How We People
As things grow and expand, one must find and establish supports in the middle.
Points of commonalities and experiences people are arguing from - shared goals, shared fears, shared joys, shared traumas. Anything can become that point.
- psychological visibility
- Refers to that feeling that you get and you are interacting with someone and you feel like they get you. This is something that people crave.
- To reduce the stress of interacting with other people, simulating the best way to say the thing such that I will establish a stronger more positive relationship with the person I'm interacting with.
- One thing that I've noticed about interactions on social media: a lot of conflict arises when people treat each other as different types as how they want to be treated. Not out of malice, necessarily.
These words are being used interchangeably in this conversation
- Language/ Grammar/ Automaton
- If you are trying to operate with one type of automaton, and they don't respond in the pattern you expected, It's just not going to work.
- What type of steps will actually work with somebody?
communication between people
- players don't get to interact
Signalling Game introduced when players CAN actually communicate with each other
- Example: Interview generalizes this type of interaction
- Manager work with employee vs. Employee working with manager
- "Is this person I'm interacting with acting in good faith or not?"
A type of a person roughly corresponds to the preferences that they hold?
example of signalling game
- Real men vs. wimps
- Bully or defer based on their preference when they enter the bar
- Beer vs. eat quiche
- Do you order beer to be deferred to?
- Choosing signal that you do not prefer in order to not be bullied.
- "A game theorist that is married to someone who doesn't like games might play to lose so that they can play more games"
What's all the things that matter to you in priority order? Not something that we do.
Optimality theory - sounds that make up different languages can all be described by a single set of rules but ordered differently for different languages. This model translates to people, too, but instead of rules about sounds, it's preferences.
How is the person actually behaving in earnest (good faith, optimal space for everyone, not trying to screw them)?
- You must be an Active, local observer (cf. Pfitzmann/Hansen anonymity glossary)
- Not passively
- Send out signals that people respond to
- How they respond tells you more about them
As you learn things about the person that people signal off of them, you do not necessarily "build judgements" but "predictive models for how someone is likely to respond to something we present them with"
What does active observer look like?
- Active vs. Passive - does observer send out any signals?
- End goal: When trying to resolve a dispute that comes up. It's not for one side to win. Victory is not the goal. Success criterion
Satisficing: Make both people happy as possible - attempts to find the best possible situation for both. A lot of different preferences // let's find some threshold that people can agree on that satisfies as many preferences as possible for people to generally feel happy Different between compromise: Offer a solution for both.
"Trying to keep games from being zero sum"
Some people assume games are zero sum, if they don't get all the cookies they don't want to play at all
Being able to recognize when playing someone that is playing that game
It is hard to negotiate with someone who draws line in the sand and refuses to move from it
Or, get them to think on more than 1 axis
Signal universally that it's time to quit for someone else
Zero tolerance policy for circular arguments - once I spot the person isn't listening, I step out It depends on the person - Comes back to the trust thing
Signaling is hard on social media.
It's hard to make inferences on the fly with limited information.
How to generalize with really big social graphs?
What happens when the rules keep changing?
Set of moves available to each player of the game
Chaos that is possible
Information that is available
"Tell me something interesting" "Tell me something funny" "Send out signals that add to their information" "Maybe I'll continue to say different things"
With Twitter, the players change too.
Social Media can be compared to nature. It's nature's move.
You get to choose, "What platform, and what game you are playing?"
Is this someone you want to continue model building on?
Common value ground with?
Common Stakes ground?
What is it that one person is doing and how can I figure out if I can trust them? And what errors on how I can trust them?
- David Redish - "The mind within the brain" - neuroscience of learning and behavior
- Observe/ Decide/ Act: People's actions and preferences depends on context
- Herbert Simon
- Clay Shirky - "Here comes everybody"