look like through the lens of this person?

12.1.2. However, it's not like literature There's no end game in literature Every perspective is important in literature If you're taking a decision There's an end game You only want to be inclusive to things that are actually going to help you do

13. Collective Intelligence

13.1. Experiment

13.1.1. Physicists

13.1.2. Hayekians (decentralized market)

13.1.3. Bees (do the equivalent of waggle dance)

13.2. On an easy problem and a problem with five peaks

13.2.1. the bees did just as well as physicists

13.2.2. bees can take a derivative!

13.3. When it's super hard the market is not going to work

13.3.1. You need coordination

14. AMA