Thread by @evolvingcalm

For the past month or so I've been taking @normonics ACS101 which I can thoroughly recommend.

Seeing as we're about half the way through #ACS101 I thought I give a recap of my notes of some of the highlights so far.

/thread

I've being going back through the sessions so far of ACS101 #SpringA2021 in my spare time.

It's remarkable how much more everything clicked that much more the second time around after the time absorbing everything.

1/n

Truly no book is read twice by the same man.

2/n

Preliminaries & Scene Setting:

(I must admit I remember feeling nerves on the first day. Hearing myself speak also made me immediately redo my entire audio/visual set up).

3/n

There are commonly held hidden (reductionist, mechanistic) assumptions that need to be unlearned.

These permeate how we think the world works.

4/n

Once you have a sensitivity to these assumptions you will see them everywhere (true) and how they corrupt our thinking.

5/n

Unlearning is hard because we stuff everything into boxes until reality ensues and we have a set of experiences that is painful enough to break them down.

6/n

You can build intuition through soaking in the set of interrelated topics with the same theme in different contexts.

7/n

Models cannot capture systems in their full richness.

It's not reasonable to use a model to make a naive prediction about complex systems.

8/n

No single viewpoint yields a complete picture.

You must vary parameters of scope and resolution to even try.

9/n

A blurry big picture is often better than a crisp micro-picture.

10/n

Deep enough but no more.

11/n

A sensitivity to complexity leads to asking different questions and getting different answers.

12/n

Joe likes being interrupted.

13/n

Commonly unstated assumptions are insufficient they provide models that breakdown & can't handle basic properties we see in the world.

14/n

Models can be used in an allegorical sense to grasp archetypes of properties/relationships in systems.

15/n

We talk about what our favourite colours are.

16/n

'Any dead thing can flow with river. Only a living thing can swim against it'

- G.K Chesterton

17/n

Session One: Reductionisms & Emergence

This (xkcd 435) is wrong.

18/n

Moving through the hierarchy of science or the different scales of systems introduces new properties, behaviours, laws as you go.

(Consider elements to chemicals to biology)

19/n

You can't just extrapolate the rules of chemistry to get biology.

Biology contains things you can't get just by simply studying chemicals or chemistry.

20/n

We parse them perceptually & intuitively - but what is a system?

Generally, it's drawing a boundary around a part of ambient environment and calling it a system in whatever way it matters/makes sense to us (perspective).

21/n

There is an interchange/exchange with systems and the environment of inputs and outputs.

22/n

We often try to insulate & isolate systems to study them without reference to what is happening in the environment but we can destroy properties the system by doing so.

23/n

A lot of science is based on this contrived assumption of being able to isolate a system.

24/n

What makes a system complex?
When the interaction of the parts is essential to understanding the properties and behaviours of the system rather than the parts being examinable in isolation (i.e reductionism).

25/n

Complexity is the science of Humpty Dumpy

A non-trivial integrated whole.

26/n

Reductionism is an all-too-common assumption about how the world works (things are the sum of their parts).

27/n

The idea you can understand the world by decomposing it into parts, study the properties of parts in isolation, putting parts back together into the big picture is trivial, and thinking everything can be understood this way.

28/n

Scope and resolution are useful concepts to grok complex systems and can vary independently.

(Theoretically anyway as they often converge practically, but not necessarily)

They are independent orthogonal ways to adjust our lens on a system.

29/n

The sidedness of a mobius strip is a good example of emergence.

30/n

No one piece of the strip that you remove won't be two-sided and double-edged.

But together in their entirety one sided & one edged.

31/n

The details you lose by lowering the resolution might be important.

32/n

An agent is always bandwidth limited.

33/n

Scope is limited by the bandwidth of the observer.

34/n

There are fewer unique pieces to account for the lower the resolution.

Resolution is a function of the observer.

35/n

Weakly emergent properties are observer/resolution dependent.

Strongly emergent properties are based on scope.

36/n

Can you see the forest for the trees?

https://twitter.com/evolvingcalm/status/1365434110395572227?s=20