The art of dividing people

This is a follow-up to the tricky questions discussion.

The recipe seems to be to construct a no-brainer question that prompts an intuitive System 1 "yes" response. For example, Do you prefer $1 to $1,000,000? or Why do we need people from shithole countries? or Do we need to deport illegal immigrants? etc. The question can be easily mapped to object 2 {0,1}.



The tricky part comes from constructing a question so that it it maps to a 2x2 object (rather than 2) with skewed values: A = {0,1}; B = {1,0} -- note that objects {0,0} and {1,1,} are missing. Therefore, by choosing A or B we are always "right" in one dimension and "wrong" in another dimension. Understanding this portion requires System 2 thinking.

Example 1. Do you prefer state {$1, $2} to state {$1, $1,000,000}? The no-brainer System 1 dimension is, e.g. the total welfare of the state or Pareto improvement from a transition from the first to the second state. The System 2 (slow thinking) dimension is extreme inequality. The object is constructed so that a preference for welfare requires a preference for extreme inequality {0,1} or vice versa: a preference for socially acceptable inequality requires a preference for poverty {1,0}.

Example 2, Do we need more people from shithole countries or those from Norway? The no-brainer System 1 dimension is quality of the country, which is obviously bad vs good, i.e {0,1} The System 2 dimension is the quality of an individual from that country, which American history shows is often {1,0}. Being "right" in the first dimension reveals System 1 bias.

Example 3, Should we deport illegal immigrants?

As a communication medium, Twitter conditions the public to fast "no-brainer" System 1 responses.

I need to show this formally with proper arrows and objects.

What it takes to be a stable genius.

http://www.bbc.com/news/world-us-canada-42717163

US President Donald Trump has shown no abnormal signs following a cognitive exam and is in excellent health, his White House doctor says. But what sort of questions was he asked? See if you can answer them all.

I answered all 5 questions correctly.

(no subject)

Probably we need to formally introduce concepts of cognitive System 1 and System 2 preferences, e.g.

A & B are states of welfare. A = {1,2}; B = {1, 10000000}

f: A -> B - System 1 preference. ( A1+A2 << B1+B2) - one-step evaluation -> better-off in welfare, "true".

g: A -> A; endomapping for A ( Amax/Amin)
h: B -> B; endomapping for B (Bmax/ Bmin)

j: g -> h ( Amax/Amin << Bmax/Bmin) - one-step evaluation -> better off in inequality, "true".

k = j f - System 2 preference - two-step evaluation after the discovery of the 2nd dimension.

Reinventing the wheel

It seems like I've been trying to reinvent the wheel with my "tricky exercises." Basically, all I needed to do was to count 2-valued functions.

The preceding principle will have a great many uses. It enables us, for example, to count parts by counting 2-valued functions instead.

-- Lawvere, Rosebrugh. 2006.

That is, having two "no-brainer" 2-valued functions tells us that there are two parts in the set.

upd. The tricky aspect is restricting the set of functions to two by eliminating {0,0} and {1,1} choices.

upd1. Any two-dimensional choice space is prone to tricky either-or questions. E.g. System 1 vs System 2, short- vs long-term, шашечки или ехать, quality vs price, etc. Under the circumstances, we should start by expanding the choices by expanding from product AxB to disjoint union A|_|B.

(no subject)

Чему реально надо учиться у хороших инженеров/программистов - это желанию и способности делать умное, а не говорить умное. Умный разговор, как правило, получается по итогам сделанного.