Hacker News new | past | comments | ask | show | jobs | submit login
The Contagion of Concern: Game theory of how anxiety spreads (adamjuliangoldstein.com)
137 points by goldfish 35 days ago | hide | past | favorite | 13 comments

Hi all, this is a follow-up to two essays:

1) The parallels between anxious ideas and immune system threat prediction (discussed at https://news.ycombinator.com/item?id=22475370)

2) The game theory of determining which imagined threats are worth our attention (discussed at https://news.ycombinator.com/item?id=22760540).

Thoughts and feedback welcome as always :)

I wonder if the model here is missing something important - it assumes there is no survival penalty for a false positive attack. In real life this isn’t true. Activating a fight or flight response in humans and animals has an energy (survival) penalty.

It does include a survival penalty for a false attack. I laid out the assumed payoff matrix here: https://www.adamjuliangoldstein.com/blog/paranoia-parameter/.... The penalty for false positives is also baked into the simulations (see first few lines of code here: https://github.com/adamjuliangoldstein/anxiety_algorithms/bl...).

You are right. It would work well if the models included a utility function and, instead of measuring likelihood of survival, the measurement were expected utility.

I’ve always thought anxiety is simply from different patterns of thought than people without anxiety. We can all share thoughts and fears but people with anxiety have a pattern that results in an unpleasant experience. Maybe one suffering anxiety should analyze different approaches to how one arrives at an outlook and compare to their own for maybe adapting. I’m writing with ignorance because I’ve never had bad anxiety.

From what I've read, and from my personal experience, is that anxiety comes from negative, inward-looking and expansive understanding of events. That is, anxious people tend to assume the worst, make things more personal (maybe it happened because of me or some innate human characteristic) and let the negativity leak over into other parts of their lives (I failed at this, I'll probably fail at something unrelated).

Conversely, people who remain positive and can isolate negative news to external, transient events are less anxious and end up healthier and happier. That might be contrary to what is actually happening; evidence shows realists tend to be pessimists.

Well, it's alright to be a pessimist. Not everyone has to be an optimist and wrong to be happy. I consider myself a pessimist and I don't suffer from anxiety. So maybe the assumption is somewhat true.

Yeah, I suppose you can be a realist or pessimist to a degree and live without anxiety. I think a lot comes down to how you attribute negativity. Take the riots going on at the moment.

An extreme optimist might say they're just another bad-luck event from 2020 (transience) that'll be sure to cause societal reform (positive outlook) and it doesn't have anything to do with their behaviour (externalising).

A realist on the other hand would recognise that the riots are a sign of a deep rooted problem with American society, that the riots may not lead to immediate reform but have the possibility to act as catalyst for changing some things that are wrong with society. Depending on who the person is, they may recognise they could be doing something to help with the problem.

An anxious person would take the negative aspects of a realist's position and make them personal, important and extrapolate out in time and domain. They might think the riots highlight how deeply racist humans are by nature (personalising & extrapolating), that the riots will cause economic instability or violence that will affect the person's life (extrapolating and importance), or that there's really nothing this person can do to change anything (generally feeling helpless).

This an excellent short explanation for group decision dynamics. If this subject is of interest to you I strongly recommend: How We Decide https://www.amazon.com/How-We-Decide-Jonah-Lehrer/dp/0547247...

This subject, at least to me, calls into question the behavior associated with recent events. The subjects of police brutality, civil rights, systemic racism, and so forth are much older than two weeks. I appreciate there are people and groups who have been actively working for years to raise awareness of these concerns, and then there is everybody else. Many people in that second category are crazy impassioned about these subjects as seen in both HN activity as well as the protesting and riots seen in the media. If this segment of people are honestly that concerned and these subjects were clearly concerns for more than two weeks ago then were these peoples’ level of concern then? I suspect this sudden call to concern is more the result of contagious group behavior dynamics and social psychology than a premeditated cognitive concern of independent consideration.

I am glad someone has at least fake-formalized the infectious nature of anxiety.

That's called modeling.

What makes this formalization fake?

Which axioms or assumptions do you disagree with?

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact